基于雙邊型直線電機(jī)的故障信息融合診斷方法的研究 

徐鳴飛，余海濤，程帆，胡敏強(qiáng) 
東南大學(xué) 電氣工程學(xué)院，江蘇 南京 210096 
 

摘 要：針對雙邊型直線電機(jī)的氣隙偏心故障、繞組匝間短路故障，提出了一種基于雙邊型直線電機(jī)的故障信息融合診斷方法，該法通過數(shù)據(jù)層融合、特征層融合、決策層融合以及兩級診斷的方式來實(shí)現(xiàn)，其中數(shù)據(jù)層采用了Park矢量融合方法以及改進(jìn)的基8FFT算法，特征層采用了模糊人工神經(jīng)網(wǎng)絡(luò)法。將改進(jìn)的基8FFT算法、模糊人工神經(jīng)網(wǎng)絡(luò)法結(jié)合在一起，提高了雙邊型直線電機(jī)各種故障的診斷能力。
關(guān)鍵詞：雙邊型直線電機(jī)；故障信息融合診斷；改進(jìn)的基8FFT算法；蝶形單元；模糊人工神經(jīng)網(wǎng)絡(luò)
中圖分類號：TM359.4 文獻(xiàn)標(biāo)識碼：A 文章編號：1007-3175(2015)06-0015-04

Study on Fusion Diagnosis Method of Fault Information Based on Bilateral Linear Motor 

XU Ming-fei, YU Hai-tao, CHENG Fan, HU Min-qiang 
School of Electrical Engineering, Southeast University, Nanjing 210096, China 
 

Abstract: Aiming at air gap eccentricity fault and winding turn-to-turn short circuit of bilateral linear motor, this paper proposed a kind of fusion diagnosis method of fault information based on bilateral linear motor. This method consisted of the data level fusion, feature level fusion, decision level fusion and the mode of two level diagnosis, of which the data layer adopted the Park vector fusion method and improved 8FFT algorithm, and the feature level adopted the fuzzy artificial neural network. The improved 8FFT algorithm was combined with the fuzzy artificial neural network to raise the diagnosis ability of various kinds of faults of bilateral linear motor.
Key words: bilateral linear motor; fault information fusion diagnoses; improved 8FFT algorithm; butterfly unit; fuzzy artificial neural network

參考文獻(xiàn)
[1] 王海鋒.基于Park矢量與信息融合技術(shù)的電機(jī)故障診斷[D].曲阜：曲阜師范大學(xué)，2012.
[2] 馬宏忠.異步電動(dòng)機(jī)轉(zhuǎn)子繞組故障后轉(zhuǎn)矩轉(zhuǎn)速的分析與計(jì)算[J].電工技術(shù)學(xué)報(bào)，2004，19(4)：23-27.
[3] 張建文.電氣設(shè)備故障診斷技術(shù)[M].北京：中國水利水電出版社，2006.
[4] 唐貴基，葉進(jìn)生，胡愛軍.基于諧波小波包的微弱信號頻域提取[J].工控技術(shù)，2009(9)：156-159.