Nonlinear system control via feedback linearization using neural networks

This paper addresses the problem of feedback linearization of nonlinear systems. The existing linearization methods require complete knowledge of the system model. A new method for feedback linearization, avoiding this requirement which is rarely satisfied in practice, is proposed. The method is based on artificial neural networks (ANNs) that emulate the plant's Lie derivatives. Simulation results show satisfactory performance when the proposed ANN-based feedback linearization is included in a tracking control system.