Iterative Learning Fault Estimation for Stochastic Time-varying Systems

The article investigates the fault estimation problem for a class of stochastic time-varying systems based on iterative learning methods. Unlike most traditional works, the system investigated in this paper is affected by disturbances, noises and initial state deviations. A type of iterative learning fault estimator is designed utilizing the historical output estimation errors for receiving the fault information of the system. Furthermore, the existence condition is analyzed in order to meet boundedness with respect to the A-norm of fault estimation errors based on the proposed iterative learning scheme. At last, a case study is applied for verify the effectiveness of the studied scheme.