四轮驱动电动汽车稳定性预测控制器快速实现

Fast realization of stability prediction controller for four-wheel drive electric vehicles

中文关键词  车辆横摆稳定  模型预测控制  并行牛顿算法  一体化控制

英文关键词  vehicle yaw stabilization  model predictive control  parallel Newton algorithm  integrated control

中文摘要

      为解决四轮驱动电动汽车在高速情况下易发生甩尾失控的安全性问题, 针对整车和执行器间的动力学耦合、控 制系统非线性、多变量、实时性等问题, 本文采用集中式的控制策略, 设计了一种车辆横摆稳定的快速非线性预测控制 器, 实现了整车横摆稳定和电机转矩分配的一体化控制. 为了控制系统的实时实现, 将非线性规划问题转化为代数方程 组求解, 通过解耦预测时域间方程组的耦合关系, 实现时域间优化问题的并行求解, 提高了控制器的计算速度. 最后给 出了控制器的硬件并行加速实验, 完成了控制系统的硬件在环实验, 实现了车辆横摆稳定系统的实时控制. 实验结果表 明该控制器不仅具有良好的控制性能, 而且明显提升了系统实时性.

英文摘要

      The safety problem of the four-wheel drive electric vehicle is easy to lose control of tail swing at high speed. In view of the vehicle and the dynamic coupling between the actuator and control system of nonlinear, multi-variable and real-time problem, a fast nonlinear prediction controller for vehicle yaw stability is designed by adopting a centralized control strategy. This strategy realizes the integration of vehicle yaw stability and motor torque distribution control. In order to realize the real-time control system, the nonlinear programming problem is transformed into algebraic equations to solve the problem. By decoupling the coupling relation of the predict horizon equations, the parallel solution of the time-domain optimization problem is realized, and the computational speed of the controller is improved. Finally, the hardware parallel acceleration experiment of the controller is given, the hardware in loop experiment of the control system is completed, and the real-time control of the vehicle yaw stability system is realized. The experimental results show that the controller not only has good control performance, but also significantly improves the system real-time performance.