基于BP神经网络和证据理论的超声检测缺陷识别

王力; 周志杰; 赵福均

doi:doi:10.3969/j.issn.1671-637x.2018.01.014

电光与控制, 2018, 25 (1): 65, 网络出版: 2018-01-30

基于BP神经网络和证据理论的超声检测缺陷识别

Flaw Identification in Ultrasonic Testing Based on BP Neural Network and Evidence Theory

论文大纲

王力周志杰赵福均

作者单位

火箭军工程大学，西安 710025

数据融合证据理论神经网络超声检测证据可靠性 data fusion evidence theory neural network ultrasonic identification evidence reliability

摘要

针对超声检测缺陷性质识别中准确率较低的问题，研究了一种基于BP神经网络和证据理论进行超声检测缺陷识别的方法。首先，提出了一种基于BP神经网络和证据理论的融合模型，利用BP神经网络进行特征层融合，将其输出作为证据源的概率分布函数。其次，在决策层融合中针对传统D-S证据理论易出现证据冲突的情况，考虑到不同传感器获取数据的可靠性差异，给出获取证据源可靠性因子的方法。通过引入可靠性因子λ衡量不同证据源的可靠性，使得所有证据源经过可靠性评估后再进行数据融合。最后，通过超声检测手段获取某航空材料的缺陷数据，并对提出的方法进行了验证。研究结果显示，该方法能够更加准确地进行缺陷识别，与传统D-S证据理论相比提高了缺陷识别的准确性。

Abstract

In order to solve the problem of low accuracy in ultrasonic characterization of defect detection,a new method based on BP neural network and D-S evidence theory is proposed to identify defects in ultrasonic inspection.Firstly,a fusion model based on BP neural network and evidence theory is proposed.The BP neural network is used to fuse the feature layer,and its output is taken as the basic probability distribution function of the evidence source.Secondly,in the fusion of decision-making level,the validity of the evidence source is evaluated by introducing an effective factor λ to evaluate the reliability of the evidence source.Therefore,all the evidence sources are fused after the reliability evaluation.Finally,the defect data of a certain type of aeronautical material are obtained by means of ultrasonic testing,and the method proposed in this paper is verified.The results show that:compared with the traditional D-S evidence theory,this method can identify the flaws more accurately and improve the accuracy of flaw recognition.

参考文献

[1] 张道富，曾燕，周晓军，等.基于信息融合技术的无损检测缺陷模式识别［J］．仪器仪表学报,2003,24(5): 449-452.

[2] OUKHELLOU L，DEBIOLLES A，DENOEUX T,et al.Fault diagnosis in railway track circuits using Demster-Shafer classifier fusion［J］.Engineering Applications of Artificial Intelligence,2010,23(1):117-128.

[3] SI X S,HU C H,YANG J B,et al.On the dynamic evidential reasoning algorithm for fault prediction［J］.Expert System with Applications,2011,38:5061-5081.

[4] XU X B,LIU P,SUN Y B.Fault diagnosis based on the updating strategy of interval-valued belief structures［J］.Chinese Journal of Electronics,2014,23(4):753-760.

[5] 潘泉，王增福，梁彦，等.信息融合理论的基本方法与进展(II)［J］.控制理论与应用,2012，29(10): 1233-1244.

[6] 韩崇昭，朱洪艳，段战胜，等.多源信息融合［M］.2版.北京:清华大学出版社，2010: 19-44.

[7] 徐晓滨，王玉成，文成林.评估诊断证据可靠性的信息融合故障诊断方法［J］.控制理论与应用,2011,28(4): 504-510.

[8] 何友，王国宏.多传感器信息融合及应用［M］.2版.北京:电子工业出版社，2010: 59-63.

[9] 周志杰,杨剑波，胡昌华，等.置信规则库专家系统与复杂系统建模［M］.北京:科学出版社，2011: 9-30.

[10] 徐晓滨，郑进，徐冬玲，等.基于证据推理规则的信息融合故障诊断方法［J］.控制理论与应用,2015，32(9): 1170-1182.

[11] YANG J B,XU D L.Evidence reasoning rule for evidence combination［J］.Artificial Intelligence,2013,205: 1-29.

[12] 胡文刚，刚铁.基于多探头源数据融合的焊缝缺陷识别［J］.焊接学报,2013，34(3): 45-48.

王力, 周志杰, 赵福均. 基于BP神经网络和证据理论的超声检测缺陷识别[J]. 电光与控制, 2018, 25(1): 65. WANG Li, ZHOU Zhi-jie, ZHAO Fu-jun. Flaw Identification in Ultrasonic Testing Based on BP Neural Network and Evidence Theory[J]. Electronics Optics & Control, 2018, 25(1): 65.

基于BP神经网络和证据理论的超声检测缺陷识别

关于本站 Cookie 的使用提示

全站搜索

基于BP神经网络和证据理论的超声检测缺陷识别

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索