本文選題：動(dòng)平衡機(jī)測量系統(tǒng)　切入點(diǎn)：不平衡信號提取　出處：《上海大學(xué)》2014年博士論文　論文類型：學(xué)位論文

【摘要】：不平衡信號提取的精度和效率是影響動(dòng)平衡機(jī)測量系統(tǒng)精度和效率的主要因素，確定不平衡信號和轉(zhuǎn)子校正面上等效不平衡量之間關(guān)系的系統(tǒng)標(biāo)定的準(zhǔn)確性也是影響動(dòng)平衡測量精度的關(guān)鍵因素之一。本文從提高硬支承動(dòng)平衡機(jī)通用測量系統(tǒng)的精度和效率出發(fā)，在論述動(dòng)平衡測量系統(tǒng)原理和誤差分析的基礎(chǔ)上，分別對近頻干擾和轉(zhuǎn)速波動(dòng)環(huán)境下不平衡信號的高效、高精度提取方法以及永久標(biāo)定法開展了深入研究。主要研究內(nèi)容和特色貢獻(xiàn)包括： （1）根據(jù)硬支承平衡機(jī)測量原理，給出了通用動(dòng)平衡測量系統(tǒng)的組成，分析了硬支承平衡機(jī)測量系統(tǒng)的幾種誤差來源，重點(diǎn)討論了近頻干擾和轉(zhuǎn)速波動(dòng)對動(dòng)平衡測量精度和效率的影響。 （2）針對低速平衡系統(tǒng)的信噪比差、易產(chǎn)生近頻干擾導(dǎo)致的不平衡提取精度和效率不高等問題，通過仿真分析了近頻干擾時(shí)快速傅立葉變換（FFT）法和相關(guān)法兩種傳統(tǒng)不平衡信號提取方法的性能，結(jié)果表明這兩種方法都需要較長的數(shù)據(jù)樣本才能保證一定的不平衡信號提取精度，且對近頻干擾信號的相位、頻率等參數(shù)變化敏感。為克服傳統(tǒng)法的不足，提出一種基于諧波小波和Prony算法的不平衡信號提取方法，并分別進(jìn)行了仿真和現(xiàn)場實(shí)驗(yàn)分析。結(jié)果表明，即便有多個(gè)近頻干擾信號時(shí)，該方法只需采用較短數(shù)據(jù)樣本序列便能準(zhǔn)確提取出不平衡信號的幅值和相位，且提取精度對干擾信號的相位、頻率等參數(shù)的變化不敏感，具有較高精度和魯棒性。 （3）對于轉(zhuǎn)速波動(dòng)的轉(zhuǎn)子系統(tǒng)，振動(dòng)信號具有明顯的非平穩(wěn)性，如用平穩(wěn)信號分析法提取不平衡信號會帶來較大誤差。由此本文提出非平穩(wěn)轉(zhuǎn)速下兩種不平衡信號提取方法：一是基于等相位重采樣的不平衡信號提取方法，即根據(jù)瞬時(shí)頻率對等時(shí)采樣的振動(dòng)信號進(jìn)行等相位重采樣，可以改善傳統(tǒng)快速傅立葉變換（FFT）法的相位提取精度；二是基于經(jīng)驗(yàn)?zāi)B(tài)分解（EMD）和最小二乘法（LSA）的不平衡信號提取方法，即根據(jù)瞬時(shí)頻率構(gòu)造不平衡信號，再從EMD濾波后的振動(dòng)數(shù)據(jù)中截取樣本數(shù)據(jù)，由LSA識別出不平衡信號的幅值和相位。仿真和實(shí)驗(yàn)結(jié)果表明，當(dāng)轉(zhuǎn)速波動(dòng)時(shí)，兩種方法都可以明顯改變不平衡信號提取的精度和穩(wěn)定性，當(dāng)信噪比較小時(shí)，基于EMD和LSA的不平衡信號提取方法的性能更好，適合于工程應(yīng)用。 （4）永久標(biāo)定法是硬支承動(dòng)平衡機(jī)通用測量系統(tǒng)中常用方法，具有簡單、實(shí)用等特點(diǎn)。本文詳細(xì)分析了硬支承平衡機(jī)永久標(biāo)定法的原理誤差，采用數(shù)字仿真的方法給出了原理誤差與轉(zhuǎn)子質(zhì)量和頻比之間的關(guān)系。針對傳統(tǒng)永久標(biāo)定法忽略慣性力和阻尼帶來的測量誤差，提出一種基于Thiele連分式插值法的永久標(biāo)定改進(jìn)方法，實(shí)現(xiàn)了寬轉(zhuǎn)速范圍內(nèi)的較高的測量精度。 （5）研制了具有自主知識產(chǎn)權(quán)的高精度動(dòng)平衡機(jī)通用測量系統(tǒng)。硬件系統(tǒng)由模擬信號處理板、數(shù)據(jù)采集卡和工業(yè)單板機(jī)組成，軟件系統(tǒng)采用WindowsXP環(huán)境下的C++Builder編制，界面友好，具有很強(qiáng)的擴(kuò)展功能。系統(tǒng)的最小可達(dá)剩余不平衡量優(yōu)于0.1gmm/kg，一次不平衡減少率優(yōu)于95%，能滿足大部分旋轉(zhuǎn)機(jī)械動(dòng)平衡試驗(yàn)的要求，具有很強(qiáng)的通用性。
[Abstract]:Don't balance the precision and efficiency of signal extraction are the main factors influencing the dynamic balance machine measurement system accuracy and efficiency, determine the unbalanced relationship between the signal and the accuracy of the calibration system of rotor unbalance on the equivalent positive school is the key factors affecting the measurement accuracy of a dynamic balance. This paper improve the efficiency and accuracy of hard bearing balancing general measurement system based on dynamic balance principle and error analysis of measurement system based on of near frequency interference and speed fluctuation under unbalanced signal, high precision, extraction methods and permanent calibration method is researched. The main research contents and contributions include the characteristics:
(1) according to the measuring principle of hard support balancing machine, the composition of universal dynamic balance measurement system is presented. Several error sources of the measurement system of hard bearing balancing machine are analyzed. The influence of near frequency interference and speed fluctuation on the accuracy and efficiency of dynamic balance measurement is mainly discussed.
(2) poor signal-to-noise ratio for a low speed balancing system, easy to produce near frequency interference unbalance extraction accuracy and efficiency is not high, through the simulation analysis of nearly frequency interference fast Fu Liye transform (FFT) method and the correlation method of two kinds of traditional extraction methods of unbalanced signal, the results show that the data sample can these two methods will require a longer guarantee certain unbalanced signal extraction accuracy, and the phase of near frequency interference signal, frequency and other parameters sensitive. In order to overcome the deficiency of the traditional method, this paper presents an approach to extract unbalance signal and harmonic wavelet based on Prony algorithm, and analyzes the simulation and field the experimental results, even if there are multiple near frequency interference signal, the method only needs to use short data sequence can accurately extract unbalance signal amplitude and phase, and the extraction accuracy of interference signal phase The changes of the bit, frequency and other parameters are not sensitive, and have high accuracy and robustness.
(3) for rotor speed fluctuation, vibration signals are clearly non-stationary, such as non-stationary signal analysis method to extract unbalance signal will bring great errors. This paper puts forward two kinds of non-stationary unbalanced signal extraction methods: one is based on the unbalanced signal extraction method of phase re sampling. According to the sampling of vibration signals of instantaneous frequency equal phase resampling, can improve the traditional fast Fu Liye transform (FFT) method, the phase extraction accuracy; the two is based on empirical mode decomposition (EMD) and least squares (LSA) method to extract unbalance signal, according to the instantaneous frequency of structural unbalance signal, and then from the vibration data of the EMD filtered interception of sample data by LSA identify the amplitude and phase unbalance signal. Simulation and experimental results show that when the speed fluctuation, the two methods can significantly change the unfair The precision and stability of the signal extraction, when the signal to noise is relatively small, the performance of the unbalanced signal extraction method based on EMD and LSA is better. It is suitable for engineering applications.
(4) permanent calibration method is widely used, hard bearing balancing machine universal measurement system has the advantages of simple, practical and so on. This paper analyzes the principle error of hard bearing balancing machine permanent calibration method, using the method of digital simulation is presented and the principle error of rotor mass and frequency ratio between the traditional measurement error. Permanent calibration method ignores inertia and damping caused by the proposed Thiele continuedfraction interpolation method based on improved permanent calibration, the measurement accuracy of a wide speed range high.
(5) developed with independent intellectual property rights of the high precision dynamic balancing machine general measuring system. The hardware system consists of analog signal processing board, data acquisition card and industrial microcomputer, software system using WindowsXP under the environment of C++Builder software, friendly interface, strong expanding function. The minimum achievable residual unbalance system the amount is less than 0.1gmm/kg, an unbalance reduction rate is less than 95%, can satisfy the majority of rotating machinery dynamic balance test, has a strong versatility.

