ࡱ> tvsxhM &xbjbj== WWr lD$,,,$--|$Zf,p.////<0<0<0eeeeeee$h je-<080<0<0<0e>//4f>>><0( 8//e><0e>>FNAad/d. $#r$,d9b0d4*f0Zfcttk=tkd>$$An Inheritance Procedure for a Knowledge Representation Scheme Based on Fuzzy Petri Nets Slobodan Ribari, Nikola Paveai* Faculty of EE and Computing, University of Zagreb, Unska 3, 10000 Zagreb, Croatia  HYPERLINK "mailto:slobodan.ribaric@fer.hr" slobodan.ribaric@fer.hr *Faculty of Electrical Engineering, University of Ljubljana, Tr~aaka 25, 1000 Ljubljana, Slovenija  HYPERLINK "mailto:nikola.pavesic@fe.uni-lj.si" nikola.pavesic@fe.uni-lj.si ABSTRACT Knowledge about facts from the real world is very often uncertain, ambiguous and vague. As a result, the inference mechanism based on such facts generates uncertain and vague conclusions. In this paper we describe a formal model of knowledge representation for uncertain and vague domains. The model is based on the theory of high-level fuzzy Petri Nets (FPNs), and a fuzzy inheritance procedure that uses the dynamical properties of a FPN is proposed. The upper-time complexity of the proposed inheritance algorithm is O(nm), where n is the number of places (concepts) and m is the number of transitions (relations) in the scheme. An illustrative example of the inheritance procedure is presented. INTRODUCTION The crucial component of an intelligent (knowledge-based) agent is its knowledge base [1]. Informally, a knowledge base is an abstract representation of a working environment or a world in which the agent (or agents) has to solve tasks. It contains collections of facts and objects (in an abstract sense); spatial and temporal relations between facts or objects; descriptions of typical situations and the agents behaviour; the rules of the world; common-sense knowledge; heuristics and problem-solving methods and procedures; knowledge about states, the actions, motivations and goals of the agent; and knowledge about knowledge (meta-knowledge). In many real-world tasks knowledge is based on vague, uncertain and/or fuzzy information, and agents have to deal with the information in such a form. Therefore, in order to properly represent real-world knowledge and support vague, uncertain and fuzzy-knowledge representation, reasoning, learning and decision-making, different knowledge schemes were developed [1-9, 11]. Among these knowledge schemes there is a class of schemes based on the theory of Fuzzy Petri Nets (FPNs): Looney [4] and Chen et al. [5] proposed FPNs for rule-based decision making; Scarpelli et al. [6] described a reasoning algorithm for a high-level FPN; Chen [7] introduced a Weight FPN model for rule-based systems; Li and Lara-Rosano [8] proposed a model based on an Adaptive FPN, which is implemented for knowledge inference, but it also has a learning ability; Ha et al. [12] described knowledge representation by weighted fuzzy production rules and inference with generalized FPN; Ke [13] defined an inference mechanism for G-nets; Lee et al. [14] introduced a reasoning algorithm based on possibilistic Petri Nets as a mechanism that mimics human inference; Canales et al. [15] described a method of fuzzy-knowledge learning based on an Adaptive FPN; and Guo-Yan [16] proposed a hybrid of the PN and the Fuzzy PN to support an inference procedure for the natural extension of fuzzy expert systems. 2. A KNOWLEDGE REPRESENTATION SCHEME BASED ON FUZZY PETRI NETS 2.1. A Formal Definition The knowledge representation scheme named KRFPN (Knowledge Representation Scheme based on the Fuzzy Petri Net theory) is defined as 12-tuple: KRFPN = (P, T, I, O, M, (, (, f, c, (, (, C), (1) where P, T, I, O, M, (, (, f, and c are components of a generalized FPN, as follows: P = {p1, p2, ... , pn} is a finite set of places, T = {t1, t2, ... , tm} is a finite set of transitions, P ( T = (, I: T ( P( is an input function, a mapping from transitions to bags of places, O: T ( P( is an output function, a mapping from transitions to bags of places, M = (m1, m2, , mr(, 1 ( r < (, is a set of tokens, (: P ( ((M), is a mapping, from P to ((M), called a distribution of tokens, where ((M) denotes the power set of M. Using (0 we denote the initial distribution of tokens in the places of the FPN. (: P ( N is a marking, a mapping from places to non-negative integers, N. A mapping ( can be represented as an n-component vector ( = ((1, (2, ... , (n), where n is a cardinality of the set P. Obviously, ((pi) = (I, and ((pi) denotes the number of tokens in the place pi. An initial marking is denoted by the vector (0. f: T ( [0, 1] is an association function, a mapping from transitions to real values between zero and one. c: M ( [0, 1] is an association function, a mapping from tokens to real values between zero and one. The complete information about the token mi ( M is given by a pair (pj, c(mi)), where the first component specifies the location of the token, and the second one, its value. The bijective function (: P ( D maps a set of places onto a set of concepts D. The set of concepts D consists of the formal objects used for representing objects and facts from the agents world. The elements from D = D1 ( D2 ( D3 are as follows: elements that denote classes or categories of objects and represent higher levels of abstraction (D1), elements corresponding to individual objects as instances of the classes (D2) and those elements representing the intrinsic properties of the concepts or values of these properties (D3). The surjective function (: T ( ( associates a description of the relationship among facts and objects to every transition ti ( T; i = 1, 2, ... , m, where m is a cardinality of a set T. The set ( = (1 ( (2 ( (3 consists of elements corresponding to the relationships between the concepts used for partial ordering of the set of concepts ((1), the elements used to specify the types of properties to which values from subset D3 are assigned ((2), and the elements corresponding to the relationships between the concepts, but not used for hierarchical structuring ((3). For example, elements from (3 may be used to specify the spatial relations among the objects. The functions ( and ( give a semantic interpretation to the scheme. The semantic interpretation requires the introduction of a set of contradictions C that contains elements corresponding to relations from ( that mutually contradict each other, for example, is_a and is_not_a. Also, there are elements in the set D that are mutually contradictory if they are inherited for the same concept or object. For example, the object cannot simultaneously inherit properties such as "Quadruped" and "Biped". Both types of contradictions should be explicitly expressed in the KRFPN scheme. Formally, a set of contradictions C is a set of pairs of mutually contradictory relations and/or concepts: C = {((i, (j), ... , ((r, (s), (dk, dn), ... , (dl, dm)}, were (u; u = i, j, ..., r, s are from the set (, and dw; w = k, n, ...,l, m are from the set D. The inverse function (-1: D ( P, and the generalized inverse function (-1: ( ( (; ( ( T are defined in the KRFPN scheme. The KRFPN knowledge scheme can be graphically represented in a similar way to Petri nets: circles represent the places, while bars are used for the transitions. The relationships from places to transitions and from transitions to places are represented by directed arcs. Each arc is directed from an element of one set (P or T) to an element of another set (T or P). The relationships between the elements from P and T are specified by the input and output functions I and O, respectively. The tokens in the KRFPN are represented by labelled dots: c(mi) ( . Denoting the places by elements of D, the transitions by elements of ( and the values of the association function by f, the graphical representation of a knowledge base designed by the KRFPN is obtained. Tokens give dynamical properties to the KRFPN, and they are used to define its execution, i.e., by firing an enabled transition tj, tokens are removed from its input places (elements in I(tj)). Simultaneously, new tokens are created and distributed to its output places (elements of O(tj)). In the KRFPN, a transition tj is enabled if each of its input places has at least as many tokens in it as arcs from the place to the transition and if the values of the tokens c(ml), l = 1, 2, ... exceed a threshold value ( ( [0, 1]. The number of tokens at the input and output places of the fired transition is changed in accordance with the basic definition of the original PN [10]. The new token value in the output place is obtained as c(ml).f(ti), where c(ml) is the value of the token at the input place pj ( I(ti) and f(ti) is the degree of truth of the relation assigned to the transition ti ( T. Figure 1 illustrates the firing of the enabled transition of the KRFPN. The inheritance, as an inference procedure defined for the KRFPN, uses the dynamical properties of the KRFPN.  EMBED CorelDRAW.Graphic.10  a) Before firing; c(ml) ( ( b) After firing; c(mp) = c(ml).f(ti) Fig. 1. Firing an enabled transition. 