摘要
基于模糊运算和人工神经网络在模式识别中的应用,分别对两种模式识别方法的基本原理及其在管道无损检测中的应用进行了分析。通过建立管道缺陷识别试验,根据检测信号选取适当的参数,分别用这两种方法对不同类型的缺陷进行识别。结合识别结果对其进行比较与总结,证明了两种识别方法的有效性,为识别方法的选择与应用提供了理论依据。
Based on the application of fuzzy algorithm and neural networks in pattern recognition, the basic theory and application of them in nondestructive testing were analyzed. The experiment of pipeline defect recognition was designed, appropriate parameters were determined, while two methods were applied to effect recognition. According to the comparison and conclusion of experimental result, the validity of two methods was provided which would be a theoretical basis of selection and application with recognition methods.
出处
《微计算机信息》
2010年第2期26-27,72,共3页
Control & Automation
基金
中国人民解放军总后物资油料部科研项目(项目名称
基金编号不公开)
关键词
缺陷识别
管道
模糊算法
神经网络
Defect Recognition
Pipeline
Fuzzy Algorithm
Neural Networks