Optimization design of noise reduction in pipeline systems

• Main Authors: Che-Ren Yeh, 葉哲仁
• Other Authors: Der-Yuan Liou
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/03658534815252048073

碩士 === 國立臺灣海洋大學 === 系統工程暨造船學系 === 94 === The pipeline ventilating system is common in the factory, office building and household, among them the air blower is the noise source and the noise can propagate along the pipeline into the inner room or radiate from the tube surface. This paper tries to design the acoustic parameters of pipeline system by using Taguchi method, grey relational analysis and artificial neural network. We first sort information of silencers in the market and make a database, next establish firmly the influential parameters(factors) of the silenced pipeline system, and give these parameters reasonable levels, finally we can get the best group to match the results by using the orthogonal table of the Taguchi method and analysis of variance(ANOVA). This paper sorts the pipeline ventilating system into a single qualitative characteristic problem and multiple qualitative characteristics problem in dealing with the silenced ventilation system. The orthogonal table of the Taguchi method deals with the single qualitative characteristics. We use the grey relational analysis to simplify the multiple qualitative characteristics problem to single qualitative characteristics problem. The test results showed that the silencer is the best influence factor for noise attenuation in ventilation systems, next are the length of pipe and bulged tube length, the elbow pipe diameter is the least. Because the orthogonal table of the Taguchi method is the partial factor experiment, in order to confirm the result of the best group, this paper uses the artificial neural network to establish the ventilation system forecast model, and we found that the obtained results both are to closely meet, therefore the results in this paper show that the Taguchi method process the ventilation system is feasible based on the grey relation in grey theory. Also prove that we can use the fewer time to obtain many related ventilation system silencers information and the best parameters design.