The application of RBF network based on the principle of immune in the condenser's fault diagnosis

For the specific problems which is determined by the Center value and the width of the RBF neural network function, a new network training method is proposed. It is based on the immune theory and combines immune genetic algorithm with RBF neural network. And it is applied to realize the failure diagnosis of condenser. The result of verification shows that the proposed method can improve the accuracy and the speed of failure diagnosis effectively and has certain application value in engineering.