Optimization based fault detection for nonlinear systems

As systems become more complex and interconnected, it becomes more difficult to monitor and maintain them. Components wear or fail, or operating conditions change, causing a degradation of performance. To meet this challenge, a theory of fault detection and isolation has been developed to enable automatic detection of faulty conditions. The focus of the paper is on a general, practical method of nonlinear fault detection which is easily implemented on input/output models that are generated by modern nonlinear system identification methods. The method proposed is different in that a neural network is used to model the process dynamics, while a dead-beat observer is implemented by solving a set of coupled nonlinear equations. This allows us to introduce constraints into the problem that can improve the power of the fault detection test