To improve the control performance of traditional equivalent fuel consumption minimum strategy (ECMS) under the real complex road conditions,an adaptive equivalent consumption minimum strategy (A-ECMS) of energy management was proposed for a parallel hybrid electric vehicle,which can adjust the equivalent factor online according to the change of driving cycle.The significantly different characteristic parameters of driving cycle were extracted by statistical method.The driving cycles were classified with cluster analysis method,and the database of typical driving cycles was constructed.Then,the optimal equivalent factor for each typical driving cycle can be calculated.A recognizer of driving cycles was designed by learning vector quantization method,and its accuracy of recognization was proved up to 98.8％.The actual driving condition was identified to be one of the typical driving cycles by the recognizer,and the corresponding optimal equivalent factor was adopted as the optimization input of ECMS.Thus,the A-ECMS control strategy based on the driving condition online identification was established.The simulation results show that the optimization effect of A-ECMS is similar to ECMS under the given single cycle,the fuel economy of A-ECMS is decreased by 0.8％ and SOC increased by 0.13％.Under the multi-driving cycles,the fuel economy of A-ECMS strategy is improved by 4.18％,and the fluctuation of SOC is reduced by 43.26％,which can prove the superiority of the A-ECMS.%为改善传统等效燃油消耗最低策略(ECMS)在真实复杂路况下的控制效果,以并联混合动力汽车为研究对象,提出了一种依据工况变化在线调整等效因子的自适应等效燃油消耗最低(A-ECMS)控制策略.首先,提取差异化显著的工况特征参数,采用聚类分析方法来完成工况分类,构建典型工况库,计算出各典型工况对应的最优等效因子;然后,采用学习向量量化(LVQ)神经网络设计了工况识别器,经充分训练后识别器准确率达到98.8％;最后,在线采集选定的车辆行驶特征参数,将当前实际工况识别为典型工况库中某一种,采用对应典型工况下的最优等效因子作为当前优化输入,建立了基于工况识别的A-ECMS控制策略.仿真结果表明:与ECMS相比,在单一给定工况下,A-ECMS燃油经济性降低了0.8％,而电池组荷电状态(SOC)提高了0.13％,能取得近似优化效果;在多工况联合工况下,燃油经济性提高了4.18％,且SOC波动减小了43.26％,证明了A-ECMS控制策略的优越性.
展开▼
