Acoustic Analysis And Noise Reduction Scheme Optimization In The Large Space Engine Room

When the equipment works in the plant or engine room, it can make a lot of noises. Theperson who works under this high decibel noises for a long time will suffer mental stress,distraction and other problems that could have bad effect on the health and efficiency. Thenoise is a big threat for the safety of the buildings, because the vibration which transmitsfrom equipment to the floor and building envelope could cause secondary vibration noiseand the acoustic vibration coupling could threaten the safety of construction.Now, theprecision of calculation in large space sound field is low. This makes the noise control planharder to optimize. So there is no ideal control result. In this study, by collecting noise signal,using blind source separation algorithm can accurately separate the major equipments’ noisesignals in the large space, optimized noise reduction scheme by the spectal characteristics ofthe noise signals, and included in the building envelope and the vibration-acoustic couplingof the mechanical equipment, achieved vibration and noise reduction purposes.First of all，by comparing different kind of blind source separation algorithm andconsidering the advantage and the noise signal characteristic of different kinds of algorithm.The blind source separation algorithm which is small amount of computation, lowcomplexity, and can effectively resist interference is the chosen one. Based on the simulationresult, theHigher-order statistics of blind source separation algorithm is higher accurate. Inorder to better evaluate the accuracy of blind source separation, paper selected crosstalk errormethod for spatial noise within the signal characteristics.Because crosstalk error method forsome special global matrix can not make an accurate evaluation, the traditional methodscrosstalk error is optimized, the global matrix of variance and correlation coefficients of eachelement is introduced. Therefore, the crosstalk error method to evaluate the accuracy for aparticular matrix is improved,the impact of strong interference signals in the space isexcluded, and the accuracy of calculation of the noise field is improved.Secondly, collect real noise signal mixed in the laboratory, different noise signalsseparated, evaluate the separation results by using optimized crosstalk error method, andprove the effect ofblind source separation.Finally, based on the data of large supermarket realmixed noise groundwater source heat pump room, separated from the noise signal generatedby different devices for acoustic signal feature extraction, it’s combined withmethod of a vibration absorber and pipeline damping down to solve noise problems. Optimize the impactabsorber absorption coefficient of the perforated plate aperture, perforation rate andthickness of the structure parameters. Meanwhile, separate the secondary signal containsnoise components heat pump for floor vibration, and build room Vibration and Acousticcoupling model included the impact vibration and acoustic coupling on the sound field.Based on the numerical simulation results, the two intermediate massdamperis the greatestimpact on the damping effect. The last intermediate mass dampers optimized parameter is200kg. After the testing, we know that the vibrationsare significantly weakened by using theoptimized parameters damper.Pump room noise level reduces14.6dB, vibration amplitudefrom4mm/s reducesto0.7mm/s, vibration and noise reduction achieves good results. Thisresearch improves the multi-device spatial noise signal separation accuracy and provides atheoretical basis for noise control optimization program.