2.2. Selection of values f(ti) and ( A humans knowledge about facts from the real word is very often uncertain, ambiguous and vague. The inference mechanism based on such facts, as well as the human interpretation of such facts and conclusions, generates uncertain and vague conclusions. There are many different approaches to representing knowledge in uncertain domains, such as belief networks (Bayesian or probabilistic networks) [1], qualitative probabilistic networks, rule-based methods for uncertain reasoning, schemes based on the Dempster-Shafer theory [19], fuzzy logic and the fuzzy Petri net theory. In our approach we have used the last of these, in which the uncertainty and confidence related to the facts, concepts and the relationships between them are expressed by means of the values of f(ti), ti( T, and c(mi), mi (M, association functions. For example, the value of the function f, as well as the value of the function c, can be expressed by truth scales and by their corresponding numerical intervals depicted in Table I [5]. The threshold value ( ( [0, 1] defines the sensitivity of the scheme during the inference procedures. By using different values for (, the user can specify different degrees of truth for inherited concepts properties. Table I TRUTH SCALES AND THE CORRESPONDING NUMERICAL INTERVALS [5] Truth scalesNumerical intervalsAlways true[1.0, 1.0]Extremely true[0.95, 0.99]Very true[0.80, 0.94]Considerably true[0.65, 0.93]Moderately true[0.45, 0.64]More or less true[0.30, 0.63]Slightly true[0.10, 0.29]Minimally true[0.01, 0.09]Not true[0.0, 0.0] Example 1 In order to illustrate the basic components of the KRFPN, a simple example of the agents knowledge base for a scene (Fig. 2) is introduced. Shortly, the scene may be described as follows: Fred, who is a human, is in front of the elephant Clyde. Both are mammals, but Fred is biped, while Clyde is a quadruped. The knowledge base designed by the KRFPN = (P, T, I, O, M, (, (, f, c, (, (, C), has the following components (Figure 3): P = {p1, p2, ... , p10}; T = {t1, t2, ... , t13}; I (t1) = {p1}; I(t2) = {p3}, ... ; I(t13) = {p1}; O(t1) = {p2}; O(t2) = {p4}, ... ; and O(t13) = {p9}. The set of tokens is M = (m1, m2, , mr(, the initial distribution of tokens is (0 = {{m1}, (, ..., (}, where c(m1) = 1.0, and ( denotes an empty set. The vector (0 = ((1, (2, ... , (10) = (1, 0, ... , 0) denotes that there is only one token in the place p1. The function f is specified as follows: f(t1) = 0.9; f(t2) = 0.9; f(t3) = 1.0; f(t4) = 1.0, ...;  Fig. 2. A simple scene with Fred and the elephant Clyde. The set of tokens is M = (m1, m2, , mr(, the initial distribution of tokens is (0 = {{m1}, (, ..., (}, where c(m1) = 1.0, and ( denotes an empty set. The vector (0 = ((1, (2, ... , (10) = (1, 0, ... , 0) denotes that there is only one token in the place p1. The function f is specified as follows: f(t1) = 0.9; f(t2) = 0.9; f(t3) = 1.0; f(t4) = 1.0, ...; f(t12) = 0.6; and f(t13) = 0.8 (Fig. 3). f(ti), i = 1, 2, ... , m indicates the degree of our pursuance in the truth of the relation ((ti). The set D = D1 ( D2 ( D3 is defined as follows: D1 = {Elephant, Human, Mammal, Biped, Quadruped}, D2 = {Fred, Clyde}, D3 = {White, Soul, Kind_hearted}. The set ( = (1 ( (2 ( (3 is {is_a, is_not_a} ( {has_colour, has_characteristic, has} ({is_in_front_of, is_behind}. Functions (: P ( D and (: T ( ( are (Fig. 3): (: p1 ( Fred, (: t1 (is_a, (: p2 ( Human, (: t2 ( is_a, ... ... ... ... (: p10 ( Kind_hearted, (: t13 ( has. Let C be a set of contradictions defined as C = {(is_a, is_not_a), (is_in_front, is_behind),(Quadruped, Biped)}. For the initial distribution of tokens, (0, the following transitions are enabled: t1, t9, t11 and t13. 3. FUZZY INHERITANCE In general, inheritance is a form of reasoning that allows an agent to infer the properties of a concept on the basis of the properties of its ancestors in the networks hierarchical structure [17]. More precisely, the inheritance can be described as the process of determining the properties of a concept di ( D, by looking up the properties that are locally attached to the  EMBED CorelDRAW.Graphic.10  Fig. 3. The agents knowledge base designed by the KRFPN. concept di. If such local information is not available (or is insufficient), the process will continue by looking up properties attached to the concepts that lie at higher levels in the conceptual hierarchy. The inheritance procedure in the KRFPN is based on its dynamical properties and the determination of the inheritance set of the KRFPN. The inheritance set for the KRFPN is based on concepts similar to the reachability set of the ordinary Petri nets (PNs), where the reachability relationship is the reflexive, transitive closure of the immediately reachable relationship [10]. The reachability set of the PN is graphically represented by a reachability tree. The main differences between the inheritance set of the KRFPN and the reachability set of the PN [10] are as follows: (i) After firing all the enabled transitions for the distribution of tokens in the KRFPN, where the transitions are related to the elements in the subsets (2 and (3, the created tokens at the corresponding output places have to be frozen. Recall that the elements in (2 and (3 are used to specify the properties and the non-hierarchical structuring, respectively. A frozen token in the output place is fixed and it cannot enable a transition; (ii) An inheritance tree, as a graphical representation of the inheritance set, is bounded by k + 1 levels, where k is a predefined number of levels. Such an inheritance tree is called a k-level inheritance tree; (iii) A k-level inheritance tree has the following additional types of nodes: a k-terminal node, a frozen node, and an identical node. A k-level inheritance tree consists of nodes (pj, c(mi)), j = 1, 2, ... , n and i = 1, 2, ... , v where n is a cardinality of the set of places P, and 0 ( v ( r, where r is a cardinality of the set of tokens M, and directed, labelled arcs. In order to simplify and make the notation uniform in the inheritance algorithm, the nodes in the tree are denoted by n-component vectors in the form ( = ((1, (2, , (n). Each component (i ; i = 1, 2, ..., n of ( is represented by an empty set ( for ((pi) = 0, i.e., there is no token(s) at the place pi, or by a set {c(mk), ... , c(ml), ... , c(ms)}, where c(ml) is the second component of the pair (pi, c(ml)) and represents the value of the token ml at the place pi. The number of elements in the above set is ((pi) (1. The vector ( is called the distribution vector, or simply the distribution. The directed arcs in a k-level inheritance tree are labelled by tj ( T and f(tj), where tj denotes a fired transition and f(tj) the value of its association function f, respectively. A directed labelled arc leads from the node at the i-th level of the inheritance tree to the corresponding node at the i+1-th level; i = 1,2, ... , k-1, where k is (pre)defined as the final level of the inheritance tree. During the creation process of the k-level inheritance tree each node is classified either as a frontier node, a terminal node ( denoted by T), a k- terminal node (k-T), a frozen node (F), an identical node (I), a duplicate node (D), or an interior node. Taking into account above particularities, the inheritance tree can be constructed by applying the slightly modified algorithm for reachability tree given in [10]. The 3-level inheritance tree for the KRFPN scheme, where the token m1 is initially at the place p1, generated by the algorithm is shown in Fig. 4. By using the components of the k-level inheritance tree: a node at the level i - 1, labelled arc, a node at the level i (successor of the node at level i - 1), the functions ( and (, and a triplet named an inheritance assertion is formed. For example, a node (11 at the level 1, a labelled arc t4; f(t4) = 1.0, and the successor node (21 at the level 2, and ((p2) = Human, ((t4) = is_a, and ((p5) = Mammal, define the inheritance assertion: (Human is_a Mammal). Note that the tokens are at the places p2 (for (11) and p5 (for (21). The strength of the assertion is defined as the value of the token at the successor node, i.e., as a product of the token value at the node at level i - 1 and the value of the association function of the corresponding transition.  EMBED CorelDRAW.Graphic.10  Fig. 4. The 3-level inheritance tree for the KRFPN scheme depicted in Fig. 3. For example, the strength of the inheritance assertion (Human is_a Mammal) is 0.90 . 1.0, where 0.90 is the value of the token at place p2 (see the distribution (11; Figure 6.). The inheritance paths, starting from the root node of the inheritance tree and finishing at the leaves of the tree (terminal node, k-terminal node, frozen node, identical node, duplicate node), represent sequences of the inheritance assertions. An inheritance path is interpreted as the conjunction of the inheritance assertions in which the redundant concepts connected by AND are omitted. The strength of an inheritance path is defined by the value of the token at the node that is a leaf of the inheritance tree. For example, the inheritance path defined from the root node (0 to the leaf node of the tree (22 (see Fig. 4) is: Fred is_a Human AND is_a Biped. The strength of the inheritance path is 0.72. In network-based knowledge representation schemes there is the well-known problem of the conflicting multiple inheritance [18]. The problem of the conflicting multiple inheritance in the KRFPN scheme is expressed as follows: Two inheritance paths, represented by sequences of the inheritance assertion, are in conflict if the same concept inherits the mutually contradictory elements from D, i.e., dk and dn, where (dk, dn)( C, C is a set of contradictions. Also, two inheritance paths are in conflict if the same concept inherits the concept/property ((pk) ( D, but in one inheritance path it inherits the concept/property by means of ((tr) ( (, and in another by means of ((ts) ( (, where ((tr) = (r, ((ts) = (s, and ((r, (s) ( C. To resolve the situations involving conflicting multiple inheritance in the KRFPN scheme we used Touretzky's principle of inferential distance ordering [18]: In the situation of conflicting multiple