Control of Flexible Ankle Joint Based on Singular Perturbation Theory

In order to solve the strong impact of the biped walking robot from the foot during the walking process and enable the foot to adapt to the bumpy road surface, the flexible ankle joint design scheme of the robot is proposed. The singular perturbation theory is used to reduce the model. Combined with state feedback and The fuzzy PID theory design fast and slow dual controllers. While satisfying the stability requirements of the system, minimizing the internal oscillation of the fast-changing elastic system and eliminating the stable residual tension of the system due to the friction. The experimental results show that the fuzzy PID make the system reach the expected position and stabilize the convergence at a faster speed. The fuzzy PID better adapts to the nonlinearity of the system and compensates for the modeling error. Because of the existence of state feedback fast-changing controller, the singular perturbation system can effectively suppress the oscillation of the system during the adjustment process and after the system is stabilized, and eliminate the residual tension of the spring caused by the friction of the ankle joint after the system is stabilized. The slow-changing fuzzy PID controller with the fast-change state feedback controller can effectively improve the overall response performance of the system. If there is no fast-changing controller, system oscillations are more likely to occur, and the system has longer settling time and greater spring residual force.