Predictive Control For A Hybrid Electric Vehicle Based On Driving Cycle Recongnition

Plug-in Hybrid Electric Vehicle (PHEV) has the advantages of pure electrics andconventional vehicles, and obvious advantages in energy conservation and emissionsreduction. Energy management strategy is the core of the development of PHEV, itsquality affects the vehicle performance directly. The paper regards PHEV as theresearch object, the research on the control strategy of PHEV was conducted, with theobjective of improving based on the future state information of vehicles provided byon-board navigation system. The main contents of this paper are as follows.①The mathematics model of PHEV’s power train were established,.In order toimprove computing efficiency of dynamic programming,the calculation formula ofengine fuel consumption rate,battery internal resistance and electromotive force werefitting out based on particle swarm optimization(PSO).②The fuel consumption of engine and power consumption of motor were unifiedwith price, the global optimization strategy of PHEV was put forward to achieve thelowest cost in every whole mileage.The additional cost function was add into theobjective function to reduce the additional fuel consumption caused by the frequentstart-stop of engine. The simulation calculation of global optimization under differentconditions of UDDS, NEDC,1015were conducted respectively. The change rules ofSOC curve under general conditions were able to be obtained by the simulation. thechange rules of SOC under general conditions and special conditions were obtained bythe simulation and analysis on the new constituted cycle conditions,such as the longuphill, long downhill and idle stop conditions.③In order to make predictive control approximate to global optimization controleffect,The vehicle working conditions were classified to general condition,transientcondition and special condition,recognition methods about future conditions wereproposed.Control methods about future conditions were proposed. The modelpredictive control strategy based on majorized SOC Trajectory under generalconditions, the fuzzy control strategy under transient conditions, and the controlstrategy based on rules under special conditions were studied. The predictive control ofHEVs can be realized by the alternate use of these three control strategies.④The predictive control model of PHEVs based on driving condition reco-ngnition by language C and MATLAB/Simulink was established, The simulate d SOC curve of model predictive control under several standard conditions canfollow the optimal trajectory of SOC better in the global optimization. The simulation analysis of predictive control under special working conditions indicatedthat all kinds of modes can be switched reasonably,in the early to charge anddischarge function can play under transient conditions,the engine and motor mostly work in the region of the low oil consumption and high efficiency, whichsuggests that the predictive control of PHEV the author put forward based on the recognition of working conditions can effectively improve the fuel economy.Finally, operation time of predictive control strategy based on the dSPACE real-time simulation system was tested, and the results have shown that the proposed predictive control algorithm can meet the requirements of real-time control.