基于紧约束鲁棒模型预测控制的无人车辆轨迹跟踪控制

Trajectory tracking control of unmanned vehicles based on tight constrained robust model predictive control

针对无人驾驶车辆的轨迹跟踪问题,基于车辆非线性动力学模型,设计了一种紧约束鲁棒模型预测控制(MPC)方法。首先,利用魔术公式轮胎模型,结合车辆3自由度系统,构建了具有有界干扰的车辆线性离散误差模型;其次,采用紧约束控制策略设计系统的鲁棒优化问题;最后,通过在线优化得到最优控制序列,并离线构造之后多个时刻满足约束条件的可行控制序列。仿真结果表明,所提算法能够使车辆稳定快速跟踪上参考轨迹并有效提高系统计算资源利用率。

A tight constrained robust model predictive control(MPC)method is proposed for unmanned vehicles to track trajectory based on the nonlinear dynamics model of the vehicle.Firstly,a linear discrete error model of the vehicle with bounded disturbances is constructed by using the magic formula tire model and the vehicle 3-degree of freedom system.Secondly,a tight constraint control strategy is used to design the robust optimization problem of the system.Lastly,the optimal control sequence is obtained through on-line optimization,and the feasible control sequence that satisfies the constraint conditions at multiple times is constructed off-line.Simulation results show that the proposed algorithm can make the vehicle track the reference trajectory stably and quickly,and effectively improve the utiliza-tion of system computing resources.