inheritance, a concept inherits the property of the nearer individual (concept) or class. Nearer is defined as follows: Concept or class A is nearer to class B than to class C if A has an inheritance path through B to C. In many cases the inferential distance ordering fails and reports an ambiguity (for example, see the well-known Quaker's problem [18]) because it is based on a measure of between-ness. In such situations, in the KRFPN scheme, the concept inherits the property for which one can find the more direct inheritance path, i.e., the shorter path. If two or more inheritance paths have the same length the concept inherits property that corresponds to more strength in the inheritance path. 4. INHERITANCE ALGORITHM The inheritance algorithm for the KRFPN is presented as follows: Input: A concept of interest di ( D, the depth of the inheritance k; 0( k <(, and (. Output: The properties of the concept di ( D, obtained by looking up the properties locally attached to the concept di, and the properties attached to the concepts that lie at higher levels in the conceptual hierarchy. The properties are expressed by means of semantic interpretations of the inheritance paths. Step 1. For a given concept of interest di ( D, by using the inverse function (-1, find the corresponding place pk: (-1: di ( pk. If di ( D stop the algorithm and send the message: di is an unknown concept. Step 2. Define the initial marking (0 = ((1, (2, ... , (n), where for j = 1, 2, ... , n  EMBED Equation.3  (j = EMBED Equation.3  EMBED Equation.3  Step 3. Define the initial distribution of tokens (0 = (0 = ( (, (, (, ... {(pj, c(m1)}, ... , (, () and set c(m1) = 1.0; Step 4. For the initial distribution of tokens (0 = (0 construct k levels of the inheritance tree. Step 5. By using the elements of the inheritance tree: a node at level i 1, a labelled arc, a corresponding node at level i, and the functions ( and (, form the inheritance assertions. Step 6. Find the inheritance paths, starting from the root node of the inheritance tree and finishing at the leaves of the inheritance tree. Determine the strengths of each inheritance path. Step 7. If there are situations involving conflict due to the conflicting multiple inheritance use the inferential distance ordering concept or (if that fails) make a decision on the basis of the more direct inheritance path. If two or more inheritance paths have the same length the concept inherits the property that corresponds to the stronger path. On the basis of the above criteria, remove the inheritance assertion that is the source of conflict and all the inheritance assertions that follow it. Example 2 For the knowledge base represented in Fig. 3. infer the properties of the concept Fred (with the depth of inheritance k = 3) and ( = 0.1. A set of contradictions C = {(is_a, is_not_a), (is_in_front, is_behind_of), (Biped, Quadruped)}. Input: Fred (D, k = 3, ( = 0.1. Step 1. (-1: Fred ( p1 Step 2. The initial marking is (0 = (1, 0, 0, 0, 0, 0, 0, 0) Step 3. The initial distribution of tokens is (0 = (0 = ({(p1, c(m1)}, (, (, (, ..., ... , (, () and c(m1) = 1.0; Step 4. The inheritance tree is shown in Fig. 6. Step 5. The inheritance assertions are (see Fig. 6): Level 1: Fred is_a Human; (((p1) ((t1) ((p2)); (strength = 0.90) Fred is_in_front Clyde; (((p1) ((t9) ((p3)); (strength = 0.90) Fred has_characteristic Kind_hearted; (((p1) ((t11) ((p10)); (strength = 0.50) Fred has Soul; (((p1) ((t13) ((p9)); (strength = 0.80) Level 2: Human is_a Mammal; (((p2) ((t4) ((p5)); (strength = 0.9) Human is_a Biped; (((p2) ((t5) ((p6)); (strength = 0.72) Human is_not_a Quadruped; (((p2) ((t6) ((p7)); (strength = 0.9) Level 3: Mammal is_a Quadruped; (((p5) ((t7) ((p7)); (strength = 0.72) Step 6. The inheritance paths and their interpretations are: (i) Fred is_a Human AND is_a Mammal AND is_a Quadruped; (0.72) (ii) Fred is_a Human AND is_a Biped; (0.72) (iii) Fred is_a Human AND is_not_a Quadruped;(0.90) (iv) Fred is_in_front Clyde; (0.90) (v) Fred has a Soul; (0.80) (vi) Fred has_characteristic Kind_hearted; (0.50) Step 7. According to the elements of the set of contradictions C and the definition of the contradiction, the inheritance paths (i) and (ii) generate contradictions because (Biped, Quadruped) ( C. The paths (i) and (iii) introduce a conflicting multiple inheritance due to (is_a, is_not_a) ( C. There is an ambiguity about the above inheritance paths: Is Fred a quadruped or a not? To resolve the above ambiguity we can determine whether the inheritance path (iii) is more direct than the path (i). The inheritance path (i) is a path from the concept Fred through the concepts Human and Mammal to the concept Quadruped, simultaneously there is the inheritance sub-path Human is_a_not Quadruped. The result of resolving the conflict situation on the basis of the more direct inheritance path is: Fred is_a Human AND is_a Mammal; (0.90) Fred is_a Human AND is_a Biped; (0.72) Fred is_a Human AND is_not_a Quadruped; (0.90) Fred is_in_front Clyde;(0.90) Fred has Soul; (0.80) Fred has_characteristic Kind_hearted; (0.50). Therefore, the inheritance process resulted in the following statements: Fred is a human and a biped, he has a soul, and he is kind hearted. The strength attached to each inheritance path has to be interpreted as the confidence of the statement. With reference to Table I, for instance, the statement: Fred is_a Human AND is_a Mammal; (0.90) is interpreted as very true, while the statement Fred has_characteristic Kind_hearted; (0.50) is interpreted as moderately true. 5. CONCLUSION An original fuzzy inheritance procedure for the knowledge-representation scheme KRFPN based on the Fuzzy Petri Net (FPN) theory is proposed. The inheritance uses a k-level inheritance tree, which is generated on the basis of the dynamical properties of the FPN, i.e., on firing-enabled transitions and changing the values of the tokens according to the association function f that specifies the degree of assurance for the relation assigned to the transitions. The very important properties of the inheritance algorithm are that the k-level inheritance tree is finite and that the upper bound of the time complexity of the fuzzy inheritance algorithm is O(nm), where n is the number of places and m is the number of transitions. These allow the efficient execution of the inheritance procedure. Adaptation of the knowledge-representation scheme KRFPN for time-varying fuzzy knowledge and spatio-temporal reasoning are our future directions. 6. REFERENCES [1] S. Russel, P. Norvig, Artificial Intelligence, A Modern Approach, Prentice Hall, Upper Saddle River, 1995. [2] C. G. Looney, L. R. Liang, Inference via Fuzzy Belief Petri Nets, Proceed. of the 15th Int. Conf. on Tools with Artificial Intelligence (ICTAI '03), pp. 510-514, Nov. 2003, [3] J. Luo, H. Tu, K. Pattipati, L. Qiao and S. Chigusa, Graphical Models for Diagnosis Knowledge Representation and Inference, IEEE Proceed. on the Autotestcon, pp. 483-489, Sept. 2005. [4] C. G. Looney, Fuzzy Petri Nets for Rule-based Decisionmaking, IEEE Trans. on the System, Man and Cybernetics, Vol. 18, No. 1, pp. 178-183, 1988. [5] S-M. Chen, J-S. Ke and J-F. Chang, Knowledge Representation Using Fuzzy Petri Nets, IEEE Trans. on Knowledge and Data Engineering, Vol. 2, No. 3, pp. 311-319, Sept. 1990. [6] H. Scarpelli, F. Gomide and R. R. Yager, A Reasoning Algorithm for High Level Fuzzy Petri Nets, IEEE Trans. on Fuzzy Systems, Vol. 4, No. 3. pp. 282-294, 1996. [7] S-M. Chen, Weighted Fuzzy Reasoning Using Weighted Fuzzy Petri Nets, IEEE Trans. on Knowledge and Data Engineering, Vol. 14. No. 2. pp. 386-397, 2002. [8] X. Li, F. Lara-Rosano, Adaptive Fuzzy Petri Nets for Dynamic Knowledge Representation and Inference, Expert Systems with Applications, 19, pp. 235-241, 2000. [9] B. C. Avila, F. Hembecker, Inconsistencies, Exceptions and Frame Systems, Proceed. of the XXI Int. Conf. of the Chilean Computer Science Society, SCCC 2001, pp. 23-32, 2001. [10] J L. Peterson, Petri Net Theory and Modeling of Systems, Prentice-Hall, Englewood Cliffs, 1981. [11] T. Hong, H. Rybinski, A New Approach to Computing Weighted Attributes Values in Incomplete Information Systems, Proceed. of the Int. Conf. on the Hybrid Information Technology, ICHIT'06, pp. 268-273. Nov. 2006. [12] M-H. Ha, J. Li, H-J. Li and P. Wang, A New Form of Knowledge Representation and Reasoning, Proc. of the Fourth Int. Conf. on Machine Learning and Cybernetics, Guangzhou, pp. 2577-2582, 2005. [13] Y-L. Ke, Distribution Feeder Reconfiguration for Load Balancing and Service Restoration by Using G-Nets Inference Mechanism, IEEE Trans. on the Power Delivery, Vol. 19, No. 3, pp.1426-1433, 2004. [14] J. Lee, K. F. R. Liu and W. Chaing, Model Uncertainty Reasoning With Possibilistic Petri Nets, IEEE Trans. on Systems, Man and CyberneticsPart B: Cybernetics, Vol.33, No. 2, pp. 214-224, April 2003. [15] J.C. Canales, X. Li, W. Yu, Fuzzy Knowledge Learning via Adaptive Fuzzy Petri Net with Triangular Function Model, Intelligent Control and Automation, The Sixth World Congress on WCICA 2006. Volume 1, pp. 4249 4253, June 2006. [16] H. Guo-Yan, Analysis of Artificial Intelligence Based Petri Net Approach to Intelligent Integration of Design, 2006 International Conference on Machine Learning and Cybernetics, pp. 1691 1695, Aug. 2006, [17] L. Shastri, Semantic Networks: An Evidential Formalization and Its Connectionist Realization, Pitman, London, 1988. [18] D. S. Touretzky, The Mathematics of Inheritance Systems, Pitman, London, 1986. [19] R. R. Yager, M. Federizzi and J. Kacprzyk (eds.), Advances in the Dempster-Shafer Theory of Evidence, J. Wiley, 1994.     (*TVX   ? I %&HINOTU¾ jWOJQJmH sH 5OJQJmH sH OJQJmH sH 5OJQJ\mH sH  CJOJQJCJOJQJmH sH 5\mH sH 6]5jCJU0JZCJjCJU jCJUCJ 6CJ]5CJ 5CJ\1^, +  $`a$ $7$8$H$`a$$`a$$ & Fa$$a$$ a$$a$x#x%xxAz#X^. Q!u###f$$`a$]`]]$xxa$\$xa$ !"#GIKLVW|}'()*-.56댂v j{OJQJmH sH H*OJQJmH sH  jH*OJQJmH sH  jOJQJmH sH  jOJQJmH sH  jOJQJmH sH H*OJQJmH sH  jWOJQJmH sH  jbOJQJmH sH  jaOJQJmH sH OJQJmH sH  jmOJQJmH sH ,67;<ABXY]^_`}~!"pqǥ~뙥뙥뙥뙥 jm5OJQJ\mH sH 6OJQJ]mH sH  jmOJQJmH sH H*OJQJmH sH  jOJQJmH sH  jOJQJmH sH  jWOJQJmH sH  jOJQJmH sH  jOJQJmH sH OJQJmH sH  j}OJQJmH sH 1*+YZ[\cdXYZ[styz67쏇w jSmH sH  jmH sH  jbmH sH mH sH  jOJQJmH sH  jaOJQJmH sH  jOJQJmH sH  jOJQJmH sH 5OJQJ\mH sH 5H*OJQJ\mH sH  jm5OJQJ\mH sH OJQJmH sH H*OJQJmH sH .qrstHIJ)*+IJKZ[gqw~A B C E F H* js6OJQJ]mH sH  jSOJQJmH sH  CJmH sH OJQJmH sH  jbmH sH  jamH sH  jmH sH  jSmH sH 6]mH sH  jmH sH  H*mH sH mH sH ;F G P Q R S T V W X ] ^ a b m n q r { | }  !! !"!#!$!%!&!'!)!*!+!,!P!Q!~##OJQJaJmH sH OJQJmH sH  jmH sH  jtmH sH  jSmH sH  jbmH sH  jmH sH  H*mH sH  jamH sH mH sH  jS js6]H*;#######e$f$$$$$#%$%%%%%%%&&<&=&@&A&g&h&i&j&E'F'G'H'K'L'X'Y'''''''''''''z((((սսսսշկթաՙՐՊսՐսՙսսսՙ H*mH sH 6H*]mH sH  jmH sH  jlmH sH  6mH sH 6H*mH sH  ]mH sH  H*mH sH 6]mH sH \ jSmH sH mH sH  jOJQJaJmH sH OJQJaJmH sH H*OJQJaJmH sH 6f$((()*)+),)Q)R)$.%.-.h.u..$xx$x&`#$/Ifa$b$$h`ha$$a$$a$` ]$@ ^@ `a$]$a$`((((((((((((((((((())) ) )*),)H)I)O)P)R)W,X,\,],{njej^j\\H* jl5\ 5H*\5\B*OJQJmH phsH 5OJQJ\mH sH H*OJQJmH sH H*OJQJmH sH  jlOJQJmH sH  j>OJQJmH sH 6H*OJQJ]mH sH OJQJmH sH mH sH jCJUmH sH 'jII OJQJUVmHnHsHtH CJmH sH jCJUmH sH "],^,i,j,n,o,p,q,[-\-]-^---h..///001111 1111F1G1J1K1T1V1_1`1c1d1m1o1v1w1}1~111111111111111111111111111 j{mH sH  H*mH sH  jbmH sH  jamH sH  jmmH sH  jWmH sH B*mH phsH mH sH 5\CJ\ jlH* jF....~`ll$$x&`#$/Ifa$b$$$Ifl0, q E  t 6`x0644 lae4....~tll$$x&`#$/Ifa$b$$$Ifl0, q E  t 6`x0644 lae4........///'/4/~`ll~ll~xll~ll$$x&`#$/Ifa$b$$$Ifl0, q E t 6`x0644 lae4 4/5/C/P/Q/`/m/n/w//~pll~tll~Tll$$x&`#$/Ifa$b$$$Ifl0, q E t 6`x0644 lae4 /////A3B3D3E3~33s56~xxxrpkkpkrr$a$]]``$$Ifl0, q E  t 6`x0644 lae4 1222*2+2,222326272>2?2K2L2Z2[2}2~2222222222222 3 333#3$30313@3A3B3C3E3K3M3}3~333B*OJQJmH phsH B*OJQJmH phsH OJQJmH sH 5OJQJ\mH sH jUmH sH CJ jmmH sH 5H*\mH sH  jm5\mH sH  jmH sH  jWmH sH  j}mH sH  H*mH sH mH sH 13333333333333333333344#4$4%4)4*4+4-4.4/47484:44444444444444455k5l5n5o55뮡 jbOJQJmH sH  jmOJQJmH sH 5H*OJQJ\mH sH  jm5OJQJ\mH sH  jOJQJmH sH  jWOJQJmH sH  j}OJQJmH sH H*OJQJmH sH OJQJmH sH  j{OJQJmH sH 655555555555555555555566666666666 6!6"6#6(6-6.6668696;6^6`6a6b6{666666666666666666 jbOJQJmH sH  jOJQJmH sH  jaOJQJmH sH  jSOJQJmH sH 6OJQJ]mH sH  jOJQJmH sH OJQJmH sH H*OJQJmH sH @666666&77788899999:<@^CD$^`a$ $^`a$$`a$]$a$$a$$a$]``]]`666666666666666666666666666777777 7777777 7!7$7&7X7f7j777777777777777778 jWOJQJmH sH 6]mHsH jaOJQJmH sH  jOJQJmH sH H*OJQJmH sH  jbOJQJmH sH OJQJmH sH 6OJQJ]mH sH ?888 88I9J9K9L999999999999999u<<======== > > >>>>??+?,?t?u?????@@@D@E@J@K@_@`@e@f@u@v@}@~@µCJH* 6CJ]CJ jS6]B*OJQJmH phsH OJQJmH sH 5OJQJ\mH sH  jzUjՁI UV jUmHsH jH*5\5\mHsH?~@@@@@@@@@@@@{A|AAAAAAAAAAAAAAAAAAAAAAABBBB5B6BHBIBUBVBbBcBpBqBBBBBBBBBC C C CCCCCCPC]CuCvCCC jCJ 6CJH*] jmCJ jCJCJH* jpCJ 5CJ\ jp5CJ\ CJmH sH  6CJ] jCJCJHCCCCCCCCCC D(DHDLDDDDDDDDDEEEEEEFFFF*GIGJGwGxGGGGGGGGGG H-H.H0H1HQHRHWHXHyHzH|HHHHHHHHHHHHHHHHHHHHH*5\ 5H*\ jp5\ jb ja6] B*CJph 6CJ] CJmH sH  jCJCJH*CJLDEF*GHAI'JIJJJJKKOMNNPzT{TTT$a$$a$$`a$ $d`a$$a$$`a$$`a$ $`a$$a$$^`a$HHHH!I"I(I)I+I2I3I9I:II?III'J(JEJFJGJHJIJPJJJJJJJ"K#K:K;K=KKKaKKKMMMMMMMMMMMMMMOOOOOOOOOO7P8P:P;P=P>PPPPPPPP jb ja jCJ H*5\ jUj I UV jU 5H*\ jp5\H*6]MPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPRRRRSSNSOS{TTTTTTTUUUUU!U"U(U)U+U3URUSUTUUUUUaVbVhViVVVVVVVVVVVVVVH* ja jl j j5\ B*phf6] js jH* jb jSRTT+UbVVV8WWWWXXXwY6ZS[0\1\;\<\\(]H]a]]^B^w^$`a$$a$$a$VVVVVVVVVWW8W?W[W\W]WaWbWcWeWfWgWpWqWrWWWWWWWWWWWWWWWWWWWWWXXXXXXXXXX X"X#X,X-X2XŻ j 5H*\ jp5\jn5EHU\jhG UVj 5EHU\jYG UVj5U\ jEHUjwG UV jU jm jW5\ j jH*<2X3X=X>X@XAXNXOXXX^X_X`XXXXXXXXXXXXXYY7Y8YLYMYRYSYwY~Y6Z=Z>Z1\;\i\j\\\\\\\\\\\\]]]]]]$](].]3]4]5]8]:]?]@]H]Q]R]S]U]W][]\]H* j jl B*ph jb ja 5H*\ jp5\ jW6] 56\]5\ jH*K\]]]_]a]h]]]]]]]]]]]]]]]]]]]]]]]]]]]^^^^^^B^I^J^w^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^!_"_#_%_&_(_)_+_ jb jaCJ6] 56\] j jp5\ jW 5H*\ jm5\5\H* jNw^^^^K____`B`L``` a5aiaaaa\cdd!e$xa$ ^`$`a$ $r^`ra$ ^`$a$^+_-_/_0_2_4_K_[_\_]___`_b_c_e_g_i_j_l_m_____________________________````` `!`#`$`&`'`)`*`,`-`L`c`d`e`g`h`j`k`m`n`p`q`s`t`````````aa a aa"a-a4a5\ B*phf jb6] jaH*\4a5a:aKaOabamaaaaaaaaaaabbbbbbbbbbbc@cDcGcPcd dd%d)d/d?dHd{ddddd e ee e1e5e@eGeXe\eoeveeeeeeee#f(f-f3f9f?fIfOfYfff ggg)gCgNgbggggggg5\mH sH  B*CJphCJ j5\6]W!eHeweeeehfggggpkqkkkklbmmnSoo'p\p `^`$da$]$a$$`a$ $x`a$$xa$$a$gghihiijjQjVj^j_j|j}jpkqkkllmmmmn7o8o>o?oQoSoKqqrQsvwwxxxxxxx x%x&x jU5B*CJOJQJphB*CJOJQJph CJOJQJ B*ph B*aJ ph5aJmH sH 5\6OJQJ]mH sH B*OJQJph6OJQJmH sH OJQJmH sH -\ppKqqrQsttuv*wwwwwwwwxxxxxxxxx x$a$ `^`$a$ x x x x xxxxxxxxxxxxxxxx!x"x#x$x%x&x$a$ &P . A!8"}#$ %+ 0&P . A!8"}#$ % P ' 0&P . A!8"}#$ % + 0&P . A!8"}#$ % P DyK slobodan.ribaric@fer.hryK >mailto:slobodan.ribaric@fer.hrDyK nikola.pavesic@fe.uni-lj.siyK Fmailto:nikola.pavesic@fe.uni-lj.si Dd: >@J  C A? "2* IcCQ3 ~`! IcCQ3"` xڽX pUy\TBf$ [|Œ&( 3ϣO$*@Ci *SuAJ D眽Zϻq6^ (ŮU>ʯ֣ȵ?xˊת@X먉 $.Pgϓ28$P.GP!%P(DS֡ 6 诔R`\|]ryr`c00}AHh@`8㐌 t2#j86'O NJ,H Kxfe$FԈZ`b(l*qdKCrd.Y$%L*e\"zimV cl:93)EnIb,{l^џFi rj-c刴򀜗h(Mx4GcC#ZXT4| >A39'-L☭2<X0v(J*e 9h+KQ I֡lDd,C3 tr#pj Z`I>R yfϗs38OcO #nX-stV#o˗/_՝/WU;57;/ްizG҇ғ;krrirY%]sg)\nvV:p˥'y{Miyy5MKlf;o_7M҅u1bc6c4֞1aj>bcާzu;>&&ƗMlrY ֗͜s[s\\W1}%Y\˳֟dj%S1{ͽj>:\yI€xFF;N^tZў  ?1$t =36v%VX[R;~ nTWpoօn[g/%XZ|pa[EȤ?-4ƥ_2430<]w J9(wഛC;$@TQYw:]Olex 7뢔=)#-=^-29nTArGS#jY\7[k$}[q/47,\>#UI\^۬1خel49|b.9yRln9˧^9l gT9٢u?9ے 51C眔y\xωC.j'kh΃sdsN 2)rJ1Tlq`3%81g>u6F rM'8r@ex JXHWk]msUN *"w [wO_Wbx{I/>7*Mܓ^hV}*UkOsE W_8zY ըWԎER>B՝=V#0Ԍ5[F:`cNlGsm>coG|ݼO__,yJn?5LMi~/-^ ,nZXڎm׃ {8.:y=nc]OT6a]lV98.Fރ6(?M>/ե(' }~b0ZsOEGaD= B?xtÃ~&9Vlg=ccш\_)1)?*?(}?,WHG$!r7F2<.JUNrD~%[J+XEK^*E:SJj ۭ1Y1BL29O6$NoD?\ϒs"_2<J}JR\ 5B1!߇㟴oCq 5<.2<Ch cb( =A=Mt: mtZMhsҙGtr'j Z`Id$,sYƽxžC\"{$NZ׫oO[;~ӆu?dx[{I?պ;?̸#s?ӟA|2|O;iú7ޢeb~ug~=h񎵞X{!KTӺrȵ;^';C?LGoc8. YD  rζs:5M&s`39y6s71o|yakџ11 C (PzꆂPKZ ح-FLfR끙( 3-i)0ccw;z'0a;I+ֻ#׳+CH!Z}Xu0b=!UaڄjH?I*G: Π; cTf0fbJR5 @dȢQJ|<*oy21MX-cT{Ɛ.lU)3)J:(>,U.o BnN8ex v*Ur$VeIZ)!5XN[$Z-TګLIR%!jVۭ1دr[)^|:obPxeO{% {[Vlw̧>U)cK2k^ ^wZR0ܻ d2<70Ko*Ҽ腑ea&Lr0ho9#1j]m*oy;b(yksAwE߫R|o2J^,k"y+!ǃ u3!vHSDCH0Z_upb$u/P]ˑbԟ__)f{79s~DV6Ԇ߫7%W7%Pojd{^7pEV fXIlDdLx  C .Afred&clydeS"`b}k$HcLۧYk~nQk$HcLۧPNG  IHDR^D0PLTEde萊nnp;=KJKڎʔǸ100gR   !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijlmnopqriuz{|}~Root Entry2 F;rwData kժWordDocument1ObjectPool4Er;r_1234803529"3H觜Er rOle ObjInfoCorelDRAWya  !"#$%&).1259:;<=>?ABCDEFGHIJKLNOh+'Oh+'0  ( 4 @ LX`hpfeRIFFaCDRAvrsnLISTNdoc pfrdmcfg~ Q-p8WF]tE?F]tE??8888 i' LastUsed/6a.?AVCLastUsed/6a.?AVCLastUsed/6a.?AVCdd@@@@?}4V?}4V?C:\Corel\WebSite\LastUsed/6a.?AVCDocTemplate@@/6a.?AVCMultiDocTemplate@@/6a.?AVWDrawMultiDocTemplate@@Initializing the App Recovery systemPLGT3280.DLLcoreldrw.exeRAVE.exedrawintl.dll.tlbDrawBrowser\Rave@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va(@ptrte@LLIST$fnttfontX@ ArialXwcXWesternfontX%@Arial Unicode MSXCHINESE_BIG5fontX@Arabic TransparentArabicLISTarrtarrw4`V>  DDH" "arrw4V>  DDH" "arrw4ܚV>  DDH" "arrw4V>  DDH" "LISTXfiltLIST0filcfild$4dX|@<LISTfilcfildLISTotltoutlXd??d@<outlÙd??d@<outlęd??d@<`outlhřd??d@<outl0d??d@<outl@ƙd??d@<ܚoutlǙd??d@<LISTkstlt8Ϟ4`4c50@X%iii@@@ZC@B@B@B' -`c9 V? c5  ncdH5`ՒDefault Graphic`4@8Default Artistic Text8Ϟc5ZC`H588Ϟc5ZC`H5`88Ϟc5ZC`H588Ϟc5ZC`H5M88Ϟc5ZC`H5諗88Ϟc5ZC`H588Ϟc5ZC`H5(Ւ`Ւ`4@Ԓ`Ւ`4@Ԓ`Ւ`4@Ԓ`Ւ`4@Ӓ`Ւ`4@XӒ`Ւ`4@88Ϟc5ZC`H5088Ϟc5ZC`H588Ϟc5ZC`H5 Ӓ`Ւ`4@Ғ`Ւ`4@Ғ`Ւ`4@xҒ`Ւ`4@@Ғ`Ւ`4@Ғ`Ւ`4@LISTuil LISTbpageflgsbbox@_j( obbx @_j( @_j( LISTgobjLISTllayrflgs LISTTlgoblodaGG$ 04@G.GuidesLISTllayrflgsLISTTlgoblodaHH$04@H.dDesktopLISTllayrflgsLISTTlgoblodaHH$04@H.dLayer 1LISTjlayrflgsLISTRlgoblodaEE$ 04@E.GridLIST0lgobloda$$ $.LISTBpageflgsbboxsx\eobbx sx\ee@$sLIST|AgobjLIST\layrflgs LISTDlgobloda88 (,8.LIST\layrflgsLISTDlgobloda88 (,8.dLISTD@layrflgsLISTDlgobloda88 (,8.dLISTXobj spnd Tflgs bbox)&gobbx )&g)&gusdn6LIST8lgobloda <`lptx| d..>J஗40ftil0??LIST\trfltrfdPPP(?)6A?+Atxsm T?)6A?+AUK@pktxtjLISTXobj spndSflgs bbox & 3obbx & 3 & 3usdn7LIST8lgobloda <`lptx| d..>J`40ftil0??LIST\trfltrfdPPP(?;%A?+AtxsmS?;%A?+A`UK@pjtxtjLISTXobj spndpSflgs bboxobbx usdn5LIST8lgobloda <`lptx| d..>J8(w40ftil0??LIST\trfltrfdPPP(??AtxsmpS??AUK@pktxtjLISTXobj spndSflgs bboxmobbx mmusdn4LIST8lgobloda <`lptx| d..>J8;^40ftil0??LIST\trfltrfdPPP(?*?AtxsmS?*?A8UK@pjtxtjLISThobj spndRflgs bbox x; k-yobbx x; k-y x; k-yusdn'LIST8lgobloda <`lptx| d..>JMp40ftil0??LIST\trfltrfdPPP(? x5A?2"AtxsmR? x5A?2"AMUK@c(mp)txtjLISTXobj spndhRflgs bboxb'[.obbx b'[.b'[.usdn)LIST8lgobloda <`lptx| d..>J諗x40ftil0??LIST\trfltrfdPPP(?b'1A?l4AtxsmhR?b'1A?l4A諗UK@titxtjLISThobj spndRflgs bbox( IW obbx ( IW ( IW usdn*LIST8lgobloda <`lptx| d..>J`ę40ftil0??LIST\trfltrfdPPP(?(0A?u'AtxsmR?(0A?u'AUK@f(ti)txtjLISTXobj flgs bbox*2  u obbx *2  u b$  P ߳usdn+LISTlgobloda4P\`dhl d..J(Ւƙ^CÙLIST\trfltrfdPPP4(?Td"A? &ALISTobj flgs bboxg  obbx g   cusdn,LIST0lgobloda4P\`dhld..JԒHgr4ftil0??LIST\trfltrfdPPP4(?1A@33?&ALISTbobj flgs bboxu obbx u u > Qusdn-LISTlgobloda4P\`dhl d..JԒxʟyDǙLIST\trfltrfdPPP7(?*A?l!ALISTbobj flgs bbox!obbx ! > !Qusdn.LISTlgobloda4P\`dhl d..JԒ(CyD@ƙLIST\trfltrfdPPP7(?1A?l!ALISTgrp flgsbbox! \eobbx ! \e6 $s߳usdn/LISTlgoblodaDD$0<@Dd.JLIST\trfltrfdPPP?%FA?ЎLISTobj flgsbboxL hobbx L h u-usdn0LIST8lgobloda8Xdhlpt d..JӒ^C4Ùftil0??LIST\trfltrfdPPP4(.?@VK7A.?Sn"ALISTXobj flgsbbox! \eobbx ! \e6 $s߳usdn1LISTlgobloda4P\`dhl d..JXӒ8^CÙLIST\trfltrfdPPP4(?"!5A?@ &ALISThobj spndPOflgs bboxI\7 ԧobbx I\7 ԧI\7 ԧusdn$LIST8lgobloda <`lptx| d..