论文部分内容阅读
针对目前基于工况识别的混合动力汽车能量管理策略中工况识别算法的局限性和缺点,应用K均值聚类算法进行工况识别,并结合等效燃油最小能量管理策略(ECMS)实现对整车的能量管理。具体方法为:选定4种典型城市循环工况,根据等效燃油最小能量管理策略,得到4种典型工况下不同等效燃油系数与油耗之间的关系,根据所得每种典型工况均有相应的最优等效燃油系数和最优需求功率分配方式的分析结果,对某随机行驶工况采用遗传优化后的K均值聚类算法进行识别,获得随机工况当前所属的工况类别,结合所属工况类别对发动机和电机的功率进行实时优化分配。仿真结果表明:所制定的能量策略同未采用工况识别的能量管理策略相比,车辆综合油耗下降了7.47%,电池荷电状态变化更加平稳,而且能更好地维持在电池效率较高的区域。
Aiming at the limitations and shortcomings of current condition recognition algorithms in hybrid vehicle energy management strategy based on condition recognition, K-means clustering algorithm is used to identify working conditions. Combined with the equivalent fuel minimum energy management strategy (ECMS) Car energy management. The specific method is as follows: According to the minimum energy management strategy of equivalent fuel, the relationship between different equivalent fuel efficiency and fuel consumption under four typical operating conditions is obtained. According to each typical working condition With the corresponding optimal fuel efficiency and optimal power allocation analysis results, a random driving condition using genetic algorithm to identify the K-means clustering, get the working conditions belong to the current conditions, combined with The type of service under which the power of the engine and the motor is optimally distributed in real time. The simulation results show that compared with the energy management strategy without working condition identification, the comprehensive fuel consumption of the vehicle is reduced by 7.47%, the battery state of charge is more stable, and can be better maintained at higher battery efficiency area.