基于聯(lián)合最大均方差異的軸承跨工況故障診斷

打開(kāi)文本圖片集

引用格式：，，.基于聯(lián)合最大均方差異的軸承跨工況故障診斷[J].現(xiàn)代電子技術(shù)，2025，48（10）：179-186.

關(guān)鍵詞：滾動(dòng)軸承；跨工況任務(wù)；故障診斷；聯(lián)合最大均方差異；距離度量函數(shù)；特征提取中圖分類號(hào)：TN911.7-34；TP277 文獻(xiàn)標(biāo)識(shí)碼：A 文章編號(hào)：1004-373X（2025）10-0179-08

Abstract：Inalusion to the problem thatroling bearing vibration signalsapproximatelyobeythe normal distributionand thefeaturedistributionisinconsistentunderdiferentworkingconditions，adistancemetricfunctionthatcancomprehensively reflectthe meanand variance informationof thedata samples inthereproduced kernel Hilbert spaceandsimultaneouslyfit the marginaldistributionofthedepthfeatureandtheconditionaldistributionisproposed，i.e.，thejointmaximalmeansquare diference（JMMSD）.byusing theoriginal vibrationsignalasinput，thesourceandtargetdomainsarefeature extractedbyfeature extractiolayerwiththesame structure;Thejointmaximum meansquare diffrenceisintroduced tocombine thefeatureand outputlabelinginformationbetweendiferentdomains，soastoachievethesimultaneousalignmentofthemarginaland conditionaldistributionsinanunsupervisedfieldcross-conditionscenario.Thediagnosticperformanceofthelgorithmisverifid tooutperformthigatonlagpoachhichussolyonlobaldoindaptation，inireto-coiofult diagnosis tasks using the Case Western Reserve University bearing fault dataset.

Keywords：rolling bearing;cross-condition task；faultdiagnosis；joint maximummeansquaredifference；distance measurement function;feature extraction

0 引言

旋轉(zhuǎn)機(jī)械廣泛應(yīng)用于能源供應(yīng)、航天航空、軌道運(yùn)輸?shù)阮I(lǐng)域，滾動(dòng)軸承作為旋轉(zhuǎn)機(jī)械的核心部件，長(zhǎng)期工作在惡劣環(huán)境下，且運(yùn)行工況多變，及時(shí)評(píng)估滾動(dòng)軸承的工作狀態(tài)，能夠確保設(shè)備的運(yùn)行安全，極大地減少人力和物力的損耗。（剩余8306字）