LQR Compensated by Fuzzy for Kicking Balance Control of A Humanoid Robot

Robots have various shapes optimized to fulfill specific functions and tasks. One type of robot based on its form is a humanoid robot with a human-like figure. With a similar body shape, humanoid robots make it possible to carry out activities with humans. One of them is a humanoid robot playing soccer. For this reason, humanoid robots must be able to perform basic movements in soccer, such as kicking a ball. When kicking, the robot tends to fall during swinging leg movements due to extreme speed changes. The fixed gain value in control given to the robot is not optimal for overcoming this problem because of the setpoint changes that occur. For the needs of motion stability simulation, we use the inverted pendulum model as the basis for the mathematical model of the humanoid. This mathematical model is combined with Linear Quadratic Regulator (LQR), which is compensated by the Fuzzy method, and then simulated. This simulation aims to see the response of the control system on the robot. Providing LQR-based and Fuzzy controls has been shown to play a role in reducing excessive torque when kicking. This condition makes the humanoid robot succeed in kicking stably without falling