混合动力汽车补偿模糊神经网络能量管理策略

Compensation Fuzzy Neural Network Power Management Strategy for Hybrid Electric Vehicle

摘要:

以上海大众汽车公司某型号混合动力电动汽车(HEV)的设计要求为基础,提出了一种基于补偿模糊神经网络的能量控制策略,并采用动态调整步长的梯度下降法加快算法的收敛速度.分析了样本数据选取、输入、输出模糊分割和模糊规则提取对控制器性能的影响.利用ADVISOR2002仿真平台进行二次开发,完成了基于补偿模糊神经网络的控制策略、并联电力辅助控制策略和模糊控制策略的仿真比较.仿真结果表明,基于补偿模糊神经网络的控制器具有较强的自适应能力,可以较好提高混合动力汽车的燃油经济性和排放性.

Abstract:

A power control strategy of compensation fuzzy neural network (CFNN) is put forward based on the design requirement of a hybrid electric vehicle (HEV) of Shanghai Volkswagken company,the proposed algorithm speeds up the convergence rate by adjusting step dynamically.An analysis is made of the influence of sample data selecting,fuzzy sets for input and output and the determination of fuzzy rules.Based on the redevelopment of the ADVISOR2002,a comparison is conducted on the CFNN control strategy,the electricassisted control strategy and the fuzzy logic control strategy.The simulation results show that the CFNN controller is of good selfadaptive ability and helpful to fuel economy and emission.