插电式混合动力汽车车速预测及整车控制策略

Plug-in hybrid electric vehicle speed prediction and control strategy

中文关键词  插电式混合动力汽车  模糊推理  NAR神经网络  车速预测  混合逻辑动态模型

英文关键词  plug-in hybrid electric vehicle  fuzzy inference  nonlinear auto-regressive neural network  vehicle speed prediction  mixed logical dynamical model

中文摘要

      本文针对插电式混合动力汽车(plug-in hybrid electric vehicle, PHEV)这一典型混杂系统, 提出了一种基于 车速预测的混合逻辑动态(mixed logical dynamical, MLD)模型预测控制策略. 首先, 通过对发动机和电动机能量消 耗模型进行线性化, 建立双轴并联插电式混合动力城市公交车的动力传动系统数学模型; 其次, 运用模糊推理进行 驾驶意图分析, 提出基于驾驶意图识别和历史车速数据相结合的非线性自回归(nonlinear auto-regressive models, NAR)神经网络车速预测方法进行未来行驶工况预测. 然后, 以最小等效燃油消耗为目标建立PHEV的混合逻辑动 态模型, 运用预测控制思想对车速预测时域内最优电机转矩控制序列进行求解. 最后, 通过仿真实验验证了本文所 提出控制策略在特定的循环工况下与电动助力策略相比, 能够提高燃油经济性.

英文摘要

      Focusing on plug-in hybrid electric vehicle (PHEV), a classical hybrid system, a model predictive control (MPC) strategy based on mixed logical dynamical (MLD) model and vehicle speed prediction is proposed. Firstly, the dynamic model of parallel plug-in hybrid electric city bus is established using the linearized energy consumption models of engine and motor. Then, the driving intention is recognized through fuzzy inference, and a vehicle speed prediction method using nonlinear auto-regressive (NAR) models neural network is proposed based on the driving intention and the past vehicle speed data. Next, the MLD model is established with the objective function of minimum equivalent fuel consumption, and the optimal motor torque sequence within the vehicle speed prediction horizon can be solved with the predictive control theory. Finally, the simulation experiment is implemented, and the result shows that the proposed energy control strategy can improve the PHEV fuel economy compared with electric assist strategy under certain driving conditions.