Research on generating detector algorithm in fault detection

The essence of fault diagnosis is pattern recognition to characters of fault. The non-dimensional parameter immune detector as the recognition fault detector is constructed with some non-dimensional parameter being more hypersensitive to fault combined with negative selection mechanism of artificial immune system. And two parts of immune vaccine and immune learning produce the excellence detector to diagnose fault directly. The part of immune response show fault identification and the result of the k-nearest neighbor classify method to diagnose fault. The algorithm of the design frame graph and the concrete implementing approach is given in detail. In the end, the simulation examples show that the detector by the algorithm is valid to fault detection.