基于鲁棒控制的推土机轨迹跟踪设计

Trajectory Tracking Design of Bulldozer Based on Robust Control

本文针对推土机在未知有界扰动下的轨迹跟踪问题,提出了一种有效的鲁棒模型预测控制(RMPC)方法。提出的RMPC是通过考虑线性化的跟踪误差模型,结合前馈和最优控制动作来实现提出的轨迹。校正反馈控制器作为时变有限时间线性二次型调节器(LQR)的存在,通过调节标称系统周围的不确定性来抑制作用于实际系统的不确定性。控制反馈所需的姿态估计是基于传感器数据融合,由基于扩展卡尔曼滤波器(EkF)地图的定位器进行,该定位器处理惯性测量单元(IMU)和光探测和测距(lidAR)测量。利用一个真实的推土机进行了实验,验证了所提出的控制方案的性能。实验结果表明,该控制器能够安全地跟踪目标运动轨迹,具有处理时间短、跟踪误差小、控制动作平稳等优点。最后,将所提出的控制方案与相关技术进行了比较,在跟踪精度上优于相关技术。

In this paper,an effective Shandong rod model predictive control(RMPC)-RRB-method is proposed for trajectory tracking of a bulldozer under unknown bounded disturbances.The proposed RMPC realizes the proposed trajectory by considering the linearized tracking error model,combining feedforward and optimal control action.As a time-varying finite-time linear quadratic regulator(LQR),the correction feedback controller is used to control the uncertainties around the nominal system.The attitude estimation required for control feedback is based on sensor data fusion,performed by Extended Kalman filter(EKF)map-based locators,the locator handles inertial measurement unit(IMU)and light detection and ranging(LIDAR)-RRB-measurements.The performance of the proposed control scheme is verified by an experiment with a real bulldozer.The experimental results show that the controller can track the target trajectory safely,and has the advantages of short processing time,small tracking error and stable control action.Finally,the proposed control scheme is compared with the related techniques,and the tracking accuracy is better than the related techniques.