基于改进D一S证据理论的刮板输送机断链隐患诊断与预警方法

打开文本图片集

中图分类号：TD528.3 文献标志码：A 文章编号：1671-5276（2025）05-0287-07

Abstract：In responsetothe occurances of scraperconveyor chain breakage and block，amethod for diagnosing and warning hidden dangers ofscraper conveyorchain breakage and block basedonthe characteristics ofchanges in scraper positionand chain tension is proposed.With tension sensors，tension information of the scraper conveyor sprocket is collcted to implement dimensionalityreduction processing onthetension signal characteristics.Usingthe processed tensionsignalfeaturesas the inputof theElman neuralnetwork，thepreliminary judgment resultsof theeasily broken points in the chain areobtained through training and learning.The maximum and minimum closeness method is applied to improve theD -s evidence theory and make a decision level fusion for the preliminary judgment results. The diagnostic signal atthevulnerable pointisconvertedintothescraperoffset distanceandtheheadandtailcurrentdiference，and the thresholdrangeof thewarning signal is divided toachievechainbreakage hazardwarning.Theexperimentalresults show thatthe proposed method can accurately determine whether there isa hidden danger of chain breakage in the scraper conveyor based on the tension of the chain when breaking，thus a timely pre一warning being issued.

Keywords：scraper conveyor; kernel principal component analysis method; Elman neural network; D-S evidence theory; chain breakage hazard warning

0 引言

成威胁。（剩余9267字）