A Tracking Control Approach With Sequence-Scaling Lyapunov-Based MPC for Quadruped Robots

Abstract

This article studies the tracking control problem of an autonomous quadruped robot (AQR). A new kinematic model is presented to describe both the translation mode and the rotation mode of the AQR using the same number of inputs as compared with the linear inverted pendulum model. We newly propose a sequence-scaling Lyapunov-based model predictive control algorithm for the AQR to improve the tracking performance. Within the control framework, the sequence-scaling strategy is introduced to optimize the tracking, and simultaneously guarantee the closed-loop stability. The practical constraints, such as dynamic balance of motion and speed limits, are explicitly considered to provide a feasible tracking trajectory to the AQR. We analytically address the closed-loop stability, where a contraction constraint is built within the gait sequence. Simulation and hardware experiments on Unitree Aliengo witness the superior real-time control performance via the proposed algorithm.