A Novel Fault Diagnostics and Prediction Scheme Using a Nonlinear Observer With Artificial Immune System as an Online Approximator

In this paper, an observer-based fault diagnostics and prediction (FDP) scheme for a class of nonlinear discrete-time systems via output measurements is introduced by using artificial immune system (AIS) and a robust adaptive term. Traditionally, AIS was considered as an offline tool for system identification and pattern recognition whereas here AIS is utilized as an online approximator in discrete-time (OLAD) in a fault detection (FD) observer. A fault is detected when the output residual exceeds a predefined threshold. Upon detection, the OLAD is initiated to learn the unknown fault dynamics online while the robust adaptive term ensure asymptotic convergence of the output residual for a state fault whereas a bounded result for an output fault. Additionally, a mathematical equation is introduced to estimate the time-to-failure (TTF) by using the output residual and the estimated fault parameters. Finally, the performance of the proposed FDP scheme is demonstrated on an axial piston pump hardware test-bed.