A framework for robust neural network-based control of nonlinear servomechanisms

A framework for robust neural network-based control of nonlinear servomechanisms is proposed. It utilizes a general controller structure comprising a nonlinear compensation block and a robust control block. Two strategies for designing the control laws are discussed and uniform stability of the overall system in the presence of modelling mismatches and nonparametric uncertainty is achieved. This framework is demonstrated in implementation experiments for position control in a servomechanism with asymmetrical changing load.<
>