Study On Hydraulic Robot Actuator Control Based On Fuzzy Algorithm

In recent years, the electro-hydraulic servo system has been widely used. In the bionic hydraulic quadruped robot, the electro-hydraulic servo control system has loaded the important task of driving power system. In the hydraulic quadruped robot, the actuator is as the drive component, the motion of each joint is realized through the control of the hydraulic cylinder piston rod extended and retracted. Therefore, it is the foundation and key of the robot motion to realize the accurate control of the actuator position system.The electro-hydraulic position servo system model of the actuator of the hydraulic quadruped robot is studied and modeled. And the model structure or order on position servo system is determined based on servo valve current input signal and hydraulic cylinder displacement output signal. The identification experiments are finished on hardware-in-loop simulation experiment bench based on MATLAB model identification toolbox, in which model parameters are obtained and identification accuracy is verified by comparing the experimental data with the identification model simulation curve, which provide the conditions for the simulation of control strategy.And the intelligent weight function algorithm is used to controller to vary error and error rate weight function, in which the weight function and fuzzy control rules can be in real time on-line automatic adjusted automatically according to the input variables. In order to reduce system steady-state error, the humanoid intelligent integral control algorithm is used to the control strategy, in which the integral control is action when the deviation increased, and the integral control is stopped when the deviation decreased. Fuzzy control rules are designed according to the principle of the servo control. In MATLAB/Simulink, the hybrid control strategy is designed and simulated, and the controller parameters are acquired in simulation based on the the position servo system identification model, which can be as the foundation for the experimental verification of the controller.The designed controller is verified on the hardware-in-loop simulation experiment bench, and the controller parameters are adjusted. Combined with the actual motion of the robot, the experiments are finished under different conditions, with the variable inertia load, the variable elastic load and the variable input displacement signals. The experiment results show that the designed controller can adapt to these changes and can suppress the overshoot and improve the steady state accuracy better than PID controller and fuzzy controller. The accuracy of the hybrid controller on the position control of the actuator is verified. The study can be a reliable basis for the design of fuzzy control rules and correction factors. The successful debugging of humanoid intelligent fuzzy controller can be used as a reference for other related research.