A Control Method for a Hung-Up Snake Robot Based on Proximal Policy Optimization and the Time-Reversed Method

Due to the presence of multi-joint synergistic movement, the control of the tail-swing locomotion of a snake robot is a very challenging problem. In this paper, a control method based on proximal policy optimization and the time-reversed method is proposed to solve this problem. Specifically, this method first simplifies the snake robot model to the two-link model; then, the trajectory planning is carried out by using the time-reversed method to obtain the reference trajectory; then, the trajectory tracking method based on proximal policy optimization is used to track the reference trajectory and control the tail of the snake robot to reach the target state; then two traditional nonlinear controllers are designed as a contrast. Finally, some experiments results are collected and analyzed, which shows the superior performance of the proposed method.