>J840ftil0??LIST\trfltrfdPPP(?nG-?"AtxsmPO?nG-?"AUK@c(ml)txtjLISTXobj spndNflgs bboxUobbx UUusdn!LIST8lgobloda <`lptx| d..>J0B40ftil0??LIST\trfltrfdPPP(??AtxsmN??A0UK@titxtjLISThobj spndNflgs bbox W+[ obbx  W+[  W+[ usdn LIST8lgobloda <`lptx| d..>Jx=40ftil0??LIST\trfltrfdPPP(??T|'AtxsmN??T|'AUK@f(ti)txtjLISTXobj flgs bbox[E Fobbx [E F  TusdnLISTlgobloda4P\`dhl d..J Ӓ^CÙLIST\trfltrfdPPP4(?(?&ALISTobj flgs bbox^E Cobbx ^E C jusdnLIST0lgobloda4P\`dhld..JҒ̙r4ftil0??LIST\trfltrfdPPP4(?n@33?&ALISTbobj flgs bboxV^obbx V^V: ^/usdnLISTlgobloda4P\`dhl d..JҒ@yDhřLIST\trfltrfdPPP7(?6R%?(!ALISTbobj flgs bboxu[obbx u[u: [/usdnLISTlgobloda4P\`dhl d..JxҒ8yDęLIST\trfltrfdPPP7(?x(?(!ALISTgrp flgsbboxE Vobbx E V  dusdnLISTlgoblodaDD$0<@Dd.JPLIST\trfltrfdPPP??#A?ЁLISTobj flgsbbox Y5obbx  Y5  _gmusdnLIST8lgobloda8Xdhlpt d..J@Ғ^C4Ùftil0??LIST\trfltrfdPPP4(.?SV).?S"ALISTXobj flgsbboxE Vobbx E V  dusdnLISTlgobloda4P\`dhl d..JҒL^CÙLIST\trfltrfdPPP4(?@-?@&ALIST\layrflgsLISTDlgobloda88 (,8.LISTlgobloda@@$04<@d . יLIST\trfltrfdPPP(??sumi8e@L7Ϛ)`DRAWBITMAPS r rSTREAMSLISTSTREAMSLISTDRAWPATTERNS r rSTREAMSLISTSummaryInformation(_12332456073H觜r`rOle  |x CORELDRWx1@@@c_}@ήQ}CDRAObjInfoCorelDRAWkDRAWBITMAPS`r`rSTREAMSLIST      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdgjklmnopqrstuvwxyz{|}~RIFFkCDRAvrsnLIST!doc pfrdmcfg Q-p8WF]tE?F]tE??8888 7c' LastUsed/6a.?AVCLastUsed/6a.?AVCLastUsed/6a.?AVCdd@@@@?}4V?}4V?C:\Corel\WebSite\LastUsed/6a.?AVCDocTemplate@@/6a.?AVCMultiDocTemplate@@/6a.?AVWDrawMultiDocTemplate@@Initializing the App Recovery systemPLGT3280.DLLcoreldrw.exeRAVE.exedrawintl.dll.tlbDrawBrowser\Rave@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va@l2va(@ptrt#P$'3LIST$fnttfontX@ ArialXwc3XXWesternfontX%@Arial Unicode MSXXCHINESE_BIG5fontX@Arabic TransparentXArabicLIST@arrtarrw4PՠV>  DDH" "LISTXfiltLISTfilcfildxOLIST0filcfild$@tdX|@<LISTTotltoutl#d??d@<outld??d@<outl͟d??d@<Pՠoutlnd??d@<LIST stltH@txOnn#\%JJJ@@@ؑX @B@B@B' -c9 V? c5  ncdDefault GraphicaDefault Artistic Textn\ؑX P2an\ؑX an\ؑX an\ؑX Xba\ba\ba\]a\0^a\x^a\0a\`0a\0a\X2a\_a\^a\0a\81a\1a\1a\0an\ؑX ȅan\ؑX 8an\ؑX an\ؑX an\ؑX an\ؑX 膠an\ؑX Xan\ؑX an\ؑX an\ؑX `an\ؑX an\ؑX 8an\ؑX 0   h   H֟֟֟h0(ן"" 8PpHh`LISTuil LISTbpageflgsbbox@_( ( @_obbx @_( ( @_@_( ( @_LISTgobjLISTllayrflgs LISTTlgoblodaGG$ 04@G.GuidesLISTllayrflgsLISTTlgoblodaHH$04@H.dDesktopLISTllayrflgsLISTTlgoblodaHH$04@H.dLayer 1LISTjlayrflgsLISTRlgoblodaEE$ 04@E.GridLIST0lgobloda$$ $.LISTFpageflgsbbox1'3obbx 1'3#P$'3LISTEgobjLIST\layrflgs LISTDlgobloda88 (,8.LIST\layrflgsLISTDlgobloda88 (,8.dLISTDlayrflgsLISTDlgobloda88 (,8.dLISTRobj spndflgs bboxUZobbx UZUZusdnLIST8lgobloda <`lptx| d..>JaGP@tnftil0??LIST\trfltrfdPPP(??"txsm??"aUK0.5txtjLISTgrp flgsbboxI,4oֲobbx I,4oֲo}usdnLISTlgoblodaDD$0<@Dd.JLLIST\trfltrfdPPP?9A?r7LISTobj flgsbbox)Yr$obbx )Yr$!\usdnLIST8lgobloda8Xdhlpt d..JL^C@tftil0??LIST\trfltrfdPPP4(.?L A.?YgLISTXobj flgsbboxI,4oֲobbx I,4oֲo}usdnLISTlgobloda4P\`dhl d..JPP$^CLIST\trfltrfdPPP4(?n(@?/'LISTjobj spnd flgs bbox}_c.obbx }_c.}_c.usdnLIST8lgobloda <`lptx| d..>JaP%Q@tnftil0??LIST\trfltrfdPPPB(?A#?txsm ?A#?a UKis_fronttxtjLISTVobj spndflgs bbox9iobbx 9i9iusdnLIST8lgobloda <`lptx| d..>Ja(S@tnftil0??LIST\trfltrfdPPP>(??@txsm??@aUKis_atxtjLISTTobj spndpflgs bboxN(~Zobbx N(~ZN(~ZusdnLIST8lgobloda <`lptx| d..>Ja(@tnftil0??LIST\trfltrfdPPPk(?t?|zAtxsmp?t?|zAaUK@p5txtjLIST`obj spndflgs bbox#9obbx #9#9usdnLIST8lgobloda <`lptx| d..>JaŸ8ß@tnftil0??LIST\trfltrfdPPP6(?f"?@@txsm?f"?@@aUKMammaltxtjLISTVobj spndflgs bboxJiobbx JiJiusdnLIST8lgobloda <`lptx| d..>Ja8@tnftil0??LIST\trfltrfdPPP2(?l(?@txsm?l(?@aUKis_atxtjLISTTobj spndhflgs bbox)6obbx )6)6usdnLIST8lgobloda <`lptx| d..>Ja@ȇ@tnftil0??LIST\trfltrfdPPP+(?&?/Atxsmh?&?/AaUK@t3txtjLISTRobj spndflgs bbox:ffr#tobbx :ffr#t:ffr#tusdnLIST8lgobloda <`lptx| d..>Ja`蓡@tnftil0??LIST\trfltrfdPPP,(?3-?Atxsm?3-?AaUK0.9txtjLISTjobj spndflgs bbox*0Bobbx *0B*0BusdnLIST8lgobloda <`lptx| d..>Jax˝̝@tnftil0??LIST\trfltrfdPPP (?8?h~Atxsm?8?h~Aa UKElephanttxtjLISTTobj spnd`flgs bboxx\)"obbx x\)"x\)"usdnLIST8lgobloda <`lptx| d..>Jax  @tnftil0??LIST\trfltrfdPPP(?-?@txsm`?-?@aUK@p4txtjLIST\obj spndflgs bbox!7[anobbx !7[an!7[anusdnLIST8lgobloda <`lptx| d..>Ja`@tnftil0??LIST\trfltrfdPPP(?}1?q%txsm?}1?q%aUKClydetxtjLISTTobj spndflgs bbox[>gobbx [>g[>gusdnLIST8lgobloda <`lptx| d..>Ja"@tnftil0??LIST\trfltrfdPPP(?l=?Otxsm?l=?OaUK@p8txtjLIST\obj spndXflgs bboxҞjobbx ҞjҞjusdnLIST8lgobloda <`lptx| d..>Jap@tnftil0??LIST\trfltrfdPPP(?.a=?%txsmX?.a=?%aUKWhitetxtjLISTtobj spndflgs bboxSO֥4obbx SO֥4SO֥4usdnLIST8lgobloda <`lptx| d..>Ja0}P~@tnftil0??LIST\trfltrfdPPP(?8?.L&txsm?8?.L&a UK has_colourtxtjLISTRobj spndflgs bbox%2|Goobbx %2|Go%2|GousdnLIST8lgobloda <`lptx| d..>Ja>X?@tnftil0??LIST\trfltrfdPPP(?r4?"txsm?r4?"aUK0.7txtjLISTTobj spndPflgs bbox&gobbx &g&gusdnLIST8lgobloda <`lptx| d..>JaKA@tnftil0??LIST\trfltrfdPPP (?5?OtxsmP?5?OaUK@t8txtjLISTVobj spndflgs bbox&qwQCobbx &qwQC&qwQCusdnLIST8lgobloda <`lptx| d..>JaF@tnftil0??LIST\trfltrfdPPP (?5?txsm?5?aUKis_atxtjLISTTobj spndflgs bbox"7obbx "7"7usdnLIST8lgobloda <`lptx| d..>Ja@@tnftil0??LIST\trfltrfdPPP(?$2?txsm?$2?aUK@t2txtjLISTRobj spndHflgs bboxbNW\obbx bNW\bNW\usdnLIST8lgobloda <`lptx| d..>J28QQ@tnftil0??LIST\trfltrfdPPP(?+?txsmH?+?2UK0.9txtjLISTRobj spndflgs bbox54qobbx 54q54qusdnLIST8lgobloda <`lptx| d..>Ja_@tnftil0??LIST\trfltrfdPPP(? )?txsm? )?aUK0.9txtjLISTTobj spndflgs bbox9_Vobbx 9_V9_VusdnLIST8lgobloda <`lptx| d..>JaP=5=5@tnftil0??LIST\trfltrfdPPP(?z!?txsm?z!?aUK@t9txtjLISTZobj spnd@flgs bbox'?t7JKobbx '?t7JK'?t7JKusdnLIST8lgobloda <`lptx| d..>Ja@tnftil0??LIST\trfltrfdPPP~(?)? txsm@?)? aUK@t12txtjLISTZobj spndflgs bboxz~t^Hobbx z~t^Hz~t^HusdnLIST8lgobloda <`lptx| d..>Ja55@tnftil0??LIST\trfltrfdPPPv(?? txsm?? aUK@t10txtjLISTpobj spndflgs bbox$^30Wobbx $^30W$^30WusdnLIST8lgobloda <`lptx| d..>JaH@tnftil0??LIST\trfltrfdPPPy(?C!?q%txsm?C!?q%a UK is_behindtxtjLISTRobj spnd8flgs bbox5Ūobbx 5Ū5ŪusdnLIST8lgobloda <`lptx| d..>JaX@tnftil0??LIST\trfltrfdPPPp(? )?Dtxsm8? )?DaUK0.9txtjLIST\obj spndflgs bbox>o obbx >o >o usdnLIST8lgobloda <`lptx| d..>Ja{`|@tnftil0??LIST\trfltrfdPPP(? PA?'@txsm? PA?'@aUKHumantxtjLISTpobj spndflgs bbox:fHNobbx :fHN:fHNusdnLIST8lgobloda <`lptx| d..>Jp@tnftil0??LIST\trfltrfdPPP(?3-?̏0Atxsm?3-?̏0A UK QuadrupedtxtjLIST\obj spnd0flgs bbox&M\ٖobbx &M\ٖ&M\ٖusdnLIST8lgobloda <`lptx| d..>Jp@tnftil0??LIST\trfltrfdPPP(?A?0Atxsm0?A?0AUKBipedtxtjLIST~obj spndflgs bbox-eobbx -e-eusdnLIST8lgobloda <`lptx| d..>JXb@tnftil0??LIST\trfltrfdPPP? (?A2?+txsm?A2?+Xb UK has_charact.txtjLISTRobj spndflgs bboxuJb@tnftil0??LIST\trfltrfdPPP? (?%?t&'txsm?%?t&'bUK0.6txtjLIST~obj spnd(flgs bboxaLA'3obbx aLA'3aLA'3usdnLIST8lgobloda <`lptx| d..>Jb@tnftil0??LIST\trfltrfdPPP (?Ь$?$1txsm(?Ь$?$1b UK Kind_heartedtxtjLISTZobj spndflgs bbox!- .obbx !- .!- .usdnLIST8lgobloda <`lptx| d..>J]@tnftil0??LIST\trfltrfdPPP(?t?/txsm?t?/]UK@p10txtjLISTobj spndxflgs bbox17\>-obbx 17\>-17\>-usdnLIST8lgobloda <`lptx| d..>J0^ @tnftil0??LIST\trfltrfdPPP0 (?x?!+txsmx?x?!+0^UKhas_characteristicstxtjLISTZobj spnd flgs bboxy`?bobbx y`?by`?busdnLIST8lgobloda <`lptx| d..>Jx^@tnftil0??LIST\trfltrfdPPP (?n?!txsm ?n?!x^UK@t11txtjLISTVobj spndflgs bboxjobbx jjusdnLIST8lgobloda <`lptx| d..>J0#x$@tnftil0??LIST\trfltrfdPPP* (?x4A?%txsm?x4A?%0UKFredtxtjLISTTobj spndpflgs bbox  obbx     usdnLIST8lgobloda <`lptx| d..>J`066@tnftil0??LIST\trfltrfdPPP (?%A?I#txsmp?%A?I#`0UK@p1txtjLISTRobj spndflgs bbox @Qobbx  @Q @QusdnLIST8lgobloda <`lptx| d..>J0IpJ@tnftil0??LIST\trfltrfdPPP)(? PA?Htxsm? PA?H0UK0.8txtjLISTZobj spndflgs bboxyo Cobbx yo Cyo CusdnLIST8lgobloda <`lptx| d..>JX2l@m@tnftil0??LIST\trfltrfdPPP(?A? txsm?A? X2UK@t13txtjLISTRobj spndhflgs bboxljyu obbx ljyu ljyu usdnLIST8lgobloda <`lptx| d..>J_ЀX@tnftil0??LIST\trfltrfdPPP(?ԳA?%txsmh?ԳA?%_UKhastxtjLISTVobj spndflgs bboxA 7Cobbx A 7CA 7CusdnLIST8lgobloda <`lptx| d..>J^ @tnftil0??LIST\trfltrfdPPP (?(A?q%txsm?(A?q%^UKSoultxtjLISTTobj spndflgs bbox\obbx \\usdnLIST8lgobloda <`lptx| d..>J0ئ`@tnftil0??LIST\trfltrfdPPP (?-A?(txsm?-A?(0UK@p9txtjLISTRobj spnd`flgs bbox 6obbx  6 6usdnLIST8lgobloda <`lptx| d..>J81x@tnftil0??LIST\trfltrfdPPP (?~@? 6txsm`?~@? 681UK0.9txtjLISTTobj spndflgs bboxI,i=@obbx I,i=@I,i=@usdnLIST8lgobloda <`lptx| d..>J1@@tnftil0??LIST\trfltrfdPPP (?@?ktxsm?@?k1UK@t1txtjLISTVobj spnd flgs bbox%q8vQCobbx %q8vQC%q8vQCusdnLIST8lgobloda <`lptx| d..>J1X@tnftil0??LIST\trfltrfdPPP(?0A?txsm ?0A?1UKis_atxtjLISTTobj spndX flgs bboxV,:obbx V,:V,:usdnLIST8lgobloda <`lptx| d..>J0 @tnftil0??LIST\trfltrfdPPP (?(sA? AtxsmX ?(sA? A0UK@p2txtjLISTTobj spnd flgs bboxy:(obbx y:(y:(usdnLIST8lgobloda <`lptx| d..>Jȅ8 @tnftil0??LIST\trfltrfdPPP(?n?LAtxsm ?n?LAȅUK@t4txtjLISTTobj spnd flgs bbox3 M{y obbx 3 M{y 3 M{y usdnLIST8lgobloda <`lptx| d..>J8`@tnftil0??LIST\trfltrfdPPP(?g?8{#Atxsm ?g?8{#A8UK@t6txtjLISTRobj spndP flgs bbox| Ӡ& obbx | Ӡ& | Ӡ& usdnLIST8lgobloda <`lptx| d..>J/!@tnftil0??LIST\trfltrfdPPP(? ?~U$AtxsmP ? ?~U$AUK1.0txtjLISTjobj spnd flgs bboxV p*5y obbx V p*5y V p*5y usdnLIST8lgobloda <`lptx| d..>JP33@tnftil0??LIST\trfltrfdPPPd(?O?u'Atxsm ?O?u'A UKis_not_atxtjLISTTobj spnd flgs bboxL: obbx L: L: usdnLIST8lgobloda <`lptx| d..>JFG@tnftil0??LIST\trfltrfdPPPg(?|2A?LAtxsm ?|2A?LAUK@t5txtjLISTRobj spndH flgs bbox%obbx %%usdnLIST8lgobloda <`lptx| d..>J膠Z0[@tnftil0??LIST\trfltrfdPPP(?0A? AtxsmH ?0A? A膠UK0.8txtjLISTVobj spnd flgs bbox+S QXobbx +S QX+S QXusdnLIST8lgobloda <`lptx| d..>JXpmm@tnftil0??LIST\trfltrfdPPPf(?A?(5!Atxsm ?A?(5!AXUKis_atxtjLISTTobj spnd flgs bboxb99 obbx b99 b99 usdnLIST8lgobloda <`lptx| d..>J@@tnftil0??LIST\trfltrfdPPP (?Jش`@tnftil0??LIST\trfltrfdPPPd(?&?v2#Atxsm?&?v2#AUKis_atxtjLISTRobj spnd(flgs bboxb_ "n obbx b_ "n b_ "n usdnLIST8lgobloda <`lptx| d..