A multi-objective criterion and stability analysis for neural adaptive control of nonlinear MIMO systems: an experimental validation

This paper presents a multi-objective indirect neural adaptive control design for nonlinear square multi-variable systems with unknown dynamics. The control scheme is made of an adaptive instantaneous neural emulator, a neural controller based on fully connected real-time recurrent learning (RTRL) networks and an online parameter updating law. A multi-objective criterion that takes into account the minimisation of the control energy is considered. The contribution of this paper is to develop a new controller parameter optimisation based on the Lyapunov stability analysis while ensuring control issues with environmental and economical objectives. Performance of the proposed approach in terms of regulation, tracking and minimisation of the control energy is evaluated by numerical simulations of a disturbed nonlinear multi-variable system. The obtained control scheme is then applied in real time to a disturbed MIMO thermal process.