ECMS-based Real-time Optimal Energy Management For Commuter Hybrid Electric Vehicle

The research work of this dissertation provides a real-time optimization energy management strategy based on equivalent consumption minimization strategy(ECMS)for non-plug-in commuter hybrid electric vehicle(HEV)to decrease fuel consumption and satisfy the battery charge-sustaining constraints.A proper value of the equivalent factor used is crucial to the optimality of power splits,since for a well-tuned equivalence factor,the performance of ECMS can be close to that of dynamic programming(DP).For both globally suboptimal solution and implementable strategy,this paper presents three adaptive ECMS energy management controllers using real traffic information.Main researches are shown as follows:Adaptive ECMS energy management strategy based on traffic information statistics: by the statistical characteristics deriving from historical driving data,the infinite-horizon stochastic dynamic programming(SDP)optimization is formulated for finding proper equivalence factor according to uncertain driving cycles on a fixed route.And then,a stochastic optimization-based equivalent factor map function on state of charge(SOC)is obtained off-line by SDP policy iteration algorithm,which is a simple causal closed-loop optimal solution in mean sense.In the implemented ECMS-based power splits online,the equivalence factor is adjusted in real-time for accommodating changes in real driving condition,so as to achieve the near global optimal control objective that fuel consumption is minimized over the whole driving route.Adaptive ECMS energy management strategy with real-time traffic status identification: unsupervised learning,k-means algorithm,is utilized to analyze and catalog the traffic information.By the statistical characteristics deriving from historical driving data,the infinite-horizon SDP optimization is formulated for finding proper equivalence factor according to each category driving cycles.And then,a stochastic optimization-based equivalent factor map is obtained off-line by SDP policy iteration algorithm,which is a simple closed-loop solution with feedback map function on SOC and average speed.The results of the K-Means algorithm are selected as identification toolbox for real-time traffic information and the equivalence factor is adjusted based on actual driving circumstances.A real-time energy management of integrated multi-mode and adaptive ECMS via particle swarm optimization(PSO): Considering the different demand power and battery SOC state,HEV dynamic system will have different working mode.Control parameters of mode switching signal,equivalence factor of ECMS,and the engine torque and speed at the engine working mode are optimized offline by PSO with the equivalent fuel consumption as the fitness function based on the historical traffic data to establish their maps on the state variables of SOC and power demand.And then,these established maps are involved in the real time energy management for commuter HEVs according to the SOC and power demand.To demonstrate the effectiveness of the proposed strategy,the simulation results and comparison with some other energy management strategies are presented by implementing the strategies on GT-SUITE and MATLAB test platform.