Dynamic energy management for hybrid electric vehicle based on approximate dynamic programming

In this paper, an approximate dynamic programming (ADP) based strategy for real-time energy control of parallel hybrid electric vehicles(HEV) is presented. The aim is to develop a fuel-optimal control which is not relying on the priori knowledge of the future driving conditions (global optimal control), but only upon the current system operation. Approximate dynamic programming is an on-line learning method, which controls the system while simultaneously learning its characteristics in real time. A suboptimal energy control is then obtained with a proper definition of a cost function to be minimized at each time instant. The cost function includes the fuel consumption, emissions and the deviation of battery soc. Our approach guarantees an optimization of vehicle performance and an adaptation to driving conditions. Simulation results over standard driving cycles are presented to demonstrate the effectiveness of the proposed stochastic approach. It was found that the obtained ADP control algorithm outperforms a traditional rule-based control strategy.