Model Based Optimization of Supervisory Control Parameters for Hybrid Electric Vehicles

Supervisory control strategy of a hybrid electric vehicle (HEV) provides target powers and operating points of an internal combustion engine and an electric motor. To promise efficient driving of the HEV, it is needed to find the proper values of control parameters which are used in the strategy. However, it is very difficult to find the optimal values of the parameters by doing experimental tests, since there are plural parameters which have dependent relationship between each other. Furthermore variation of the test results makes it difficult to extract the effect of a specific parameter change. In this study, a model based parameter optimization method is introduced. A vehicle simulation model having the most of dynamics related to fuel consumption was developed and validated with various experimental data from real vehicles. And then, the supervisory control logic including the control parameters was connected to the vehicle model. This simulation environment was used as an evaluation function in genetic algorithm to find the optimal parameter values. Additionally, a distributed computing technology was used to reduce the calculation time of the optimization process. The optimized values of the parameters are expected to be a good start point of parameter tuning on real vehicles.