基于随机模型预测控制的并联式混合动力汽车控制策略研究  

A Research on the Control Strategy of a Parallel HEV Based on Stochastic Model Predictive Control

作　　者：赵韩[1] 吴迪[1] 

机构地区：[1]合肥工业大学机械与汽车工程学院,合肥230009

出　　处：《汽车工程》2014年第11期1289-1294,共6页Automotive Engineering

基　　金：国家"863"计划节能与新能源汽车重大专项(2011AA11A202;2008AA11A139)资助

摘　　要：基于随机模型预测控制算法,对并联式混合动力汽车的转矩分配问题进行了研究。建立马尔科夫模型对需求功率进行预测,并将随机模型预测控制与动态规划相结合,提出了基于模型预测控制,并以消耗最小化为目标进行滚动优化的控制策略。在MATLAB/Simulink平台上搭建了仿真模型,并进行和逻辑门限控制策略的对比仿真。结果表明,与逻辑门限值控制策略相比,采用所提出的的控制策略时车辆能量经济性得到明显提高,说明用马尔科夫模型预测功率需求的控制策略是可行的,且具有良好的实时性。The torque distribution of a parallel hybrid electric vehicle is studied based on stochastic model predictive control （ SMPC） algorithm. A Markov model is built to predict required power, and by combining SMPC with dynamic programming, a control strategy is proposed based on SMPC and rolling optimization with minimizing energy consumption as objective. A simulation model is set up with Matlab/simulink platform and a comparative simulation between proposed strategy and logic threshold control strategy is conducted. The results show that com-pared with logic threshold control strategy, when adopting the strategy proposed the energy economy of vehicle obvi-ously improves, meaning the strategy of using Markov model to predict desired power is feasible with good real time performance.

关 键 词：并联式混合动力汽车 模型预测控制 动态规划 马尔科夫模型 

分 类 号：U469.7[机械工程—车辆工程]