基于随机模型预测控制的四驱混合动力汽车能量管理

Energy Management of a 4WD HEV Based on SMPC

基于随机模型预测控制基本原理,研究了四驱混合动力汽车的能量优化管理。采用马尔可夫模型预测加速度变化过程,通过计算得到混合动力汽车未来需求转矩。在保证电池荷电状态平衡的前提下,以燃油经济性最优为目标,建立混合动力汽车能量管理优化模型。针对建立的非线性优化模型,采用动态规划算法进行有限时域内的滚动求解。将提出的控制策略在d SPACE中进行软件在环仿真试验。研究结果表明,随机模型预测控制策略可以实现四驱混合动力汽车基本的能量管理,可在保证各动力部件良好工作状况的前提下,提升燃油经济性。与基于恒值模型的预测控制策略相比,随机模型预测控制策略下的平均燃油经济性提升了8.30%,优化结果接近有先验知识的预测控制策略。

The energy management optimization of 4 WD HEV were studied based on the basic principlesof SMPC. A Markov model was built to describe the changing processes of the acceleration,so as to predictrequired torques. The optimization problem was established to minimize fuel consumption while maintainingthe balance of battery state of charge（SOC）. This nonlinear optimization problem with finite time horizon wassolved with DP algorithm. The proposed control strategy was validated with a software-in-the-loop experimentusing d SPACE. The results show that the SMPC may realize the basic energy management of the 4 WD HEVand the fuel economy is improved while all power components are working well. Average fuel economy ofSMPC is improved by 8.30%comparing with the frozen-time MPC（FTMPC） approach,and is close to theresults of the prescient MPC（PMPC）approach.