STABILITY CONTROL OF HUMANOID ROBOT WALKING SPEED USING LINEAR QUADRATIC REGULATOR METHOD

The humanoid robot exhibits overshooting upon reaching the desired walking speed, introducing oscillations that pose a risk of the robot falling. Sustaining a signif-icant steady-state error during speed maintenance could compromise the robot’s stable posture and lead to a potential fall. Therefore, the implementation of a control system becomes essential to regulate the robot’s walking speed, aiming to eliminate overshooting and minimize steady-state error. In this study, the Linear Quadratic Regulator (LQR) method is employed for controlling the robot’s walking speed. The output from the LQR control is a torque value utilized for managing the servo motor positioned at the heel of the robot’s feet. By employing full-state feedback control, the humanoid robot, set to a speed target of 0.025 m/s, achieves the desired speed without generating overshooting and maintains a speed with a steady-state error below 2%.