An Adaptive Fuzzy Trajectory Tracking Control via Improved Cerebellar Model Articulation Controller for Electro-Hydraulic Shovel

Most of existing model-based control strategies for trajectory tracking of electro-hydraulic shovel (EHS) are limited to the condition that full state variables can be measured. This article proposes an adaptive control system consisting of a terminal sliding mode controller and a novel neural network controller for trajectory tracking of EHS, and only the system position signal is adopted in the control law. The newly designed hybrid cerebellar model articulation controller contains a radial basis function neural network (RBFNN) preprocessor and a main CMAC controller that generated final output. The RBFNN preprocessor can decrease input range and dimensions for CMAC, which can speed up learning and reduce computing cost. The control law is derived based on the Lyapunov stability theory and finite-time convergence can be realized. Furthermore, an adaptive compensation term is introduced to compensate the combination errors. Finally, comparative simulation and experimental results have confirmed that the proposed method can accommodate the lumped uncertainties well and has better performance with high computational efficiency.