基于空间MPC算法的矿用车辆轨迹跟踪研究

Spatial Model Predictive Control for Trajectory Tracking of Autonomous Vehicle in Coal Mine

针对矿井无人运输车辆轨迹跟踪控制问题,提出了一种改进的空间模型预测控制(SMPC)算法。首先,将传统的基于时间的车辆运动学模型转换为基于空间的偏差模型,降低预测方程维度,减小控制器的计算负担;其次,让目标函数最小化车辆转向曲率的1、2阶导数和横向位移偏差,通过优化车辆转向曲率的1、2阶导数减少转向角突然变化对转向部件和车辆结构造成的损伤,保证轨迹跟踪精确性的同时确保车辆行驶的平稳性;最后,将目标函数转化为二次规划问题进行求解最优曲率。模拟井下巷道条件在双移线工况和实际工况下对车辆进行仿真实验,结果表明:在2种工况下,SMPC控制算法均能保证轨迹跟踪的精确性和平稳性,且控制效果优于标准MPC。此外,SMPC算法在不同负载下也具有良好的鲁棒性。

An improved spatial model predictive control(SMPC)algorithm is proposed for the trajectory tracking control of mining auto-driving heavy vehicles.The innovation of this method is that firstly,the traditional time-based vehicle kinematics model is transformed into a space-based deviation model,which reduces the dimension of the prediction equation and reduces the computational burden of the controller.Secondly,the objective function is to minimize the vehicle curvature of First-order andsecond-order derivative and lateral displacement deviation,the first-order and second-order derivatives of the vehicle curvature can effectively reduce the damage caused by the sudden change of steering angle to the steering component and the vehicle structure,which ensure the accuracy of the trajectory tracking while ensuring the vehicle travels;Finally,the objective function is transformed into a quadratic programming problem to solve the optimal curvature.In this paper,the vehicle simulation experiments were carried out in the simulated roadway conditions under both double shifting and actual working conditions.The results showed that the SMPC control algorithm could guarantee the accuracy and smoothness of track tracking under both working conditions,and the control effect was better than the standard MPC.In addition,SMPC algorithm has good robustness under different loads.