>J`(@tnftil0??LIST\trfltrfdPPP^(?Jh@tnftil0??LIST\trfltrfdPPP(?"@?1Atxsmв?"@?1AUK@p6txtjLISTTobj spndxflgs bboxOobbx OOusdnLIST8lgobloda <`lptx| d..>J8@tnftil0??LIST\trfltrfdPPP(?F?51Atxsmx?F?51A8UK@p7txtjLISTbobj flgs bboxZD,*obbx ZD,*mUO,*usdnLISTlgobloda4P\`dhl d..Jf60 yD͟LIST\trfltrfdPPP7(P'?/)m(?j_2?_!V$LISTbobj flgs bbox=;9lobbx =;9l69lusdnLISTlgobloda4P\`dhl d..Jf6 8yD͟LIST\trfltrfdPPP7(@j[Wy)>e?ay!J6t_?sh?{&+LISTobj flgs bbox3-_obbx 3-_*5]usdnLIST0lgobloda4P\`dhld..Jf6 r@tftil0??LIST\trfltrfdPPP4(f=6m-?h%Υ\P?S^t?i4mu"LISTbobj flgs bboxG@.{obbx G@.{BKx{usdnLISTlgobloda4P\`dhl d..Jf6 (yD͟LIST\trfltrfdPPP7(P'? :m(?j_2?_!V#%LISTbobj flgs bbox"}(hobbx "}(h"}#husdnLISTlgobloda4P\`dhl d..Jf6h yD͟LIST\trfltrfdPPP7(@j[Wy?)>e+<}x J6t_?sh?{6.+LISTobj flgs bboxK.[obbx K.[61'YusdnLIST0lgobloda4P\`dhld..Jf6h r@tftil0??LIST\trfltrfdPPP4(f=6?m-翲\mΥ\P?S^t?i4}"LISTbobj flgs bboxxTPobbx xTPxusdnLISTlgobloda4P\`dhl d..Jh' xM=(D͟LIST\trfltrfdPPP7(cz?'2?=ɦΙ&x=?ѿr:~?W&LISTbobj flgs bboxTmobbx TmTusdnLISTlgobloda4P\`dhl d..Jh' D͟LIST\trfltrfdPPP7(B? re?Lx=?ѿr:~?2PkLISTobj flgs bboxO܌4~Iobbx O܌4~IYHusdnLIST0lgobloda4P\`dhld..JkE8r@tftil0??LIST\trfltrfdPPP4(Sr01=/o-@33DLISTXobj flgs bbox9_!}.obbx 9_!}.+TmfusdnLISTlgobloda4P\`dhl d..J P ^CLIST\trfltrfdPPP4(?z!?@'LISTbobj flgs bboxOobbx O(O0usdnLISTlgobloda4P\`dhl d..JkE!yD͟LIST\trfltrfdPPP7(ף@XB=Y Fe)LISTbobj flgs bbox5J,obbx 5J,5(J,/usdnLISTlgobloda4P\`dhl d..JkEH#yD͟LIST\trfltrfdPPP7(}W2eB=S#@Y)kLISTobj flgs bboxO44ֲobbx O44ֲYusdnLIST0lgobloda4P\`dhld..JkEH`$r@tftil0??LIST\trfltrfdPPP4(Sr01=/o-@33T"LISTbobj flgs bboxi;Gobbx i;Gi(;G"usdnLISTlgobloda4P\`dhl d..JkE֟%yD͟LIST\trfltrfdPPP7(= ?e)G5e)LISTbobj flgs bboxsobbx ss(!usdnLISTlgobloda4P\`dhl d..JkE֟P'yD͟LIST\trfltrfdPPP7(`4 ?Y)h FL8Y)lLISTobj flgs bbox4iֲobbx 4iֲYusdnLIST0lgobloda4P\`dhld..JkE֟(r@tftil0??LIST\trfltrfdPPP4(?Sr01 |5-@33T"LISTXobj flgs bboxV04sֲobbx V04sֲ"|usdnLISTlgobloda4P\`dhl d..J֟@*^CLIST\trfltrfdPPP4(<=*/'LISTbobj flgs bboxo{obbx o{o({0usdn~LISTlgobloda4P\`dhl d..JkEh+yD͟LIST\trfltrfdPPP7(= e)=0Ae)LISTbobj flgs bbox_0C obbx _0C _(0C /usdnLISTlgobloda4P\`dhl d..JkE0,yD͟LIST\trfltrfdPPP7(`4 Y)=`&AY)kLISTobj flgs bbox{4`ֲobbx {4`ֲݵYusdnLIST0lgobloda4P\`dhld..JkE0.r@tftil0??LIST\trfltrfdPPP4(Sr01=@h!A-@33T"LISTXobj flgs bbox/C 4ֲobbx /C 4ֲg5 VusdnLISTlgobloda4P\`dhl d..J(ןh0^CLIST\trfltrfdPPP4(=? /A=*/'LISTbobj flgs bboxhh6obbx hh6A6usdnlLISTlgobloda4P\`dhl d..J1yD͟LIST\trfltrfdPPP7(=?0= ":x=[LISTbobj flgs bboxhh:obbx hh:B:usdnmLISTlgobloda4P\`dhl d..J"3yD͟LIST\trfltrfdPPP7(=?0`4 1x=m7@LISTobj flgs bbox=G:obbx =G:Ej<usdnnLIST0lgobloda4P\`dhld..J"x4r@tftil0??LIST\trfltrfdPPP4(=@33?,Sr0!=/QLISTXobj flgs bbox=G6زobbx =G6زu9cusdnoLISTlgobloda4P\`dhl d..J"6^CLIST\trfltrfdPPP4(?ø2?('LISTbobj flgs bboxH-obbx H-Ѕ_-usdnfLISTlgobloda4P\`dhl d..Jh'X7yD͟LIST\trfltrfdPPP7(B reؿY:9ʤ'x=?ѿr:~?5<ALISTbobj flgs bbox90obbx 90<(9#usdngLISTlgobloda4P\`dhl d..Jh'8yD͟LIST\trfltrfdPPP7(-t&=j4ֿ$Bk{!0Ͽ]?t sOALISTobj flgs bboxo"0obbx o"0&m 2} usdnhLIST0lgobloda4P\`dhld..Jh':r@tftil0??LIST\trfltrfdPPP4( {\k꓊&ڿ^K'rJdXֿl##?׉ALISTXobj flgs bbox=G obbx =G u9cusdniLISTlgobloda4P\`dhl d..J ;^CLIST\trfltrfdPPP4(,?l4ALISTbobj flgs bboxMM6obbx MM6M6usdnaLISTlgobloda4P\`dhl d..J=yD͟LIST\trfltrfdPPP7(=?o A= ":x=[LISTbobj flgs bboxMM:obbx MM:N:usdnbLISTlgobloda4P\`dhl d..J8>yD͟LIST\trfltrfdPPP7(=?o A`4 1x=m7@LISTobj flgs bboxI,:oobbx I,:o*j<$qusdncLIST0lgobloda4P\`dhld..J8@@r@tftil0??LIST\trfltrfdPPP4(=@33?,ASr0!=/QLISTbobj flgs bboxIHI,obbx IHI,IR1usdn]LISTlgobloda4P\`dhl d..Jh'PAyD͟LIST\trfltrfdPPP7(B? re?E3-6nkx=?ѿr:~?5<ALISTbobj flgs bbox"}0obbx "}0"} (#usdn^LISTlgobloda4P\`dhl d..Jh'0CyD͟LIST\trfltrfdPPP7(-t?&=j4?{)x 0Ͽ]?t sOALISTobj flgs bboxS #0Gobbx S #0Gk 2I usdn_LIST0lgobloda4P\`dhld..Jh'Dr@tftil0??LIST\trfltrfdPPP4( {\?k꓊&?e)rJdXֿl##?׉ALISTbobj flgs bboxvH *obbx vH *vH 5"usdnTLISTlgobloda4P\`dhl d..Jp FyD͟LIST\trfltrfdPPP7(&/~?;?xIdNly!jL?Уo#ALISTbobj flgs bboxpד  obbx pד  pד  usdnULISTlgobloda4P\`dhl d..JGyD͟LIST\trfltrfdPPP7(r?kD?TH ABH0 ?'+ALISTobj flgs bboxL/f 3%obbx L/f 3%6-h I#usdnVLIST0lgobloda4P\`dhld..JHr@tftil0??LIST\trfltrfdPPP4(5jkN?XR?PruBIKD^?E!&ALISTbobj flgs bboxMM obbx MM M usdnQLISTlgobloda4P\`dhl d..JpJyD͟LIST\trfltrfdPPP7(=?o AD qܶ=,5!ALISTbobj flgs bboxM MzP obbx M MzP  OzP usdnRLISTlgobloda4P\`dhl d..JXLyD͟LIST\trfltrfdPPP7(=?o A@HBD=6;"s)ALISTobj flgs bboxI,{P oobbx I,{P o*Q $qusdnSLIST0lgobloda4P\`dhld..JMr@tftil0??LIST\trfltrfdPPP4(=@33?,ASr0!= "ALISTXobj flgs bboxI, oobbx I, oo}usdnOLISTlgobloda4P\`dhl d..JHHO^CLIST\trfltrfdPPP4(?@?l4ALISTbobj flgs bboxn[; r[i obbx n[; r[i  ; ǯi usdnMLISTlgobloda4P\`dhl d..JPyD͟LIST\trfltrfdPPP7(=?>@= ":x=iv2#ALISTbobj flgs bboxo[ s[O obbx o[ s[O  ȯO usdnKLISTlgobloda4P\`dhl d..JhRyD͟LIST\trfltrfdPPP7(=?><`4 1x=)ALISTobj flgs bbox9!O !}; obbx 9!O !}; F8P ~ usdnLLIST0lgobloda4P\`dhld..Jh`Sr@tftil0??LIST\trfltrfdPPP4(=@33?{ Sr0!= @$ALISTXobj flgs bbox9i !} obbx 9i !} +1CusdnJLISTlgobloda4P\`dhl d..JT^CLIST\trfltrfdPPP4(?z!?)ALISTXobj flgs bboxI,P!o obbx I,P!o /o}* usdn/LISTlgobloda4P\`dhl d..J`@V^CLIST\trfltrfdPPP4(?@?P!1ALISTXobj flgs bbox9P!!} obbx 9P!!} +/* usdn'LISTlgobloda4P\`dhl d..JW^CLIST\trfltrfdPPP4(?z!?P!1ALIST\layrflgsLISTDlgobloda88 (,8.LISTlgobloda@@$04<@d . LoLIST\trfltrfdPPP(??sumi8#P$'37)`STREAMSLISTDRAWPATTERNS`r`rSTREAMSLISTSummaryInformation(Oh+'Oh+'0  ( 4 @ LX`hp|x CORELDRWx1@@@б}@,4}CDRAOh+'Oh+'0QT謇+-ceΠtK5xX錊Ӝ_EH?Axw{}1(ln,09y$lS=Ô[]3,4߃(%%'dddSybKGDH cmPPJCmp0712HsiVIDATx^CSW6^@)X ۣ#Pk--E2.xt,a#{}; T' !ٷg&l#7oa#7@?3pr8_A Џ ܠGpWnPz#8G+7(# Jo~g~7@?3pr8_A Џ ܠGpWnPz#8G+7(# Jo~g~7@?3pr8_A Џ ܠGpWnPz#8G+7(#ݿY}#Bag/8cy}dV=ܘm͍|*Y6r77K?y8` |j۰NW jGA8̙#ΣHBd~4(uēW]]K>Kw^?45-3=~HxߺvxK"hy>SE4ciiuP~W@>b%Fr&Գ[LJ -rkp;{_/_#emqFWֺFi?JMdOOMOq/}l=Un_0";hlݗ1|- m| WŐ\d3 'qלJ6Nm)Wa{&S@-'~N7bOen+r w1OΎl<[S5vJqqGt) I{~%naͽx qD>' axhmn:@/DʍOpi[ qrua膵ѺQ0峵Tt`n=JXwSjQ؅Mkݩs=-X@ ˯/z"ɩil40*ES07Q<+@٩14oHi o#sSĻݧ3gP^cʢDOL E7_NcA, 1Iͩ@- п :n&3}EMHE 'ruANf|<(i,S]!+P~iu$kgX9C_ ]xρlT$l2V+c77i:_ySr} ꤱ>R|riS._ғx2Z*h2!Byh=cN7\{s1{:};k Of~px[f:TS/W)IRagKb+{7|.sBjjC!)#)ct+o[@o2ׯ}V._^1NyoU]^>jo&! hz_]/-00\M87: H@׵*b2ر["_נ6킱YsŧX&t ə q g=4#{zG~7CC햶J6[rՊ?zܴl˄`!ds#t /lko;y'|r?Î`ws5.+!?ufE zh,77'TN,8!hMz_GI"+r< ^JӰ*!v;&˽$q:}5&n7["7?xR]N6ufW%%=~ڢ'czô?gWq}[]+8/#N( )#.ȩ"/ )K] ft?lceW%okk5ӿǨ_j%2O?,h+P iS\AσۛNZ|mw>f228otUʶ$eŸYNb>l>1y˫u S#t(q1f:e NIs_IA @_Iu(kS*ԪzT--#Eg[O}q^]TW3zM̥ ǖpxYΨ8mrUB4w .{y49uY`Qպj@g+zA|q?i*W<%]zz8օ!&\*W!gOXC#7_vSA^N7L|40UM_nn+CX` pdB:=ڜx/C ˧wE;Txڇ)NsXٳp]_]ij^|:T\FLPxH˱^AC hD8l'UㅎKy~|/)h@$oHz ^?:5" 8¯3[+?MisbLÎ^x@/Z޺0DG'ǩ%TIk5j8yMg[F6 wq+}ŭzKu9f)8/ ٽtJXH&3A)(OMKh)&}rzʙ{ 7֢-ӷ(:l͐ʣuZ^t>! ě. xBzrc\qⵖSЕСoq9Jk7WF4R A3?x[. },Zt(W}XZӥ'w8WZ6'9G]^vvDD+1׌edž@F ߩ:8 ^t# w]ڐ(C8vg(U+_]Ń h?H'e SdϢQv`y9!c@xGeh9d}jkծ g㉔P# :ihkzY LP&K[0OtE6ԋ 1&@ >kK1M>Y~!thɉs&j(EPgJJcwS?`D)8,>+חEF¤jQg,QuO$.ve;|):-볐|]>.FVkiίӸt(Psmɯ2Y*cjv QEРfʥt^v읿ON C0!<ipf5VӅ8.4Sgpb .xN _IjqF-\1q" +eZGEp XMv||}@GNx-Il*@Kq7-޳}keh}nAؚۧ5eGUIݫXw!CHeaxP'>(1OF@[sFDBJ2Ǘ]1Adq3"Es9L 4t c/UqgbP)+y>vZ:䩩CVB#zͩґPjOˇe9wTA<~Akإ(2JR oy>ET3Uu.r 4R̓a%> C4R*p@9\n&bq4T.㾉$ `, P!8ݣӔ'L" تy3BƪVĬGq[uDkۋݾbzeC݁>pL9q(?HNεҧ@~?3*^3+*ӠIwB(L7da- ,(E_zvF9r3mih]eA/4YvBTԻRviGATZTF4/'cбU煐#wtǒ[T[z>R[ffZ+`ʩ7F[{70J>n28b&1hѹ<]ydǸu2ڹH/?=WwQw?͑F6dRzeA;;~|9C&y4 i3G.rWWOA/oO%zUblKv>2#f^BSo^?xE<% NԠ5,ԅ?Ƿ%E%Xa;@L {W7GgT-J$}(f(w_LxKt/g}mgܽ \ՠ('u,'G[F3˜aJV(r2$>JZ펭V(ʅh!`p? ,0w ەM.lt%bV6^xG쟺uP݁L5w7|TMfhU z 6JJ9ev.uՄkUX*FU[ 1U}0ΚNϮOIe}aAqMi߱M/ٻc Uk2:N$n&NxrSe^lsj kG.;zD͖2%ѱzds#8ߞQj 5թFG]ъp* lO䯟۳j,]e ![owVwNc;?UZW6Ao eGiVٻ]2$kB ccQ6}8Pbڼޝ~r;3@d:f)N,%axy眢>NߤGNAB/I‘`BdչXݑ*dU9◙Zp4mި/Ðt59"UvR`ف㓙Sl7A +W9O -&Jd՞M/kmGъ5W =vNibs/a`7%-1?*LdD 4$2y4 >j3f=*,(ހ 1BkYy3T4q# 4׸m+( ]2@=܊i~ՠxC˦T&c JT̙'or;wg^w=]숡88 ,}+6aPGv76->{MJ[$EʟN .iɕq9onjK.ntZ`!3UT,CA9Q/Mz#(N٪Gb} f<_L21-|6c zip\H [T~;,tD Rng x]kMLj՚"QKCIhOrSN hP p-$_s2 ]TAJR#a "}9/@$ W.Q)܂y 2Ψc `#g+ 5cy:gNL\-*X]@|ՆKS8C5>3LGv\LEVj>ͫN;<,í^A_P#Mn-ZT,!*Ԭ2U M(-[Wg|y]J1g*5'RRu߲tE4+E=;R??q ~pK^:Y`*{#ȍrJbkt3$S C?XxHK~hэ=._m.:tX&i$E[hfHx sɑ:fS%sԮx$nAό)|CO` PaAO" "oưlբީCc䘫Q_̜(2VN^ڡE_ fnen8*]mBD^cS_&9qt֓9[piu 3^Qs1BSdv-[+sN˿:Lz}3O;VeT2c[svO@yzpxXwڤ^Q4G20o,(gSp]ϛ)c4fy2)tULci:6u$ggz]ҥ]O~lt;&WD0;H0ؙOі%02M K?n=婖&|2K>\Գ(%١vn͏%mSKqFUGl^N4ZDX ;E*U\,wmon}p \-d*8|N/eNI`\?ɻdU*SOjN?wU/%[C&tsIrjyHmz:oM.)ULQwU,fGt߽jsQ`ʻ\D4_k1Jy uc)1`oؓa%ll|xXHh5w[BJY2JvнZm>ocUnnփxF\zk.֪-vcY^ⷫ6.:.A?RhfImi=~s항SsZ3Lm[ ~EwlI/c|pM$=Ecg x./'{̟0:E5IvN_1ѩۅ"yի#@jOΗu"?c+E'Oѓ[]X^uzóc;gk}wqE9T_m06a,"echd=_tpk{j 0;!)/i2QͨUS{K_/؇L')!zEq]5\Xj$a:{'0@?kck}ñ*VBKI°>(`\ N $S/Aۅ$Ͼ~w]Ԇ#mZ-uQ):̚sj5O 1&4Ԫo.LsyF2?,C} x_P^ 65\:6˜$xB;B_ `Ctf`j]_)Fs+;A5F)xLtzgz31TXL?9.R1YFU7/I,wUxLڹb+ Zk%K.U0AGg96$z"0\:J)%zU2>@׉.\誱e$䮋CLr]M0;zeH zd`5v>~D怒oDDא̂zA%-Ykx=kh-׹%/7xv˫X*ͼ6SgmV],3]9ɥbK*(>d`vۯ],^ AȞBEvL IvZd/W=׏녰[5OW9 J@N Uy7nY)>1F*R'pJZx!˗^Fg] 5><4y-kN2̼=!(ri\{< /}'; .*O7?tB5/ R 71W2rsy v ˋ !Nۖ`jCihaлb[m?&lQ5k c=)mnv `l2<kv.fh–/B ;V6a!DJ*ѢZ+w%&#y\; \㪑xnj+#+φ<Ƴy(z>>Z5^3b\FK< Ŗpx~-a>`'zoG.TdiF0|L=hU>>'nGB#k31aQ)t}oExm4T1xEoowU>}-^0/ٛnݺ  `/ޚnR8TR^ ?Vn }#܁35SzA]8%n ؂ٛB\b P{\5d+7g17̌Nv2Gb8cM}vt6P9iOMn+7AcB3ԋNb'S0 B<˴@[S4[=x3޲bDjG\)oބ,Y/ǚ> @e B:KȖkhZSxvKwoCE;iѨȫZ=׎yz.Qf5Cu}ĉ[cײ۞k]:?voٱGNK*C`cNz{\74H^7ХIf] 3) Nܺ =/W ](JPSYחmY@ĺ7KPuTkRU|7zSNS2i㠔CĥPGzO]}{~Llv҇ m,Pq) UI$)Δن)l,DÔ"Mx)>!Ga8׹$@7|E!Alm xƉQ 庾ˡקbn.DxKz9O N^m 轙RsQ/|@'(QS$9 (5-ӠJVضJR#6 n0/SirP#4rxn :ʶiGr0ˣ=p 8@K['+1sf=ݵ|U Ż쩋AͶO\"@}Lp-[A6Kε?ש 腵 CA/)tw^VZՒ DoWGĭ:TԜ*ȕɣeR֜",9xaݳupǒ*AoSA?<k}9&~yl1w4O"'Zוo5>T'R(ZzJt~ߝAq#ibwaaGG@W0q'G¸2a75h1q"2I`pevxꏷ{柕+:oR5Nܪ~N~I\9@VC}U5 aJ,|*i/9k(1?$^Pf UG{"JY4w5/#'5@SMGsTMվV&K"^" }F[q>{_k oRWxA{y  rouu̠wG(Ȟ^CLX tCǗqfVI/Vfgz q9;"bh w:To/,BK'|:cj8)aދtҺ[Fg/$⤣~{9z<ꢍ|Bg>IA!S˚ANMMV4A}d k2H6,ndGixGvJ@{?$n8pp}` ^* o97Ο>' oyZohmkǁ6Qw$( ψ˹άHh|InO.N`P4mNz&f]I&j"S(vOgԑ/|ЪD EyKٛ(k/EڹyYq.Du+M} G>,q^A3|4W]1--T b>K̓)o^Fܦ;AbQ:L,zym!Q,Baw6: )vzo \[,kb}Q92qtHg2KYă1P6lV =]Y\_$\[qQ1E#=X)ܫ.~ݟ2TL^sՊc/U|)Op^"LEѭ2H+~9 k1;ȁo>Œ>#Wk ]:73%aM>C0GUXR=Gu%#_\مfOL|oC[b+{lFƌ~ gu ]pRjkG$ UpbEn ǐݸQe5<Y%oD&?