Multivariable adaptive control using an observer based on a recurrent neural network

A real-time learning control technique for a general non-linear multivariab le process is presented and applied to a laboratory plant. The proposed tec hnique is a hybrid approach, which combines the ability of a recurrent neur al network for modelling purposes and a linear pole placement control law t o design the controller, providing a bridge between the field of neural net works and the well-known linear adaptive control methods. An Elman-type recurrent neural network strategy is introduced to model the behaviour of the non-linear plant, using available input-output data tan un measurable state problem is assumed). Following a linearization technique a linear time-varying state-space model is obtained, which allows simultaneo us estimation of parameters and states. Once the neural model is linearized , some well-established standard linear control strategies can be applied. With simultaneous online training of the neural network and controller synt hesis, the resulting structure is an indirect adaptive self-tuning strategy . The identification and control performances of the proposed approach are in vestigated on a non-linear multivariable three-tank laboratory system. Expe rimental results show the effectiveness of the proposed hybrid structure. C opyright (C) 1999 John Wiley & Sons, Ltd.