基于支持向量机和频域信号熵的电梯急停故障检测方法

打开文本图片集

中图分类号：TP277 文献标志码：B 文章编号：1671-5276（2025）03-0266-06

Elevator Emergency Stop Fault Detection Method Based on Support Vector Machine and Frequency Domain Signal Entropy

LI Zixiang，FAN Yu （Henan Special Equipment Inspection Technology Research Institute，Zhengzhou 4535Oo，China）

Abstract：Amethodfor elevator emergency stop fault detection basedonsupport vector machineandfrequencydomain signal entropyisproposedconceing thehighcomplexityof elevatoremergencystopfaultdetection.Usinganelevatorvibrationanalyzer to colectvibration signalsduring elevatoroperation，thevibration signalsaredecomposedintolow-frequencyand highfrequencycomponentfrequencybandsbydiscreteorthogonalwaveletmethod.Thedetailedsignalsandapproximatesignalsof eachflooroftheelevatorareextracted，theentropyenergyofthefrequencydomainsignalsofeachflooriscalculated，andthe signals are inputintoasupport vector machineupon normalization.Theimproved geneticalgorithm isappliedtooptimize the parametersof thesupport vector machine，obtainingtheoptimal support vector machine clasifier，achieving elevator frequency domainsignal entropyclassification，thuscompleting elevatoremergencystopfaultdetection.Teexperimentalresultsshowthat theoptimized SVMhas higherfitnessthanothermethods，withitselevatorfaultemergencystopdetectionresultsbeingconsistent withtheactualelevatoroperationnoes，andthatthemethodhasgoodapplicationeffectandefectivelyimprovestheabilityof elevator emergency stop fault detection.

Keywords：supportvectormachine；frequencydomainsignalentropy；elevatoremergencystop；faultdetection；improvedgenetic algorithm；wavelet decomposition

0 引言

电梯已成为高层建筑标准配置设施，为人们的生活提供便利[1-3]。（剩余5537字）