適応ニューラルネットワークによる故障診断法(機械力学,計測,自動制御)

抄録

Improving signal to noise ratio is a key problem to detect early faults of machinery under environment noise conditions. An effective method is presented for improving the signal to noise ratio by the adaptive neural network. This paper has made a comparison of failure detect-ability between least-mean-square (LMS) algorithm and adaptive neural network under heavy environment noise conditions. Experiment results have shown that using adaptive neural network is an effective means to extract early symptoms of machine fault under heavy environment noises and low rotating speed conditions.