Tracking Control of Unmanned Tracked Vehicle in Off-road Conditions with Large Curvature

In order to achieve the accurate trajectory tracking for a high-speed tracked vehicle in off-road conditions in the case where it is known that the desired trajectory is a large curvature curve, new research directions are being developed. In the framework of the Model Predictive Control (MPC) algorithm, the simplified ideal kinematics model of the tracked vehicle that ignores the sliding steering characteristics is applied to reduce the iterative solution time under high-speed driving conditions. By adjusting the weight coefficients of the objective function, the trajectory tracking accuracy is improved. This research is based on an unmanned electric drive tracked vehicle and carries out simulation experiments. Through real vehicle experiment, the control sequence of the experienced driver under the certain scene is collected, and then the vehicle control experience of human is obtained. Through simulation experiment, the vehicle tracking errors under different MPC weight coefficients of different curvature curves are obtained. In addition, the tracking control sequence is compared with the driver's control data. By the data analysis, the sensitivity of each weight coefficient to the tracking accuracy and how to create a driving mode that is closer to human by adjusting the weighting coefficients are determined.