engleski

AVL CRUISE Model-based Optimisation of Shift Scheduling Maps for a Parallel Hybrid Electric Vehicle

Determining shift scheduling maps plays important role in the development of both conventional and hybrid automatic transmission powertrains. In order to meet ever demanding requirements on fuel economy and performance, modern optimisation methods are commonly used for generating optimal shift scheduling maps. This study focuses on the development of a new approach for optimising shift scheduling maps of a parallel hybrid electric vehicle (parallel HEV) based on interaction between AVL CRUISE and MATLAB/Simulink software tools. A mild HEV model, given in P2 configuration comprising a Dual Clutch Transmission (DCT), is adopted from the AVL CRUISE model library. The shift scheduling problem is defined as a bi-objective optimisation problem (comfort measure vs. fuel economy objective), which is solved by using the evolutionary multi-objective (EMO) genetic algorithm NSGA-II implemented in the form of MATLAB script. A two-dimensional shift scheduling map is considered, with the vehicle velocity and the engine load signal used as independent input variables. Special attention is given to the development of a sub-tool aimed to generate proper initial shift scheduling maps in a straightforward way, which are then used as a starting point for an effective genetic algorithm-based optimisation. Also, in order to account for variations in final battery State-of-Charge (SoC) when comparing performance of different shift scheduling maps, a fuel consumption correction method is introduced. Running optimisations in the presence of two conflicting objective functions results in Pareto-optimal shift scheduling solutions, thus enabling a user to select an appropriate solution according to individual needs. The performance of HEV model with optimised shift scheduling maps is examined by simulations for different driving cycles, and compared with the performance of the original AVL CRUISE model. The simulation results point out that the proposed optimisation approach is effective in generating multiple, user- selectable, optimal shift scheduling maps, which give comparable performance to that of the original AVL CRUISE maps when the emphasis is on fuel economy.

shift scheduling optimisation ; parallel hybrid electric vehicle ; multi-objective genetic algorithm ; AVL CRUISE ; modeFRONTIER ; MATLAB/Simulink

nije evidentirano