姤P夊׋WOFBo*a;D-'l\!3(,k_<}i$<66f:dzU #Dq+losqRG0F+xyEqOwĪAdB?(dm*R"SkQN'?KOV )۬9N2Ne.cG^:gꨞkѺm(ٍYR? F#"roE+HMpZhJ 8UޞYL$pM#P]/I : jFy[D镀Ov_msqeP5\ Z%O\piHݠUvu}n9n$g)NJNVdgpcy;T>dWBN = Xv ˊ)MŴ6.2G`YD>\$j:34t4ozyV[oˇ)]:nגO~TϋV0X R')vP/`o9؀zVhteσKV{8W@уIN:5LT8͎1WXClaw]$tv@l ͉y]-9>UL kp.2jY'E:P#M1Sԍr004*`-~w^@wJ@B?`rnM9UYJsTsY^fChҴe]$ DzDD#TltXQyj`z ni^"5 nY*ӟB2-b(RdJUPz853RQw>-1 Dv^, Y}D+d36Qn>xM4cX^G>xF3a7̐{50˽{Wp1Wԋvj~+s[e8{x xtzWO>/g%z TMOZ 1PrfJk`"O!-,g0BvD{h+Ibk\Ww| 9[)S('U?lΈx-< uoZQ֗(3ggn=7Ng]rԅtIj;h %o?f'm3+kn"Itf?c0NXIǚŧi1QaY(ܼ%}h.jhERQDr&{{|VTsg{.Mً) kF'TR}A.Հ%@ /;R0l<)#P p@37OV7D떆2+MФ*9|=Qhu~ U@4t˸YrL [S,L3J/2_z*Ҝ5I$KꑫN$)c#5CMa6 cG*qFx=[9 dȕ'Ar)߁tv >mA:l9rdrCG#8oإ|L8rx#jhZz?NgonYT-~ :ĻchiqNyʺbCȹe NXq7L3GyYERl\F[67yrgit3:+zs1rG3q5$w_WM}\,u' WΘx:s9o&,*Bv`8L)&*x]GRTH?]#󐩉i;D܏ Gq:&?fhz\יN뗎XP)ӾB=h ͨc}Ќh_="U~]‹δqmP/n'aW4jsFUY' ᠃f&VW.3c"* sT ^~Z]*V SSK:g_ |A3"腻"X  ;Y7ˢ 5Njx1/<#t]k@& kT-pSXXލuPW&]'ȭ% r)oR46A7Tu@=g7yڼ]53o@;ST {x'UE?_a\kږѫ6BO߼id@`maj.5>@C]cvV lףe&^)Z()!A럑Z~&9Qbx\L"s烬!a&Tqi.s@v#7M|Ȅ{\疑!r)ޒ(EQr5R6z:詴bijxfZ:޶xu-Tx \mN2~nlB&Nu>•*AӮ~ɏ1ݜҥK-I&0gvQB *\{9Ag-35:Gz"+[}l) ClƹhʼnםZx-{U$nTcԃ=7d-w8=GטJ=Ղ'Bb:U> ]'+VsMEۀҕWVJǎtʦIe܄('޾:._M ّ$q'J,tz;ăU vﴋ)9d N%%:(hvV%AG̍~-׋Yg p`t-U֩ Ǵ܇ND9ّ>Nwiޠ[}בcl@Ә/>J=/#NML Kīt3Mcر`_%"h$;3ʕ hL!_ TfB EצOaߢc<«,!ոFPSWe?r#82ʦj559>}aJpT?*|dCU3%erx\er8+gǯ734*˦"ç}Vr|#9y.@%]\5*}=w{ݙPV_x;yʸ3?֨"Q;ZE7} K쬑EZ!v7pc]y؛!/IZ1o'Ĝ7\_#7YqYZ'$q MAAFlܙ@Uz5ⶐ]sX4`-qI70!{hV蹥 ]45[? J$2"o^BИ nM"ﵔ&R쾴,rU4僴'C@L՛N֪!'fsK܁ zvt3݅tAZaDago\,̧:\eR婔:+&ZA(>7#vL3J:[= ̕pNe'kO,m i_VRsj +dPF3suWj0:ʞ{]2Sc'/U =$m4{YNT3}VC~8+.Ej$9b~u J~=QOV_%R[MK)r_}M}+}_xB:rh2/L[ޕp>2vjSV@ ]Obmb{C6]bF:3=9[C]|7>sL7o~rϔuzRse!юU{V˴n ^*];hV[C*8/n؈?"KqPyz mږ)g|T zK9lg3zdmTGɔ1.1sX& j^ %AJg+[L+]Ew$Yjv q>#ZYX镾ՉQ/JturEf5R]Pџ};8 0BJ~*5w^]TsXų|K7CyqAd$56 ?W xntFڸW~&73U u[ Rd=NCώ)7VWB}3}z 9?'ɦ o}E&a%FqM?!|at /uJ;>)tCؐez, A6Pf;_ޏ'9xg$.Q 9b7wmg !ߛvzؖVT&4զR5»t-Cz?t)cPy$;YnjG ˏ K7%>qFHvIp>N]sa=!vj^Mݘfo.mBw_ ֌W4kY-7ka{tiҋׯ|Flbz RA9SB1Wٹ "Vҝ & qvMS[JǨB{Kqq^`RM<YV9tJq5`wU&" { ,㩉`KHe4*X8JrHp6. ]9wqo|lE%e{ [hn WS|,OK7LţK*1fc0{-4⿾K#h![wC\8~N*Enx2/51͐f:?u}3SFh)픸"ݻaI9`G[<;C +Ёʏwt\AҜ7NSJyC-לz6<+/}~k_HRvK2M~l[5qmâ0u9'э;G2y{vQ}q*/0\[-4%QxhLfZh%`+g&so>ar0pOk5HVdtYy>VnXj0tg*gmel!=c&Xr˕*҉b$5]ʹe+5@[OfuIqd$Ͽu~N'%|\ tzVw{޻p$ 6bTȰr|%0nHt?Vu#5SiSĊޜN>}jrcCq/Q|v}(ׅO9؍:" Hj2fr_0: 0nR=〓MGs>ZViSk7~Xt"+Ԅ߇,wTȀKs{0>wF?7[JaO?D?5bRͰ»(R2:hRK ^mH}H Uйi.ڢK FA[~J9&nga຦**:A,%^fըNȚOX1b:#-7٦?nuם\"Ǒ8k1] G:k:F?kuyKM}`w} EQ3B'-4?\ȗlޑ[gيƠˑ4pzzuvI6(rnDyMBHBv.cR Yrc7z3wbHyeřJj"Mޕ;-%3ca[45(qbsa#~.EIv6R h뽱VLۺ9<зf vsΨn֚~|0ޯIįTz0a;؊Uru% cwЏ}8.y`.EQFVOdD#y^Fe+6KVc- 6g+w>`aڷ .B9~V78};^_w7 ls1MSRJuBw̎Gu[5MF۲?R$uB2z׽G^fRi˘_1w.Ql7 (iI:#Fr1SdoNײ l^(+sd$7.h)}^9{ka [0u $$(Lk_q#Y8hN:" ](ຍu-+{!>r@+'K)vT#?_F$/]hU`sp_=RkU:XojTSZZǮ!6"j Oo5=~x).2f?N,遨y 9#ai#b7R~_vijynpLqH_(;9ݶ:tZӤ'X0򂡲yrmt}xXz0?av&ÍޤHU"0@X'k/ݰf7n(>D %fh>T@]mXȹqpAwWIlD%̩Vf0S4O@id,9yOJ6Nm@?= 敼uín*EYm?WJ2FҥLRjXk*͈}Ձ4 m7 O}u[88gR] ު̝KYZlgM@/hu~V D@_^ЇqW0sÕlŔێ !tC6N2CW3 Fm+zp$*OD&Kd0ў61)2JXqĶn;>|1 `kY45!?WB@'{aOOowW)Z3\tzJG͸ ZzeB<觻5̟Pb բ$D}xBEkr5ׇjQ[^抻 X+cbFݒUYzyпn.>eOBf;_]w0|ɿS@wń.@+12vRb; DbS]ՍJ;)[HI8ccSwI7lovtqSSܣ$]8o\+]4~߀[*oU"S _ ktuQ]q|s}&_'w/]b+_UCSgD'JTJX~ o/7YzlMxm6i?<ҵŅtA{R#,lTkU{N!."§ҷ"!)`6 ȦaE}v8@N(ʴRV~PoWn'=WZz*2" ),)c&mU+=b >5 yeIc\Jek 7hmQ4:#&x'fGKr:jTY9R\eCҔQa,4e+ %U^+{I3]꣼ .sdDFm* ń?5KX шA떫\|a\bKuuLg$!-uU{.u~v62DHNeXw}ڂvIe{qN-fV_mDډH*Y.=A| zx (Mv<bg*2 ɠAn7vqU%΅eucF r}s¶Z7R/2Nd 2ݙI }Ne@_VXwhq뛙f4WY3Gó⍊J耠yt\ް:=u~kۿ`;r4 }e.z)~6X_C̔Z=S=}P?Ɍ9 J}/á/v]C)Z{57LWSE:{TD{>q_=̗ "rܛqn%k]FخXT ::/ݸK9bɊ=^[M T7-t)hEt%W\Ƹ_÷kAEo<\qLzoGgK|o!=l$}D0 z|/qRU\hj\?A=nw^p \Mɱ e/zμ0Bj;w4^J~r%}3mPU8:FZR9jMBڃ.]+tޣh4 EKr_Q-B%sOa{3"J>2v1eR|n:-Aɹ_cVK旱_tW1"eeƚl﷿-JrI =+uml }ט^L9?8Sjqlww8mm:Tnw|+-Mz.qxnp0 ZКTYrnOknu{ܑh[z+(6Ls3'߱#yܮ"t1aC "2Vg~0] $wۊE"6bOT0#gs"g)ko!Q2v+]ՃSU=0GxڊRoGLyl׀L{DZSy==FA /^8(oQ EC"Nu$b3E;vax fy|OidJff>-@,K:qgN:#@iZ P8[^#f mW@]UU+e0dׁ:SIׅU:{Sid^MB蚔/I[nK띋qYtPzTvO1uN\5?64{ӂ+*NUȶ@@.N"!=x 9JG9wocIPo?J>Ho7,ݾܰP7Ձ>(z ʑim`Py<\~?dAezͬukA2[:}.<"ǘZ]6"zNXǑӧEݻn΋޶ Pi "4Y{VB5EM~MfȮ@JpgKv083LkhX^{./͵^Sc3n0Q8_=gv Ȫ.ÜU\l'ZgvzG>QVAn&CWyMH 8Y[eKܹz%_j $} o+R3Oc՝7n EZlyPR2;#g+"JՅېq5ᆽz܍#z=R֦OdU7)-.+"6 AX3u5Bv"mEJ⯃X7DRel]e! (g;84'E\^@DXә%{VP];fEcfi Zv~b_^:~f8w+'~eǧ8yK6e^ݫ >ztjeBzbХuVut\kzjEmsYfFp:zܗSCrY5[·GK/t!J]D?iD(7Yh2^4f^l,}ѭ>XѕxJfK/Bnw!ՅnClΪY:=ic}U5 fAWc(]eO$G|Afsv/8@ h!;/Ծ#ρ0.[ƪb&+^).-K[_C=_31P4GL_ ,xQ/UD}ϙӱ-p (6ycәPpǟD\Zvbt+짋> 1[-K{Z5âpuhȷ͟{ -=Z^Ƣ#u]W^ v>~jՓuNy8t.Sxff]V!$ |nW8w(Kg #hsvv6b}q $\|~[M!qM 5-]&u}] -FC[[5Y;K!(E\pNhݠo<lKKS!3/Eӂj|/<55q3\&cHiO>1<ěy)8p66;tl~^yNee(φ[ɕVuj=JpTkH71J;,K ~H3ӸL3)?UUOY14@zaߠzSV'4@  Ы?zcX7h^ êAOh^V W=eBǰo)?UUOY14@zaߠzSV'4@  Ы?zcX7h^ êAOh^V W=eBǰo)?UUOY14@zaߠzSV'4@  Ы?zcX7h^ êAOh^V W=eBǰo)?UUOY14@zaߠzSV'4@ ?EIENDB`DdlA5P  C A? #" `2 hcol> l+{~`! hcol> lnFزx| \VUZzzZ~LefRKtQeG.`!ހqQ!%T2'ϔH9gQXt;sICdtz[?s~/5+iܢO ._ƹf"q{%i7r|3zAC :[h<>Lg:IN1JY)b'HKLXIKp!]2u8ލC1ރѴi-fEkh.6Zz,#K0);I&>Cfdn%n*wH=~Ld5^fkHBr=6'n^sKr $kH֑1:n%I2<&u=/#<iI BJo%K&RגuבmJ_k-cb3d.$~x7 /1_F#Y2L;:Uwl2<&v2Ecy$I4g|!+i| -lHSx_DxHZ#2<&?Ly ?Csu=WLMڷNM7g:J2<_/fyRgϞ ?ٹ_^KXn#B=G3tA Zhz54C=Hf~GT+Q=KVR|/QcbsT=]*bV Zh+(ڠ5S!ttʦj-U54_5ZĮPDsTdON!)I|ݤZJwHVg}2[Dī U# TDeGRDJr!7:bZ6Zľted'[e#( FrB'*4ȵI:*kHc+l-cbgH$r?.72U+/S:e4J%1c/!U&wyrexLlEH tASB%+YOy,xMcd!1wɇjB7'VN}CSkS7t;d'm-Y$[im-s_Y@c.!0[ I0FK~]M0<&6A%j0a"DG1ۡ3}:qC1uc={,['Y fe l>,X1f!{Э5~dk`q+ G uf]|FW@\łp &, kkg|{?6GX#jX)[Vet3_öqm|ۯ51y6gaOx~dcG@#n}?qo*m1D!/ټA)5A _k 4g]k_12348777283H觜rrOle ObjInfoCorelDRAW~RIFFvCDRAvrsnLISTldoc pfrdmcfg2 Q-p8W|?|??8888 7c' LastUsed/6a.?AVCLastUsed/6a.?AVCLastUsed/6a.?AVCdd@@@@?}4V?}4V?C:\Corel\WebSite\LastUsed/6a.?AVCDocTemplate@@/6a.?AVCMultiDocTemplate@@/6a.?AVWDrawMultiDocTemplate@@Initializing the App Recovery systemPLGT3280.DLLcoreldrw.exeRAVE.exedrawintl.dll.tlbDrawBrowser\Ravens2 =ans2 =ans2 =ans2 =ans2 =ans2 =ans2 =ans2 =ans2 =ans2 =ans2 =a(@ptrtVte56LISTfnttfontX@ ArialXwc8_X8_WesternfontX@Symbolwc8_X8_SymbolfontX%@Arial Unicode MSX8_CHINESE_BIG5fontX@Arabic Transparent8_ArabicfontX@ Arial Transparent8_Central EuropeanLISTRarrtarrwF`B DŅHXoHp_p_Xo/LISTfiltLIST0filcfild$`dX|@<LISTfilcfild0LIST0filcfild$HLISTTotltoutld??d@<outlh!2d??outlXd??d@<`outlH d??d@<LISTstlt`0H X8%JJJ@@@8%$$$@@@%JJJ@@@H@B@B@B' - c9 V? c5  ncd-Default Artistic Text8H -Default Graphic XX-LISTuil LISTbpageflgsbbox@_ ( obbx @_ ( @_ ( LISTgobjLISTllayrflgs LISTTlgoblodaGG$ 04@G.GuidesLISTllayrflgsLISTTlgoblodaHH$04@H.dDesktopLISTllayrflgsLISTTlgoblodaHH$04@H.dLayer 1LISTjlayrflgsLISTRlgoblodaEE$ 04@E.GridLIST0lgobloda$$ $.LISTBpageflgsbboxVte5obbx Vte5Vte5LIST2gobjLIST\layrflgs LISTDlgobloda88 (,8.LIST\layrflgsLISTDlgobloda88 (,8.dLISTlayrflgsLISTDlgobloda88 (,8.dLISTfobj spnd5flgs bbox03obbx 0303usdn4LIST8lgobloda <`lptx| d..>J8`ftil0??LIST\trfltrfdPPPw?_HA?l4Atxsm5?_HA?l4AUKlevel 3txtjLISTnobj spnd5flgs bbox/Y 3 obbx /Y 3 /Y 3 usdn3LIST8lgobloda <`lptx| d..>J+@,`ftil0??LIST\trfltrfdPPPw?yGA?8{#Atxsm5?yGA?8{#AUKUSlevel 2txtjLISTfobj spnd@5flgs bbox/ 6;l3Dobbx / 6;l3D/ 6;l3Dusdn2LIST8lgobloda <`lptx| d..>J@<<`ftil0??LIST\trfltrfdPPPw?yGA?Ɛ.Atxsm@5?yGA?Ɛ.AUKlevel 1txtjLISTtobj spnd4flgs bbox/Rte5+bobbx /Rte5+b/Rte5+busdn1LIST8lgobloda <`lptx| d..>JOO`ftil0??LIST\trfltrfdPPPw?GA?f4Atxsm4?GA?f4A UK zero-leveltxtjLISTtobj spnd4flgs bboxB6obbx B6B6usdn0LIST8lgobloda <`lptx| d..>Jbc`ftil0??LIST\trfltrfdPPPw?xF&?Atxsm4?xF&?A UK T, 3 -T, DtxtjLISTHobj spnd84flgs bbox"M# obbx "M# "M# usdn/LIST8lgobloda <`lptx| d..>JuHv`ftil0??LIST\trfltrfdPPPw?IEAA?  Atxsm84?IEAA?  AUKTtxtjLISTHobj spnd3flgs bbox  u  obbx  u    u  usdn.LIST8lgobloda <`lptx| d..>J`ftil0??LIST\trfltrfdPPPw?0&A?  Atxsm3?0&A?  AUKTtxtjLISTRobj spnd3flgs bbox~#&%d obbx ~#&%d ~#&%d usdn-LIST8lgobloda <`lptx| d..>J@Ȝ`ftil0??LIST\trfltrfdPPPw??AA?'+Atxsm3??AA?'+AUKF,TtxtjLISTRobj spnd03flgs bboxod obbx od od usdn,LIST8lgobloda <`lptx| d..>J`ftil0??LIST\trfltrfdPPPw?o3A?'+Atxsm03?o3A?'+AUKF,TtxtjLISTHobj spnd2flgs bbox;G ד obbx ;G ד ;G ד usdn+LIST8lgobloda <`lptx| d..>J’HÒ`ftil0??LIST\trfltrfdPPPw?v A?'+Atxsm2?v A?'+AUKFtxtjLISTbobj flgs bbox2t2obbx 2t2E6ta/usdn*LISTlgobloda4P\`dhl d..JՒbDH LIST\trfltrfdPPP w?H4?LALISTVobj flgs bbox'{P f8.wP obbx '{P f8.wP '%z f8.& usdn)LISTlgobloda0HTX\`dd..Jp֒?DLIST\trfltrfdPPP w?'7A?"ALISTbobj flgs bbox{P wP obbx {P wP S M usdn(LISTlgobloda4P\`dhl d..JגDH LIST\trfltrfdPPPw?~@?"ALISTbobj flgs bbox@+<obbx @+<+usdn'LISTlgobloda4P\`dhl d..JْDH LIST\trfltrfdPPP w+?=A?|-ALISTbobj flgs bbox B>obbx  B> usdn&LISTlgobloda4P\`dhl d..JhڒDH LIST\trfltrfdPPP w?+A?-ALISTbobj flgs bboxDBv >obbx DBv >Dv usdn%LISTlgobloda4P\`dhl d..Jے23DH LIST\trfltrfdPPPw?V?-ALISTVobj flgs bboxr[@(r obbx r[@(r [>R usdnLISTlgobloda0HTX\`dd..J(ݒ"D#LIST\trfltrfdPPPw?8?-ALISTVobj flgs bbox7nt(r obbx 7nt(r vwqt^ usdnLISTlgobloda0HTX\`dd..J`ޒ=7 DLIST\trfltrfdPPPw?$?:,ALISTVobj flgs bbox=yP =Tobbx =yP =T yP xgTusdn"LISTlgobloda0HTX\`dd..Jߒ DLIST\trfltrfdPPPw?h'?"ALISTVobj flgs bbox=n=(r obbx =n=(r "nzg(r usdnLISTlgobloda0HTX\`dd..J DLIST\trfltrfdPPPw?d'?:,ALISTobj spnd/flgs bbox'aWobbx 'aW'aWusdnLIST8lgobloda <`lptx| d..>JXP`ftil0??LIST\trfltrfdPPP?'7A?{2Atxsm/?'7A?{2AXUK@öUS$USt13;0.80txtjLISTobj spnd.flgs bboxHy]obbx Hy]Hy]usdnLIST8lgobloda <`lptx| d..>JXP`ftil0??LIST\trfltrfdPPP?X-A?51Atxsm.?X-A?51AXUK@öUS$USt11;0.50txtjLISTvobj spnd`.flgs bboxqyظ|obbx qyظ|qyظ|usdnLIST8lgobloda <`lptx| d..>JXP+`ftil0??LIST\trfltrfdPPP?A?51Atxsm`.?A?51AXUK@öUSt9;0.90txtjLIST|obj spnd.flgs bboxX֋ ._ obbx X֋ ._ X֋ ._ usdn!LIST8lgobloda <`lptx| d..>JXh>>`ftil0??LIST\trfltrfdPPP?X6A?"s)Atxsm.?X6A?"s)AXUK@öUS$USt6;1.0txtjLISTobj spnd-flgs bboxy֋ ._ obbx y֋ ._ y֋ ._ usdn LIST8lgobloda <`lptx| d..>JXhRR`ftil0??LIST\trfltrfdPPP?n?"s)Atxsm-?n?"s)AXUK@öUS$USt5;0.80txtjLISTobj spndX-flgs bboxpJBobbx pJBpJBusdn$LIST8lgobloda <`lptx| d..>JXhff`ftil0??LIST\trfltrfdPPP?0?(AtxsmX-?0?(AXUK@öUS$USt7;0.80txtjLIST|obj spnd-flgs bboxp֋ a obbx p֋ a p֋ a usdnLIST8lgobloda <`lptx| d..