【摘要】：廠房或機房內(nèi)設(shè)備工作時產(chǎn)生的噪聲很大，當人們長時間工作在高分貝噪聲環(huán)境下時，會產(chǎn)生精神緊張、注意力分散等問題，影響人們身心健康和正常的工作效率，設(shè)備工作的同時也會將振動傳遞到樓板及圍護，令其受迫振動產(chǎn)生二次噪音，，樓板等圍護結(jié)構(gòu)產(chǎn)生的振聲耦合對建筑安全存在影響。目前，大空間聲場計算精度較低，導(dǎo)致噪聲治理方案難以優(yōu)化，治理效果不夠理想。本課題通過收集噪聲信號，運用盲源分離算法有效精確地分離大空間內(nèi)主要設(shè)備的噪聲信號，根據(jù)噪聲信號頻譜特性優(yōu)化降噪方案，并計入圍護結(jié)構(gòu)和主要機械設(shè)備的振聲耦合，達到更好的減振降噪目的。 首先比較不同的盲源分離算法，考慮不同算法的優(yōu)點以及噪聲信號特性，選取計算量小、復(fù)雜程度低，且能有效抵抗干擾的盲源分離算法，通過仿真實驗結(jié)果確定基于高階統(tǒng)計量的盲源分離算法分離精度更高。為了更好地評價盲源分離精度，本文選取適合空間內(nèi)噪聲信號特性的串音誤差方法，由于串音誤差方法對于某些特殊的全局矩陣不能做出精確評價，因此優(yōu)化了傳統(tǒng)的串音誤差方法，引入全局矩陣中各元素的方差和相關(guān)系數(shù)，提高了串音誤差方法對于特殊矩陣的評價準確度，從而能排除空間內(nèi)強干擾信號的影響，使噪聲場計算精度更高。 在此基礎(chǔ)上，在實驗室中采集真實的混合噪聲信號，分離出不同的噪聲信號，運用優(yōu)化的串音誤差方法評價分離結(jié)果，證明盲源分離算法的分離效果。最后，采集某大型超市地下水源熱泵機房的真實混合噪聲，分離不同設(shè)備產(chǎn)生的噪聲信號，針對提取的聲信號特性，分別采用了組合式吸聲體和管道減振的方法對其進行減振降噪綜合治理。優(yōu)化了影響吸聲體吸聲系數(shù)的穿孔板孔徑、穿孔率和厚度結(jié)構(gòu)參數(shù)。同時，分離信號中含有熱泵機組對樓板振動產(chǎn)生的二次噪音成份，建立機房振聲耦合模型，計入振聲耦合對聲場的影響根據(jù)數(shù)值模擬計算結(jié)果，認為二級減振器中間質(zhì)量塊對減振效果影響最大，優(yōu)化了減振器中間質(zhì)量塊參數(shù)為200kg。經(jīng)測試可知，優(yōu)化參數(shù)后的減振器對樓板的振動影響明顯減弱，熱泵機房噪聲值降低了14.6dB，振動幅值從4mm/s降低為0.7mm/s，取得了良好的減振降噪效果，課題的研究提高了多設(shè)備空間內(nèi)噪聲信號分離的精度，為噪聲治理方案優(yōu)化提供理論依據(jù)。
[Abstract]:When people work in a high decibel noise environment for a long time, they will produce problems such as mental tension and distraction, which will affect people's physical and mental health and normal working efficiency. At the same time, the vibration will be transferred to the floor and enclosure, which will make the forced vibration produce secondary noise. The vibration and acoustic coupling produced by the floor structure will have an impact on the safety of the building. At present, the accuracy of large space acoustic field calculation is low, resulting in the noise control scheme is difficult to optimize, the treatment effect is not ideal. By collecting noise signals, the blind source separation algorithm is used to effectively and accurately separate the noise signals of the main equipment in large space, and the noise reduction scheme is optimized according to the spectrum characteristics of the noise signals. The vibration and acoustic coupling of the envelope structure and the main mechanical equipment is taken into account to achieve better vibration and noise reduction. First of all, comparing different blind source separation algorithms, considering the advantages of different algorithms as well as the characteristics of noise signal, the blind source separation algorithm is selected, which has the advantages of small computation, low complexity, and can effectively resist interference. The simulation results show that the blind source separation algorithm based on high order statistics is more accurate. In order to evaluate the accuracy of blind source separation better, the crosstalk error method suitable for noise signal characteristics in space is selected in this paper. Because the crosstalk error method can not accurately evaluate some special global matrices. Therefore, the traditional crosstalk error method is optimized, the variance and correlation coefficient of each element in the global matrix are introduced, the accuracy of the crosstalk error method for the special matrix is improved, and the influence of the strong interference signal in the space can be eliminated. The accuracy of noise field calculation is higher. On this basis, the real mixed noise signals are collected in the laboratory, and different noise signals are separated. The optimized crosstalk error method is used to evaluate the separation results, and the separation effect of the blind source separation algorithm is proved. Finally, the real mixed noise of a large supermarket underground water source heat pump machine room is collected, and the noise signals produced by different equipments are separated. The combined sound absorber and pipeline are adopted to reduce vibration and noise respectively. The hole diameter, perforation rate and thickness structure parameters affecting the sound absorption coefficient of the sound absorber were optimized. At the same time, the separation signal contains the secondary noise produced by the heat pump unit to the floor vibration, and the vibration-acoustic coupling model of the machine room is established, and the effect of the vibration-acoustic coupling on the sound field is taken into account according to the numerical simulation results. It is considered that the intermediate mass block of the two stage absorber has the greatest influence on the damping effect, and the parameter of the intermediate mass block of the damper is optimized to be 200 kg. The test results show that the vibration of floor slab is obviously weakened by the vibration absorber with optimized parameters, the noise value of heat pump room is reduced by 14.6 dB, the vibration amplitude is reduced from 4mm/s to 0.7 mm / s, and good vibration and noise reduction effect is achieved. The research improves the accuracy of noise signal separation in multi-equipment space and provides the theoretical basis for the optimization of noise control scheme.
【學(xué)位授予單位】：沈陽工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TB535

