Research And Development Of Polyethylene Pipe Electrofusion Joint Ultrasonic Detection Defect Automatic Identification System

As the continuous adjustment of energy structure in China,the proportion of clean energy is increasing.Polyethylene(PE)pipe is widely used in the construction of gas pipeline network due to its advantages of good toughness,corrosion resistance,anti fouling and long service life.It has been used in the pressurized water delivery system of nuclear power plants in recent years.As a result,the requirements for the safety performance of polyethylene pipes are becoming stricter.The pipe joint is the weakest part of the pipe network,where accident occured frequently.However,due to the high requirements of the professional detection knowledge for inspectors and unregularly detection process,the detection cost is high and the detection efficiency is low.In this paper,the automatic detection system for ultrasonic inspection of polyethylene pipe fusion joints is developed.The automatic identification and safety assessment of the fusion joint defects are realized based on the BP neural network algorithm and the data management module is further developed for the core function.The main contents are as follows:(1)According to the characteristics of the detection image of polyethylene pipe fusion joints,the BP neural network structure was designed to achieve the accurate classification of image detection feature areas,and the application scope of automatic defect recognition was broadened.The BP neural network method was used to study the training samples and extract the similar eigenvalues of similar images.The test proved that the method has good adaptability to the classification of the single and complex detection image feature regions.(2)According to the requirement of automatic identification of software functions for defects,a software for automatic detection of flaws in ultrasonic testing of polyethylene pipe fusion joints was prepared,and the development of an automatic defect identification module for the system was completed.This module implements the features of typical defect classification,automatic defect identification,image marking and safety assessment.Simulating the actual conditions of the project,the experimental software of the recognition software written in this paper was verified by experiments.The results show that the software has a high accuracy in defect recognition,and also has a good detection effect for non-standard detection images.(3)The research and development of the system data management module was completed,the data communication software was designed and programmed,the real-time data transmission between the testing instrument and the host computer was completed,and the remote monitoring of defect detection was realized.A fusion joint defect detection database was established,the cloud management of inspection documents was realized,and the feedback and instant reading and writing of detection results for polyethylene pipe fusion joints using NFC technology was achieved.The design and integration of the entire system standardizes the management process of inspection information and facilitates the promotion of nondestructive testing in the field of defect detection of electrofusion joints.