Adaptive Force/Motion Control System Based on Recurrent Fuzzy Wavelet CMAC Neural Networks for Condenser Cleaning Crawler-Type Mobile Manipulator Robot

In this paper, an adaptive motion/force control system is proposed for condenser cleaning crawler-type mobile manipulator robot (CCCMMR). With the merits of recurrent fuzzy wavelet cerebellar model articulation control neural networks, the unknown dynamics and parameter variations of the CCCMMR control system are relaxed by an approximation process. In addition, an adaptive robust compensator is also proposed to eliminate uncertainties that consist of the unknown approximation errors and disturbances. According to the adaptive position tracking control design, we develop an adaptive robust control strategy for the nonholonomic constraint force of the CCCMMR. The design of the adaptive online learning algorithms is derived using the Lyapunov stability theorem. Therefore, the proposed controllers prove that they not only can guarantee the stability and robustness, but also the tracking performances of the CCCMMR control system. The effectiveness and robustness of the proposed control system are verified by comparative simulation and experimental results.