Study On Control Strategy Of Hybrid Vehicles Based On Driving Condition Recognition

The sustainable development of domestic automobile industry faces thechallenges of energy crisis and environmental pollution, hybrid play an importantroles in solving this problem. Hybrid electric bus has been widely used,as a kind oftypical new energy vehicles, and made great contributions to energy conservationand emissions reduction. Research shows that driving conditions have great influenceon the economy and emissions of the vehicle. Energy management strategy is mostlydeveloped from a particular driving cycle, so it does not consider the influence of therunning condition. The adaptability of energy management strategy become worse inthe face of complex and volatile driving condition, so the hybrid bus cannot fullyexert energy saving effect,which result in higher fuel consumption. Therefore, thispaper takes the parallel hybrid electric bus of single axis as the object to study how toimprove the adaptability of hybrid electric bus under different driving conditionsfrom two aspects of driving condition and energy management strategy.Firstly, completing the algorithm of online driving condition recognition basedon the typical driving condition of hybrid electric bus. The library of typical drivingcondition used in this paper is established, through screening and correcting of theinternational library of driving condition. The characteristic parameters of drivingcondition are extracted, through the analysis of the characteristics of different typicaldriving conditions. Training and validation of driving condition online recognitionalgorithm is achieved. The algorithm can classify vehicle traffic characteristics as oneof typical driving condition by means of collecting data of vehicle driving and supplydriving condition to energy management strategy.Based on the above typical driving conditions, using the dynamic programmingalgorithm to obtain the offline optimal allocation trajectories, then completing the online application design and connecting with driving condition recognition toestablish energy management strategy. The offline optimal trajectories of energydistribution is achieved by applying global optimization program on6kinds of typicaldriving condition. then it is transformed into the rules of the online application byusing neural network method, so as to respectively set up the optimal online energymanagement strategy for six typical driving condition. Combining driving conditionrecognition programming with6kinds of energy management strategies to constructstrategy with high adaptability. Which realize automatic recognition and switching ofoptimal control algorithm under different driving conditions.Finally, the simulation have been carried out to verify the proposed energymanagement strategy under the regulations test cycle, the strategy based on singlecondition and the proposed strategy are simulated respectively under standardoperating conditions CCBC, UDDSHDV, C–WTVC. The simulation results show thatthe proposed strategy has improved significantly on fuel-saving rate and balance ofSOC and can decrease by1.99%~12.4%fuel consumption of hundred kilometerscompared with strategy based on single mode.The study of this paper shows that driving condition has a significant impact onfuel saving, using driving condition recognition can enhance the adaptability ofenergy management strategy. The traditional optimization methods have difficulty inapplying to the design of the real-time energy management strategy, due to theuncertainty and complexity of actual condition. This paper solves two key problems,the real-time driving condition classification and the optimal energy managementstrategy online. Simulation results show that proposed strategy has significant effecton exerting the potential fuel saving and improving the energy utilization rate.