Estimation of actuator faults for system with unpredictable inputs

This paper deals with a fault estimation algorithm to reconstruct faulty process inputs under actuator fault. The presented estimation approach is based on the parameter identification and a Recursive Least Square (RLS) algorithm for the identification of the model parameters. The proposed approach was essentially built for the estimation of the handwriting process inputs under actuator fault, which are the activities of the forearm muscles, responsible for the handwriting act, called Integrated Electromyography (IEMG) signals. The proposed estimation model is then using data of the handwriting process in faulty case, to estimate the actuator faults. In this paper we also present different kind of faults that could attempt handwriting process actuator during the writing act. The simulation results presented here illustrate the advantages of the proposed approach on estimating faulty inputs of the handwriting process.