Fault diagnosis of air compressor based on RBF neural network

Because the air compressor has too many fault types, so it is often difficult to make the fault diagnosis of air compressor. For example, the detected variables are too many then it is difficult to take fault classification. The method of making use of RBF neural networks to achieve the fault diagnosis of air compressor is proposed in the paper. For the sample data is used to train RBF neural networks, thus the process of handling a large number of data detected is predigested. The RBF neural network is used to automatically identify the running states of air compressors. Simulation results show that the method that using RBF neural networks to achieve the fault diagnosis of air compressor is a feasible and very effective, and can achieve a higher diagnostic efficiency.