In order to study on fault diagnosis of photoelectric detection system based on the photomultiplier tube（PMT）and current amplifier

a fault diagnosis method based on waveform characteristics of output signals of current amplifier is put forward.By analyzing on characteristics of output voltage waveform variation of current amplifier circuit under different fault conditions

a corresponding fault model are established.Using wavelet packet decomposition and a support vector machine（SVM）model based on particle swarm optimization（PSO）to extract and classify the feature vectors of the output voltage waveforms of the current amplifier

so as to judge fault type.The simulation results show that the PSO-SVM model can quickly and effectively diagnose and classify fault types under single and double fault conditions of the current amplifier circuit

and has a very high accuracy rate.