Exploring Snake Robot Motion Control: A VR-Enhanced Approach for Studying the MPC Algorithm

Abstract

This study focuses on enhancing control strategies and locomotion capabilities for underactuated snake robots. A 10-segment underactuated snake robot with 9 Degrees of Freedom (DOF) is modelled and controlled using Model Predictive Control (MPC). MATLAB is utilized to implement the snake robot model and various control algorithms. The simulation results are then integrated into a Virtual Reality (VR) application to visualize the motion of the snake robot. The VR scene generates an animated representation of the snake robot's segments, along with selected physical parameters such as segment velocities, heading angle, and "ghost robots" - predicted positions of the snake robot, along the prediction horizon, calculated by the MPC algorithm. The results demonstrate improved performance and agility of the snake robot by implementing advanced control strategies. The examined strategies enable the robot to navigate in challenging environments even with joint malfunctions.