Research on the optimal power management strategy for a hybrid electric bus

Abstract

The power management strategy for a hybrid electric vehicle uses expert knowledge and experience to determine the control rules and thresholds, but it is difficult to obtain good performance. This paper presents a dynamic programming algorithm based on a powertrain system to find the optimal solution in the specific speed cycle, but this method cannot be applied to a real-time control system. Therefore, the optimal control rules and thresholds are extracted for the rule-based strategy; the simulation results show that the refined strategy can reduce the fuel consumption by about 14.41% compared with that obtained by the original logic rule control algorithm in the Chinese urban driving cycle. An online neural network energy controller based on the dynamic programming results is provided by choosing a reasonable network structure and control parameters to improve the control precision; the numerical simulation results show that the neural network controller improves the vehicle fuel economy by 15.05%, which is closer to the dynamic programming algorithm simulation result. This control algorithm not only can be used in real-time control and reduces the computational complexity but also allows new design ideas to be proposed for the control algorithm and the vehicle fuel consumption to be reduced.