Fault Estimation for Nonlinear Repetitive Systems: An Iterative Learning Approach

In this paper, the fault estimation (FE) problem for a class of nonlinear repetitive systems with bounded initial state errors is investigated by using an accelerated iterative learning (IL) approach. In each iteration, only the boundedness condition is known, which can reduce the conservativeness of IL methods when applied to industrial production. The traditional IL algorithm is modified by adding a weight coefficient to accelerate the learning speed and weaken or even eliminate the negative influence of initial state errors on the convergence of fault estimation errors. A sufficient condition is established for guaranteeing accurate fault estimation under the proposed IL-based scheme as the iteration numbers increase. Finally, illustrative examples are presented to verify the effectiveness of the proposed modified fault estimation scheme.