【參考文獻】

相關(guān)期刊論文 前10條

1 馬英,許建華,朱翔;淺析城市電廠的噪聲治理[J];電力環(huán)境保護;2003年02期

2 林秋華;鄭永瑞;殷福亮;;基于參考獨立分量分析的語音增強方法[J];大連理工大學(xué)學(xué)報;2006年06期

3 游榮義,陳忠;一種基于ICA的盲信號分離快速算法[J];電子學(xué)報;2004年04期

4 劉海林;;基于廣義特征值的病態(tài)混疊盲源分離算法[J];電子學(xué)報;2006年11期

5 李麗君;李紅艷;剛憲約;柴山;;聲固耦合系統(tǒng)數(shù)值計算方法的研究與應(yīng)用[J];工程設(shè)計學(xué)報;2010年06期

6 孟兆龍;;建筑工程中的噪聲防治研究[J];工業(yè)安全與環(huán)保;2011年11期

7 陳永強;王宏霞;;一種強魯棒性的盲分離混合矩陣估計方法[J];電子與信息學(xué)報;2012年09期

8 肖明,謝勝利,傅予力;卷積混疊盲信號分離的一種線性化方法[J];華南理工大學(xué)學(xué)報(自然科學(xué)版);2005年03期

9 楊杰;關(guān)麗坤;任學(xué)平;;不同步長下自然梯度算法分離狀態(tài)的比較[J];機械設(shè)計與制造;2012年07期

10 趙冠軍;劉更;吳立言;;基于模態(tài)疊加法的聲固耦合噪聲仿真與實驗[J];機械科學(xué)與技術(shù);2007年12期

相關(guān)博士學(xué)位論文 前1條

1 費朝陽;多激勵系統(tǒng)隔振降噪功率流及振聲能量耦合理論與應(yīng)用[D];沈陽工業(yè)大學(xué);2012年

本文編號：2401304