A multilayer recurrent neural network for on-line synthesis of minimum-norm linear feedback control systems via pole assignment

A multilayer recurrent neural network is proposed for on-line synthesis of minimum-norm linear feedback control systems through pole assignment. The proposed neural network approach uses a four-layer recurrent neural network for the on-line computation of feedback gain matrices with the minimum Frobenius norm and desired closed-loop poles. The proposed recurrent neural network is shown to be capable of synthesizing minimum-norm linear feedback control systems in real time. The operating characteristics of the recurrent neural network and feedback control systems are demonstrated by use of an illustrative example.