원문정보
초록
영어
The acoustic emission (AE) technology can be used to assess the security condition of oil storage tank without opening pot. Signal recognition is a foundation to analyze the corrosion status for oil storage tanks. Because of inadequateness of the analysis method of parameters, a new acoustic emission signal recognition method is proposed based on wavelet transform and RBF neural network. AE signal was decomposed to 6 layers by db2 wavelet and the space energy of 6-layer detail features is regarded as the vector of the AE signal characteristics. RBF neural network is designed by considering the characteristics of AE signal. The RBF neural network is trained by using the pattern known of acoustic emission signal. RBF network is used to classify experiments to corrosion, crack and condensation acoustic emission signal. The experimental results show that the recognition rate of RBF neural network reaches 93.3%, which reveals the advantage of the acoustic emission signal of neural network recognition. It has some significance of the quantitative analysis to the safety situation of oil storage tanks.
목차
1. Introduction
2 The AE Signal Feature Extraction based on Wavelet Transform
2.1. The Theory of Wavelet Transform
2.2. The Feature Extraction of AE signal
3. The Structure of RBF Neural Network
4. The Results and Analysis of the Experiments
5. Conclusion
References