RBFNN based direct adaptive control of MIMO nonlinear system and its application to a distillation column

In this paper, a RBFNN (radial basis function neural network) based direct adaptive controller for a MIMO nonlinear system is designed. The tuning law of weights of the neural network is derived from a selected Lyapunov function. So the stability of the closed loop and convergence of weights are guaranteed. The design method is applied to the quality control of a distillation column. The dual-point control strategy is adopted instead of single-point control. A simulation is illustrated to show the validity of the designed controller.