Application of Least Squares Parameter Estimation Techniques in Fault Detection

Failure detection has been the subject of many studies in the past. Modern technology has required highly complex dynamic systems. A critical system is any system whose ‘failure’ could threaten the system’s environment or the existence of the organization which operates the system. A fault is understood as any kind of malfunction in the actual dynamic system, the plant that leads to an unacceptable anomaly in the overall system performance. Fault detection via parameter estimation relies in the principle that possible faults in the monitored system can be associated with specific parameters and states of the mathematical model of the system given in the form of an inputoutput relation. In this thesis, the focus is put on the study of fast least squares parameter estimation methods, like recursive least square (RLS) algorithm, Fast Kalman algorithm, FAEST(fast a priori error sequential technique) algorithm, FTF(fast transversal filter) algorithm and lattice filter algorithm and their fast algorithm implementation. The above algorithms are applied to a dynamic system and the performances of different algorithms in detecting different changes in the systems are compared. The MATLAB coding of these algorithms are done and their effect on first and second order dynamic systems under various conditions are verified. Statistical methods like Shewart moving range control chart, the cumulative sum control chart, the moving average control chart, the exponentially weighted moving average control chart (EWMA) etc. for analyzing the changes in dynamic systems are also studied.