【學(xué)位授予單位】：上海大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2014
【分類號】：TP274;TH877

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 常廣;鄢素云;王毅;;零相位數(shù)字濾波器在非平穩(wěn)信號處理中的應(yīng)用[J];北京交通大學(xué)學(xué)報(bào);2011年06期

2 傅俊慶;廖坤鵬;沈中偉;;旋轉(zhuǎn)機(jī)械振動(dòng)信號角域與時(shí)域采樣方法的對比研究[J];長沙交通學(xué)院學(xué)報(bào);2007年01期

3 王茂海,劉會金,彭輝,張國強(qiáng);采樣信號頻率偏離設(shè)計(jì)值情況下離散傅立葉變換的誤差分析[J];電測與儀表;2001年01期

4 陳平;李慶民;趙彤;;瞬時(shí)頻率估計(jì)算法研究進(jìn)展綜述[J];電測與儀表;2006年07期

5 江亞群,何怡剛;基于加窗DFT的相位差高精度測量算法[J];電路與系統(tǒng)學(xué)報(bào);2005年02期

6 胡勁松;楊世錫;;轉(zhuǎn)子振動(dòng)信號同步整周期重采樣方法的研究[J];動(dòng)力工程;2008年03期

7 喻勝,閆波,陳光，

本文編號：1586712