NN-ANARX model based control of nonlinear discrete-time systems with input delay

This paper provides a concise description of a dynamic output feedback linearization algorithm for control of nonlinear systems by means of Additive Nonlinear Autoregressive eXogenous (ANARX) class of models. ANARX structure of the model can be obtained by training a neural network with a specific restricted connectivity structure. Additionally, generalized control algorithm based on reference model is discussed. Linear discrete-time reference model is given in the form of a transfer function defining desired zeros and poles of the closed loop system. Neural network's based ANARX model can be linearized by the proposed linearization algorithm in a such way that the transfer function of the linear closed loop system corresponds to the given reference model. Finally, the adjustment of described algorithms for the case of systems with delays of input signals is introduced. The effectiveness of the proposed techniques is demonstrated on numerical example.