The pattern recognition of non-destructive testing based on HMM

Zhang Daofu1, Zeng Yan2, Zhou Xiaojun2, Cheng Yaodong2
1Department of Mechanism, University of Zhejiang, HangZhou, China
2Department of Information System Engineering, Zhejiang Technical University, Hangzhou, China

Tóm tắt

A method of fault pattern recognition based on signal processing and pattern recognition. A hidden signal processing and characteristic reconstructed is presented Markov model which is scatter output and one rank is in this paper. Hidden Markov model (HMM),which is initially employed in the field of speech signal processing is introduced into fault pattern recognition of nondestructive testing(NDT), which lets a fault in aviation material be reconstructed automatically with the help of computer analysis and control system. It brings practical application for aviation material quality and safety use.

Từ khóa

#Pattern recognition #Nondestructive testing #Hidden Markov models #Signal processing #Speech analysis #Scattering #Speech processing #Automatic testing #Materials testing #System testing

Tài liệu tham khảo

zhenqing, 1995, A Split Spectrum approach to strength thick crystal s ultrasonic inspection signal[J], Non-Destructive Testing, 17, 211 jiankai, 1993, Principle and method of ultrasonic inspection keji, 1997, The principle of the acoustics signal processing and practice based on neural net, 4 guangping, 1994, Ultrasonic signal processing, Non-Destructive Testing, 16, 178 jinhui, 1995, Hidden Markov models for speech processing 10.1002/j.1538-7305.1985.tb00274.x 10.1109/5.18626