Grey Wolf Optimizer Based Tuning of a Hybrid LQR-PID Controller for Foot Trajectory Control of a Quadruped Robot

Abstract

Quadruped robots have generally complex construction, so designing a stable controller for themis a major struggle task. This paper presents designing and optimization of an effective hybridcontrol by combining LQR and PID controllers. In this study, the tuning of a hybrid LQR-PIDcontroller for foot trajectory control of a quadruped robot during step motion using Grey WolfOptimizer (GWO) algorithm which is an alternative method are comparatively investigated withtwo traditional benchmarking algorithms (PSO and GA). The principal goal of this work is thetuning of the LQR controller parameters (Q and R weight matrices) and the PID controllers gains(kp, ki and kd) using the proposed algorithms. Initially, the designed solid model of the quadrupedrobot is imported into Simulink/SimMechanics which are simulation tools of MATLAB and thenobtained the mathematical model of system which is at State-Space form with Linear AnalysisTools considering the step motion of robot leg in sagittal plane. Later, the hybrid LQR-PIDcontrol system is designed and its parameters are tuned to get optimal values which guaranteebest trajectory tracing in Simulink with the three proposed algorithms. Subsequently, the systemis simulated separately with optimal control parameters which provide from the algorithms. Thesimulation outcomes are indicating that GWO algorithm is more efficiently and quickly withinsimilar torques to tuning the hybrid controller based on LQR&PID than the other conventionalalgorithms.