濾波及細化於非穩定振動之齒輪組故障診斷

Fault Diagnosis for Unsteady Vibration of Gear Train based on Vold-Kalman Filter and Zooming

張紹中(Shao-Chung Chang)

指導教授 ： 康淵 ； 張永鵬

中原大學/工學院/機械工程研究所/碩士(2010年)

https://doi.org/10.6840/cycu201001021

摘要

本文以快速傅立葉轉換(fast Fourier transform, FFT)、短時傅立葉轉換(short-time Fourier transform, STFT)、細化譜(zoom spectrum)、倒頻譜(cepstrum)及弗德卡爾曼濾波(Vold-Kalman filter)等分析方法，對行星齒輪減速機所產生之固定轉速及變動轉速的振動信號進行故障診斷分析。
對於馬達轉速為隨機變動的非穩態(unsteady)振動信號來說，直接將振動信號進行傅立葉轉換無法完整表達其時變性頻率特徵，因此利用(1)短時傅立葉轉換以短時窗函數建立振動信號之時間與頻率對應之關係，(2)細化譜分析將全域頻譜中抽出重要頻段，以相同頻率採樣數得到細微的頻率分布結構，(3)倒頻譜得到全域頻譜或細化譜重複頻線出現的頻率，有助於辨識頻率或幅值調制(modulation)的頻譜，(4)弗德卡爾曼濾波將譜線濾掉非關特徵頻率的部份，只呈現及凸顯出特徵頻率。

關鍵字

短時傅立葉 ； 弗德卡爾曼濾波 ； 倒頻譜 ； 快速傅立葉 ； 故障診斷 ； 細化譜

並列摘要

This paper use the fast Fourier transform, short-time Fourier transform, zoom spectrum, cepstrum and Vold-Kalman filter analysis method to analyze planetary gear reducer generated by fixed and variable speed vibration signals for fault diagnosis.
The motor rotational speed for the random unsteady vibration signals, the vibration signals directly to the Fourier transform can not completely express its time-varying frequency characteristics Therefore, using (1) short-time Fourier transform to short time window function to establish the time and frequency of vibration signals corresponding to the relationship, (2) zoom spectrum analysis extracts the closed region frequency spectrum in the important frequency band, obtains the slight frequency distributed structure by the same frequency sampling number, (3) cepstrum obtains the frequency which the closed region frequency spectrum or the refinement spectrum redundant frequency line appears, is helpful modulates in the identification frequency or the peak-to-peak value the frequency spectrum, (4) Vold-Kalman filter out the spectral line the non-pass eigen frequency the part, only presents and highlights the eigen frequency.

並列關鍵字

fault diagnosis ； zoom spectrum ； Vold-Kalman filter ； short-time Fourier transform ； cepstrum ； fast Fourier transform

參考文獻

1.Kang Y., Wang C. C., Chang Y. P., Hsueh C. C. and Chang M. C., “Certainty Improvement in Diagnosis of Multiple Faults by Using Versatile Membership Functions for Fuzzy Neural Networks,” Lecture Notes in Computer Science, 3973, pp. 370-375, 2006.

2.Wen, k. L. and Qiang, Z., “Deconvolutive Short-Time Fourier Transform Spectrogram”, Signal Processing Letters, Vol. 16, pp. 576 – 579, 2009.

3.Pan, M. C. and Lin, Y. F., “Further Exploration of Vold Kalman Filtering Order Tracking with Shaft-Speed Information-I: Theoretical Part, Numerical Implementation and Parameter Investigations,” Mechanical Systems and Signal Processing 20, 1134–1154, 2006.

4.董宏, 王碧琴, ”軋機齒輪箱異常振動分析及故障診斷,” 中國設備工程, 第7期, 頁46-47, 2005.

5.王俊洪, ”煤氣鼓風機組增速機齒輪故障診斷分析,” 設備管理與維修, 第12期, 頁30-32, 2002.