>JXhzz`ftil0??LIST\trfltrfdPPP?0?"s)Atxsm-?0?"s)AXUK@öUS$USt4;1.0txtjLISTrobj spnd,flgs bbox SYobbx SY SYusdnLIST8lgobloda <`lptx| d..>JXh`ftil0??LIST\trfltrfdPPP?@?{2Atxsm,?@?{2AXUK@öt1;0.90txtjLISTVobj flgs bbox kobbx k) 1kusdnLISTlgobloda0HTX\`dd..JҎ.qDLIST\trfltrfdPPPw?2)A?3ALISTVobj flgs bbox obbx  1 usdnLISTlgobloda0HTX\`dd..J.q.qDLIST\trfltrfdPPPw?H#A?3ALISTVobj flgs bbox4obbx 4"4TusdnLISTlgobloda0HTX\`dd..J0.qDLIST\trfltrfdPPPw?b/A?3ALISTVobj flgs bbox5lobbx 5l5"musdnLISTlgobloda0HTX\`dd..Jh.qDLIST\trfltrfdPPPw?ԳA?3ALISTobj spnd*flgs bbox)9T+@obbx )9T+@)9T+@usdn LIST8lgobloda <`lptx| d..>JX覓pHftil0??LIST\trfltrfdPPPw?=A?Ɛ.Atxsm5*?=A?Ɛ.AXUK @USh! h!  p14=( F, F, ..., {0.80}, F)txtjLISTobj spnd*flgs bbox M*obbx  M* M*usdnLIST8lgobloda <`lptx| d..>JX8Hftil0??LIST\trfltrfdPPPw?+A?v.Atxsm@*?+A?v.AXUK @USh!USh!   p13=( F, F,..., F, {0.50})txtjLISTobj spnd@*flgs bboxDM| obbx DM| DM| usdnLIST8lgobloda <`lptx| d..>JX0ϓϓHftil0??LIST\trfltrfdPPPw?V?v.AtxsmH@*?V?v.AXUK@USh!h!  p12=( F, F, {0.90}, F, ..., F)txtjLISTDobj spnd)flgs bbox' - obbx ' - ' - usdnLIST8lgobloda <`lptx| d..>JX8Hh!ftil0??LIST\trfltrfdPPPw?'7A?#Atxsm|)?'7A?#AXUK @USh! h!h!&       p23=(F, F, F, F, F, F, {0.90}, F,F, F)txtjLIST8obj spnd)flgs bbox ) obbx  )  ) usdnLIST8lgobloda <`lptx| d..>JX`Hftil0??LIST\trfltrfdPPPw?~@?#Atxsmp)?~@?#AXUK@USh! h!&     p22=(F, F, F, F, F, {0.72}, F, ..., F)txtjLIST>obj spnd8)flgs bbox5$F]xobbx 5$F]x5$F]xusdnLIST8lgobloda <`lptx| d..>JX:(;Hh!ftil0??LIST\trfltrfdPPPw?3?nAtxsmu8)?3?nAXUK @USh!h!'       p21=(F, F, F, F, F, F, {0.72}, F, F, F)txtjLIST*obj spnd(flgs bbox17 Qk obbx 17 Qk 17 Qk usdn LIST8lgobloda <`lptx| d..>JXTTHftil0??LIST\trfltrfdPPPw?3?#Atxsma(?3?#AXUK@USh!h!#    p21=(F, F, F, F, {0.90}, F, ..., F)txtjLISTobj spnd(flgs bbox5$Mobbx 5$M5$MusdnLIST8lgobloda <`lptx| d..>JXlHmHftil0??LIST\trfltrfdPPPw?3?v.Atxsm4(?3?v.AXUK@USh!h! p11=(F, {0.90}, F, ..., F)txtjLISTobj spnd0(flgs bboxV$obbx V$V$usdnLIST8lgobloda <`lptx| d..>JXHHftil0??LIST\trfltrfdPPPw?\A?f4Atxsm0(?\A?f4AXUK@h!p0=( {1.0}, F, F,..., F)txtjLIST\layrflgsLISTDlgobloda88 (,8.LISTlgobloda@@$04<@d . LIST\trfltrfdPPPww??sumi8Vte567)`DRAWBITMAPSrrSTREAMSLISTSTREAMSLISTDRAWPATTERNS! rrSTREAMSLISTSummaryInformation(_1206771831'$FrrOle '  ( 4 @ LX`hp|x CORELDRWx1@@@7Ea@U3aCDRACompObj#%(fObjInfo&*Equation Native +#_1206771801,)F@A r@A r FMicrosoft Equation 3.0 DS Equation Equation.39q0   FMicrosoft Equation 3.0 DS Equation Equation.39qS      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQROle ,CompObj(*-fObjInfo+/Equation Native 0n$ 1forj=i0forj`"i{ FMicrosoft Equation 3.0 DS Equation Equation.39q@l _1206771560.FrrOle 3CompObj-/4fObjInfo06Equation Native 7#1TabletkSummaryInformation(38DocumentSummaryInformation8@ DZ@#.xA+++֦u~_fqa 2<]f4gY*}sd/Vӥ+7^|]+mpaix 'ku1ؙ6YzEô vNO\h`py8뒞 [*X^5ӊR VlqVխVaUuʷ*NEqYlFZ Pk ZBܯq: }|+eURd+exL S}jR+arTY _́RժBskEk[j]EL۠/ȱP'>t-V]П/ߋCK jcgu%6[ $8"E> Lĝ(f0m>]a!RگR^Jhi`^|NvBG˴ MzφԞ_Mܴ v9 ?{6Fw˳>`0ߺ0 H`5Yia%­V̰ȱ ߲:YGXexLlraЪa4jN@PV\eA51`8T(*ȅ"Ȁ, `>gBz͹^shm-cb &@4x ^FS8‚4~u`0 nQ^pH[6vlc#ya= ,Y<,b%A9,G<2<&f*Hd+!mqf;!4ƳakHzH`0\62<&6c= ,JgC;Hkgk.ZX0eE38dk[&[ ,ZfZleYml>]Y˷a9Ͱ٭'ev6[mzޢ9v] g>淂׀i4X;tNamk0m\_0ܺ[ỹOX`P* "Hq0M̀X "RDd /X@JbKV# ̀p D9jqp5c=XB@ Qb>DZ.O7ЂЄ>h\ 4HEPP5Vs6kZ2<&6Rq$a0P+ 'l a)6- 0Q0!2<&5=6, ׳8l`IXRq,E#l>ż< }Zh{a:™YƱZ1u`;4Ƴvc$k)l-&:q.˷ fGnE" 30΂,\,b%,Jb"),PD2!&q2<&M ,M,_lc>JkzZӵUXx5@,gYZgv+gKrDA>gX/jA_1dcc|65L۠o߱|9vR&r,IL*ZR,H('Y"ZcXiJH&41Ɗh)KpZ B ѣi exL4.hȦD> Eڍe(XJڣ(:.HdP1FZĖ,XD[06a6m4ZӨlY\a1]95vnkNSq$M¡4HТM(H-xzJ:HҩIm-cbzl & $HT]L'OL:k8L]sFTIƒ$Mzn9ҪJ]kS[H;%8quG1ZgvHW1 #m8.J O]'Ѵ W"]ߨ!~o]'Ѵ rswŸoWbQxƆ_ g-.4mzH}@x?? M۠/+p|7|]|XP]y*DLUfa* gI @,SKgZuL p*!j ^j2Ta{5VQyvzDUu G\ =rn5E.WX(gasd.b,e->9m18[NDS!p8Vvs5n%He1:n$fkV>^>#>E6DYgR- O~Ttrd6ϳ ̻O5$~͇|U)Q<^eGT&qT[k I CJ:_p]}y]W }]cϼOtoY"/w_y7? ̌a\Fϥv*wW90h'BR?IV K:ΟMvR9ɪYs&Ӭ0esjP;5cgIPw6~1#XF53:M 33G̼Iwvgf޻w],g+{%:+xB>5-;{8j _ܧ ~9IlEK??%;eFylA&{s~߻ 'Yoy5to$~nCC#zl̿sOWd^d_ۯ}̋R&"؏|N^e;̎k^aNFy9ѺaAZ ܧn-V}jL_Sd,dW߈[u~G_f_Cʕ9盗C!?I5!iqfNb7\rn>m%9lA;DZfX3'{ kd,p2rIּ\}8Nj?zߙ ;& n)_:ˠ` Բ3 iE~ .ro Ͼ֟;!oC^eI?F޹跧N}'BB2gĻ蓂r'y[g*BINVbw5j^?mJPт_}%WNOo+Gm~W9[Kk-U;[KHZJW:<Μy`2ڟI.Zֿt _mޯ8Nc,̣;Ff^X4&a +J_oӘZ5MTbNfEP|mi2mkfjtJ3M;dhHܳ.Yv/9{}!s)G/Uaޏݚ>Ǿ??!^Uёtu `=^l3Մ\cjBLMg Uy}|r \P54ؕ߭WWUE6U@&m"Ϭ*hD\3bBrBv쾧D A&PT&DFݩ+ԇ#C@a}$341~Iֈa^-ʪKvPh u e09YBfXV#ՆrqP!6jC׆bޡp"'<e[\NՠXY5 t;z0IeHN(wX.&VO^UDl~\5E ?y49‚mkB5|$VE)KNSsT2cE159.NDR|\e\W*2egZK蒤9{Z)rn v=Q: f&0M(,u aO0hOU[g뱮<& fω5l*Fq}\=XQ2a@Շ$V@=9?8C}C&zq~@Fo$)S7ZbǙP|lҬ(U'gĭЯ׮TzD4^DEfՀ)Vq?"y]}d),>i_Dd;yVykr\R^ӓShuYx=.k!on2U]dTtlޭTW, tT=#dWsS@"o%T =JWJT\ۣe} FvLh(!G4szH#N/ )$7'!m@+wo7`@p93-ď嬴W;Jh3gJNCN٦NI Ўm@+B~Xwooo a&`?:ϴ0J_wHƐ/tDYO5b?C>ӿn nn>[S@SsҿF뜿via ^s#ZZ~MZMĢq nܽ>[/@%7{`> h a0h!0nw ऻqV  gc0Y]|{%֋|/q^٩wы.kh1J(23fftǍ.gp@p@;p:8'SiC n{Y}9]1r*oʱ2N^^k 89}-$Ϻ굅992M6HyE#)߽q۳n|~zɅo!I䔧;2^fed{Gqڹ{&gDdTJ  C A? "24yS ~`!4yS   ȽXJkxcdd``d!0 ĜL0##0KQ* Wä2AA?H1Zc@øjx|K2B* R\``0I3cDd<`J  C A? "2bޫ?Fv=SaIO~`!bޫ?Fv=SaI` X gxڕ?K@Ɵi[:8 nJu:؂R#P.PA:A_ݿT'x%`%>< =i1,W1]1EظKrlQT3f[Qp@ݓJU =~:TB3vVŰ}'[h띔hbq,gE%T|b*~̍K)OLf<6o4زp4/RjvG^&]2C+4߫4vhi憨'"yG)K?BgeIwG^|`듵K(oIy_Jly>:©k0,bt 4|1gDdTJ  C A? "24yS ~`!4yS   ȽXJkxcdd``d!0 ĜL0##0KQ* Wä2AA?H1Zc@øjx|K2B* R\``0I3Oh+'0   4@ \ h t HAUTHOR GUIDELINES FOR ICONIP'02-SEAL'02-FSKD'02 PROCEEDINGS MANUSCRIPTSUTHyeoeo Normal.dotE slobodantE3obMicrosoft Word 9.0O@2 @򾠟@leh@br^՜.+,D՜.+,@ px  NTU`0s HAUTHOR GUIDELINES FOR ICONIP'02-SEAL'02-FSKD'02 PROCEEDINGS MANUSCRIPTS Title 8@ _PID_HLINKSA@)Z#mailto:nikola.pavesic@fe.uni-lj.si3Fmailto:slobodan.ribaric@fer.hr-iB3 fred&clyde  FMicrosoft Word Document MSWordDocWord.Document.89qb i0@0 Normal_HmH sH tH R@R Heading 1$<@&5CJ KH OJQJ\^JaJ T@T Heading 2$<@& 56CJOJQJ\]^JaJN@N Heading 3$<@&5CJOJQJ\^JaJB@B Heading 4$<@&5CJ\aJF@F Heading 5 <@&56CJ\]aJ@@@ Heading 6 <@&5CJ\aJ:@: Heading 7 <@&CJaJ@@@ Heading 8 <@&6CJ]aJF @F Heading 9 <@&CJOJQJ^JaJ<A@< Default Paragraph FontBC@B Body Text Indent$h`ha$<T@< Block Textx]^*B@* Body Textx4P@"4 Body Text 2 dx6Q@26 Body Text 3xCJaJbM@Bb #Body Text First Indent, prva uvlaka `|N@R| 'Body Text First Indent 2, prva uvlaka 2$hx^h`a$`R@b` Body Text Indent 2, uvlaka 2hdx^h`S@r` Body Text Indent 3, uvlaka 3hx^hCJaJ0"@0 Caption xx5\*?@* Closing ^,@,  Comment TextL@ DateJY@J  Document Map-D M OJQJ^J4[@4 E-mail Signature,+@,  Endnote Texth$@h Envelope Address!@ &+D/^@ CJOJQJ^JaJ>%@> Envelope Return OJQJ^J, @, Footer ! !.@".  Footnote Text",@2, Header # !2`@B2 HTML Address$6]Be@RB HTML Preformatted% OJQJ^J2 @2 Index 1&8^`82 @2 Index 2'8^`82 @2 Index 3(X8^X`82 @2 Index 4) 8^ `82@2 Index 5*8^`82@2 Index 6+8^`82@2 Index 7,x8^x`82@2 Index 8-@8^@`82@2 Index 9.8^`8@!@b@  Index Heading/5OJQJ\^J,/@, List0h^h`02@0 List 21^`03@"0 List 328^8`04@20 List 43^`05@B0 List 54^`20@R2 List Bullet 5 & F66@b6 List Bullet 2 6 & F67@r6 List Bullet 3 7 & F68@6 List Bullet 4 8 & F69@6 List Bullet 5 9 & F:D@: List Continue:hx^h>E@> List Continue 2;x^>F@> List Continue 3<8x^8>G@> List Continue 4=x^>H@> List Continue 5>x^21@2 List Number ? & F6:@6 List Number 2 @ & F6;@6 List Number 3 A & F6<@"6 List Number 4 B & F 6=@26 List Number 5 C & F d-@Bd  Macro Text"D  ` @ OJQJ^J_HmH sH tH I@R Message HeadergE8$d%d&d'd-DM NOPQ^8`CJOJQJ^JaJ4^@b4 Normal (Web)FCJaJ6@r6 Normal Indent G^,O@, Note HeadingH4Z@4 Plain TextI OJQJ^J(K@( SalutationJ.@@. Signature K^FJ@F SubtitleL$<@&a$CJOJQJ^JaJL,@L Table of AuthoritiesM8^`8D#@D Table of FiguresNp^`pN>@N TitleO$<@&a$5CJ KHOJQJ\^JaJ H.@H  TOA HeadingPx5CJOJQJ\^JaJ@ TOC 1Q&@& TOC 2 R^&@& TOC 3 S^&@& TOC 4 TX^X&@& TOC 5 U ^ &@& TOC 6 V^&@& TOC 7 W^&@& TOC 8 Xx^x&@& TOC 9 Y@^@.U@. Hyperlink >*B*ph>V@> FollowedHyperlink >*B* ph\O\ heading1!\$$$ @*$a$5CJOJPJQJtH0O0 p1a]$a$OJPJQJtH:'@: Comment ReferenceCJaJTOT heading3_$$$ @@*$a$5OJPJQJtHPOP reference`$^`a$CJOJPJQJtH@O@  Balloon TextaCJOJQJ^JaJ&r'DD&r " O x YZ|Sy+ x Az#X^.Quf"""#*#+#,#Q#R#$(%(-(h(u((((((((((((((()))')4)5)C)P)Q)`)m)n)w))))))A-B-D-E-~--s/000000&1112223333346:^=>?@*ABAC'DIDDDDKEOGHHJzN{NNNN+ObPPP8QQQQXRRwS6TSU0V1V;V  3 7 X a \ i  !#46rtprxyBC\^`b"$J!L!!!!!!!!!##G#I#I%Q%V&X&[&]&+,--////m/o//0(0,0.060;0E0G0Y0b0p0r0{00000 11X1\1^1f1j1u1w1122H3J333555555:6F6u666667C:E:e:f:;;;;G<I<========H>L>>>u@@wAxA~AAAAAACBKBBBBBCCDDGGGGIIIIIIII9J;JJJJJJJJJMKXKNNQOSOOOPPPPPPPPPPQQ+R-RSS SS7S8SVVVVVWWWXXXXXYYYYYYYZZSZWZZZZZZZZZ[["[&[@[D[O[W[s[~[[[[[^\_\\\\\\\]]]]^^^_ __&_*_5_9_M_Q_\_d_|______aaa!agayazaa;eAeqereeeeef fffffffffff;gFggg hhhhhhhhj jjjkk[m]m=nCngntnoopp5q>qqqqqqqqqrrrrrrrrrr!r"r#r'r  OT.3-[^_vxJO^$, "9 ; B!D!""E#G#T&V&M+R+++,,--M-}---1.6...//00000000111122336668899G:I:;;;;==>>q?z?TBVBQDDDDYKbKPPiQnQQQRRTTSUZUܶ!f@E*d Xll^`.^`.88^8`.^`. ^`OJQJo( ^`OJQJo( 88^8`OJQJo( ^`OJQJo(hh^h`. hh^h`OJQJo(@ OJQJo(@ OJQJo(@ OJQJo(@ OJQJo(@ OJQJo(@ OJQJo(^`o(. ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.^`o(.^`.pLp^p`L.@ @ ^@ `.^`.L^`L.^`.^`.PLP^P`L.@^`.^`o(. ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.0^`06o(()^`. L ^ `L.  ^ `.ll^l`.<L<^<`L.  ^ `.^`.L^`L.0^`06o(()^`. L ^ `L.  ^ `.ll^l`.<L<^<`L.  ^ `.^`.L^`L.^`OJPJQJ^Jo(- ^`OJQJo(o pp^p`OJQJo( @ @ ^@ `OJQJo( ^`OJQJo(o ^`OJQJo( ^`OJQJo( ^`OJQJo(o PP^P`OJQJo(h^`OJPJQJ^Jo(-h ^`OJQJo(oh pp^p`OJQJo(h @ @ ^@ `OJQJo(h ^`OJQJo(oh ^`OJQJo(h ^`OJQJo(h ^`OJQJo(oh PP^P`OJQJo(0^`06o(()^`. L ^ `L.  ^ `.ll^l`.<L<^<`L.  ^ `.^`.L^`L.))^)`OJPJQJ^Jo(- ^`OJQJo(o   ^ `OJQJo(   ^ `OJQJo( ii^i`OJQJo(o 99^9`OJQJo(   ^ `OJQJo( ^`OJQJo(o ^`OJQJo(~}|f>])R:!f@8EX66pT3F^5                  >D        h        6H        6H        &:        @| %(g(h(u((((((((((((((()))')4)5)C)P)Q)`)m)n)w)))'r@qeqeqeqe(&r@Unknown Gz Times New Roman5Symbol3& z ArialWSimSunArial Unicode MS5& z!Tahoma?5 z Courier New3z Times7Tms Rmn;Wingdings"1hGfI^0x=!0ds& 3QHPGAUTHOR GUIDELINES FOR ICONIP'02-SEAL'02-FSKD'02 PROCEEDINGS MANUSCRIPTSyeoslobodanCompObjMj