摘要
为提高分布式驱动电动汽车在极限工况下的稳定性,提出一种基于非线性模型预测控制(nonlinear model predictive control,NMPC)算法转矩协调控制策略。该策略基于NMPC的集中式系统整体优化方法,兼顾了整车稳定性控制和转矩优化分配,可以改善车辆行驶稳定性,确保车轮滑移率维持在稳定的范围内,提高车辆驱/制动的稳定性。首先利用MATLAB建立整车7自由度动力学模型,同时基于预测控制的算法框架搭建车辆动力学控制器模型。然后根据操稳性控制需求,同时考虑了轮胎的纵向滑移约束设计系统的目标函数;结合障碍函数方法,基于估计的期望动态滑移率实现对滑移率的边界约束。最后将期望值的跟踪控制和滑移约束控制转化为带约束的优化问题,求解得到各轮最优转矩分配。仿真结果表明:该转矩协调控制系统可以有效地改善车辆的稳定性,同时将轮胎滑移率控制在稳定范围内,提高车辆安全性。
Here,in order to improve stability of distributed drive electric vehicles under limit conditions,a torque coordinated control strategy based on nonlinear model predictive control(NMPC)algorithm was proposed.The strategy was based on NMPC’s overall optimization method of centralized system,it could consider the whole vehicle stability control and torque optimal distribution,improve the vehicle driving stability,ensure tire slip rate within a stable range,and improve the stability of vehicle drive/braking.Firstly,a whole vehicle dynamic model with 7-DOF was established using the software MATLAB,and the vehicle dynamic controller model was built based on the algorithm framework of predictive control.Then,according to handling stability control requirements,considering tire’s longitudinal slip constraint,the system’s objective function was designed.Combined with the obstacle function method,the boundary constraint of slip rate was realized based on the expected dynamic slip rate estimated.Finally,the expected value tracking control and sliding constraint control were converted into the constrained optimization problem to solve and obtain the optimal torque distribution on each wheel.The simulation results showed that the proposed torque coordinated control system can effectively improve vehicle stability,control tire slip rate within a stable range,and improve vehicle safety.
作者
庄元强
吴洋
张邦基
胡文
ZHUANG Yuanqiang;WU Yang;ZHANG Bangji;HU Wen(State Key Lab of Advanced Design and Manufacture for Vehicle Body,Hunan University,Changsha 410082,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2021年第13期239-246,298,共9页
Journal of Vibration and Shock
基金
国家自然科学基金资助项目(5187051675)。
