Research On Adaptive Control Strategy For Energy Management Of Fuel Cell Vehicle Based On Power Demand Prediction

Because of pollution-free,high-efficiency,remarkable energy-saving and emission-reducing effects,the Fuel Cell Hybrid Vehicle(FCHV)is receiving worldwide attention.Energy management strategy is one of the key technologies for FCHVs,which has a significant effect on its fuel economy.Existing widely used energy management strategies of FCHVs are based on fixed rules,which lack adaptability to the complex and uncertain driving conditions,and cannot guarantee the economic efficiency of the entire vehicle.Some optimization control strategies such as Dynamic Programming algorithms,although can ensure its overall optimality.There are problems such as "dimensional disasters",make them unable to meet the requirements of actual on-board application of vehicle controller.In view of the above problems,this paper studies the adaptive control strategy for energy management of FCHV based on power demand prediction.The research content is as follows:1)The self-developed object-oriented vehicle dynamics simulation software is used to establish a fuel cell vehicle model.Based on the vehicle parameters of FCEV and design demands,along with dynamic indicators such as climb performance and economic indicators such as cruising range and fuel consumption,the parameters matching of key components such as drive motors,batteries,and fuel cells have been designed and completed.2)Aiming at the research of PFCHV energy management optimization problem,a fast dynamic programming algorithm is proposed.The computing time of algorithm is greatly shortened by reducing search domain and dimension.Based on the fast dynamic programming algorithm,the PFCHV optimal energy management control law is summarized and then the mapping relationship between the control strategy,the optimal SOC trajectory and the optimal FCS output power is constructed.3)A novel control strategy which is performed through three parts spatially and temporally is proposed.First is the whole journey reference trajectory of SOC.Second is a short-range neural network model which is established to predict the energy demand of the vehicle.And the last part is a real-time optimization adaptive control strategy.The experimental results show that the proposed control strategy can achieve near-optimal performance and significantly reduce the energy consumption of the vehicle.Simultaneously,the control strategy can meet the real-time requirements of on-board controller and can automatically adapt to the different driving conditions.