FD on Systems Type 1 and 2 Using Conditional Observers

Most of non-linear type 1 and type 2 control systems suffers from lack of detectability when model based techniques are applied on FDI tasks. This research work presents an strategy based on conditional observers implemented by means of massive neural networks based models applied on a parity space approach. Conditional observers are modeled using a novel neural network based approach. As consequence of such technique on nonlinear plants of types one and two, useful results were achieved.