基于智能驾驶员模型的电动车能源效率优化控制

Intelligent driver model based optimal control for energy efficiency of electric vehicle

中文关键词  电动汽车  智能驾驶员模型  驾驶风格识别  能源效率  模型预测控制

英文关键词  electric vehicle  intelligent driver model  driving style recognition  energy efficiency  model predictive control

中文摘要

      研究显示融合运行车辆未来动力需求预测信息的动力系统优化控制策略能够显著提高能源效率.另一方 面, 智能驾驶员模型(IDM)提供了一种利用车联网技术所提供的信息定量地估计车辆运行状态的算法.本文通过融 合驾驶风格识别提出了一种改进的IDM,以实现更高精度的前车运行状态预测.在此基础上,提出了基于模型预测 控制的电动汽车能源效率优化控制的数学描述.将能耗作为优化指标,而为了进一步扩大车辆能耗效率的优化空 间, 本文利用松弛处理方法,提出了改进的基于固定车头时距的跟车模型作为约束条件,从而在保证了运行车辆节 能与安全的前提下,进一步提升了道路的通行效率.最后,利用多模态交通场景模拟软件SUMO以及MATLAB仿真 软件构建了仿真实验平台.通过与采用IDM实现跟车控制的基准策略对比,仿真实验结果显示本文提出的基于模型 预测控制的策略不仅能够提高车辆的通行效率,同时,在3种不同的信号交叉路口运行场景下,能耗效率分别提升了 5.05%, 3.2%及4.15%.

英文摘要

      Related research shows that optimal control strategies that involve the future power demand prediction could improve the vehicle energy efficiency significantly. On the other hand, the intelligent driver model (IDM) provides an algo rithm that can provide a quantitative predictions of future vehicle states with the data obtained by using connected vehicle technologies. This paper proposes an improved IDM to predict the vehicle states with high accuracy, using this prediction algorithm, a model predictive control (MPC) formulation is proposed to deal with the energy management issue of electric vehicles. The energy consumption is taken as the objective function, and in order to further expand the optimization space, a car-following model with fixed headway is improved by introducing the relaxation processing mechanism. The study shows that the solution of the proposed problem guarantees the energy conservation and safety, and further improves the traffic efficiency. Finally, an evaluation platform is constructed by using multi-mode traffic scenario simulation software SUMO and MATLAB software. The validation results show that the proposed MPC strategy can not only improve the traffic efficiency, but also improve the energy efficiency by 5.05%, 3.2% and 4.15%, respectively with respect to three signalized intersections road operating scenarios.