NARMAX neural modelling and detecting faults using the cumulative sum statistical test

摘要

In this paper a real-time system for detecting changes in dynamic systems is designed. The Cumulative Sum (CUSUM) or Page-Hinkley test is intended to reveal any drift from the normal behaviour of the process which is established by a reliable model. In order to obtain this reliable model, the black-box identification by means of a Non-linear Auto-Regressive Moving Average with exogenous (NARMAX) neural model has been chosen. This paper shows also the choice and the performance of this neural network in the training and the test phases. A study is related to the inputs number, and of hidden neurons used and their influence on the neural model. Three statistical criterions are used for the validation of the experimental data. After describing the system architecture and the proposed methodology of the fault detection, we present a realistic application to show the technique's potential. The purpose is to detect the change presence, and pinpoint the moment it occurred.