Simultaneous Perturbation Stochastic Approximation Optimization for Energy Management Strategy of HEV

Abstract

This paper addresses optimization for hybrid electric vehicle (HEV). This project is using a single agent method to optimize the power losses under a specific driving cycle which is simultaneous perturbation stochastic approximation (SPSA) based method. For optimization process, four gain are added in four main parts of the HEV system. Those main parts are engine, motor, generator and battery. These four gain is controlled the output for each component to give the minimum power losses. The design method is applied to free model of HEV by using Simulink/MATLAB software while M-File/MATLAB is used to apply the SPSA method. The result from design method achieved a minimum reduction of power losses compared to original system. Thus, the comparison of result has been done to show the different before and after optimization.