A Nonlinear Fault Identification scheme for Reusable Launch Vehicles Control Surfaces

This paper address the problem of fault identification for Reusable Launch Vehicles (RLV) control surfaces. The identification scheme is based on a modified extended Kalman filter which is easy to implement. A solution is provided for systematic tuning the filter noise covariance matrices. It is shown that this problem can be formulated as an optimization problem using a quadratic criterion which can be solved using a Particle Swarm Optimization (PSO) algorithm. A prior trimmability deficiency analysis procedure is also proposed using a state-space modeling approach. The simulation results are quite encouraging and suggest that the proposed fault identification scheme could be an efficient tool for advanced diagnosis algorithm for RLV actuators.