Pregled bibliografske jedinice broj: 136934
Expert System for On-Line Monitoring of Streeses State in Steam Turbine Casings
Expert System for On-Line Monitoring of Streeses State in Steam Turbine Casings // DAAAM International Scientific Book 2002 / Katalinic, Branko (ur.).
Beč: DAAAM International Vienna, 2002.
CROSBI ID: 136934 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Expert System for On-Line Monitoring of Streeses State in Steam Turbine Casings
Autori
Guzović, Zvonimir ; Šerman, Maja ; Matijašević, Branimir
Vrsta, podvrsta i kategorija rada
Poglavlja u knjigama, znanstveni
Knjiga
DAAAM International Scientific Book 2002
Urednik/ci
Katalinic, Branko
Izdavač
DAAAM International Vienna
Grad
Beč
Godina
2002
Raspon stranica
ISBN
3-901509-30-5
Ključne riječi
steam turbine, turbine casing, expert system, strength margin, useful life
Sažetak
The recent load increase in steam turbine plants and insurance of their high exploitation characteristics in the increasingly complex conditions of durable exploitation require automation of monitoring, diagnostic and optimisation of steam turbine operation. The existing in practice the most frequently applied method of monitoring of steam turbine operation (i.e. of temperature stresses estimation) has disadvantages, which are negatively reflected on the safety of the steam turbine operation. Today, the mentioned problems are the most effectively solved by using artificial intelligence techniques: expert systems with built-in fuzzy logic and algorithms for self-learning using neural networks. The basic idea of this work is that the simulations of total stresses which include action of temperature field in the steam turbine casings wall by means of user software are used to create a knowledge base of an expert system for operational monitoring of stresses state. By introduction of input signals and based on the knowledge base the system estimates during operation the current values of strength margin and the remaining useful life by one of the interpolation procedures between the most similar simulated operational situations. The proposed expert system can reach the mean value of estimation error less than the existingin method. Also the presented concept of expert system is applicable for monitoring the stresses state of all critical, thick-walled components of the thermal power plants.
Izvorni jezik
Engleski
Znanstvena područja
Strojarstvo
