Research on the controller of the ground testing system for space robot

A novel PID controlling algorithm based on Radial Basic Function Neural Networks (RBFNN) is proposed for the ground testing system for space robot. The testing system is with strong non-linearity and uncertainty. The parameters of PID controller are adjusted by RBFNN online. The torque is controlled by the output of circuit loop directly, which attains constant tense to simulate microgravity environment. The experimental results show that the controlling algorithm is effective, and the system is with good dynamic performance, sound robustness and good self-adaptive performance.