仿人机器人关节模糊自适应PID控制

he joint loads of humanoid robot change in the walking process, while PID parameter of traditional PID is fixed. It is difficult to adjust PID parameter online for traditional PID controller. In this paper, a fuzzy adaptive PID controller is put forward to the servo control of humanoid robot joints. First, the fuzzy relationship model between PID parameters and current error E and error rate EC is established, and then the fuzzy control rules is determined through the experiments. Finally, the fuzzy control enquiry form is generated by using mamdani fuzzy reasoning algorithm, which compensating the PID parameters online. Through the experiments on physical knee motor, we confirmed that the fuzzy adaptive PID can adjust the PID parameters with the changes in load of knee in a fast response time. Meanwhile anti-jamming capability is improved, which satisfies requirements of humanoid robot control motor.