Iterative learning based fault estimation for nonlinear discrete-time systems

The fault estimation problem for a class of nonlinear discrete-time systems with Lipschitz condition is studied. By introducing a P-type iterative learning strategy and considering the effect of initial value deviations, we propose a fault estimation algorithm based on the iterative learning filtering. Using the lower triangular matrix theory and the singular value characteristics of matrixes, we obtain conditions for the virtual fault introduced to approach the actual fault. Simulation results show the effectiveness of our proposed algorithm.