壓縮機系統不穩定現象之偵測

Abstract

本篇論文主要是在探討壓縮機系統不穩定現象之偵測，論文中我們針對所量測到的訊號應用了不同的信號分析技巧作壓縮機不穩定現象之偵測，並與現有的錯誤偵測技術作比較。首先，我們簡述當不穩定現象發生時系統機構可能遭受的影響與破壞。其次，經由1986年Moore和Greitzer所提出的壓縮機模組，我們介紹壓縮機系統的特性。為了能夠早期偵測到不穩定現象的發生，本論文應用了Short-Time Fourier Transform(STFT)、Wigner-Ville Distribution(WVD)以及M-channel filter bank等方法來作壓縮機不穩定現象之偵測。經由電腦模擬，我們將所獲得的結果與現有的錯誤診斷濾波器(Fault Identification Filter)應用在壓縮機系統所獲得的結果作比較。

In this thesis, we study the detection of surge and rotating stall in a compression system based on 1986 Moore and Greitzer’s Model. From the measurable signals, we employ several analysis techniques in signal processing to detect such unstable behaviors, and then compare these results with those obtained by Fault Identification filter design technique. We also briefly explain the possible influence and damage to the equipment of compression system when the unstable behaviors occur. The characteristic and stability results of compression system are also included. In order to detect the occurrence of the unstable behaviors at the onset, we apply several methods including Short-Time Fourier Transform, Wigner-Ville Distribution and M-channel filter banks to process the signals which are measured from the compression system. Simulation results are also given to demonstrate the effectiveness of these employed techniques.