本論文使用聲學分析軟體SYSNOISE邊界元素(Boundary element method, BEM)法,其分析主要目的在於研究多層吸音與反射消音箱尺寸設計。 規劃分析模型組數,以I-DEAS軟體建立消音箱模型包含檔板及貼附吸音棉,匯入聲學分析軟體SYSNOISE分析多層吸音與反射消音箱,評估消音箱的傳輸損失(Transmission Loss)。 論文主要分三部份:(1) 消音箱內部檔板片數對於消音性能影響與消音箱檔板放置方式對於消音箱消音性能之改變 (2) 整合類神經網路系統與遺傳演算法之最佳化方法,並搭配聲學分析軟體SYSNOISE於多層吸音與反射消音箱之尺寸最佳化設計。 (3) 消音箱實驗探討反射消音箱內部放置不同類型之檔板,研究其中之消音性能。 經過實驗、理論數值分析的驗證,與模擬的結果相當吻合,對於工業界的應用,可有效地減少開發時間和降低生產成本,提升產品競爭力。
This paper uses acoustic analysis software – SYSNOICE, one kind of boundary element method (BEM). Its main purpose is to study the acoustical-performance analysis for an acoustical plenum with multiple sound absorption and reflection cancellation size design. The model number will be planed. Later, using the I-DEAS software, the outline model with battles and sound- absorbing material is established and imported to the SYSNOISE for further assessment of plenum’s sound transmission loss (TL). Paper is divided into three parts : (1) to investigate the influence of the acoustical plenum’s performance with respect to both the number and allocation of the battles inside the plenum. (2) to optimally design the best shape of the acoustical plenum with multiple sound absorption and reflection by using the number networks and genetic algorism in conjunction with the finite element method (FEM). (3) to assess the acoustical performance for the acoustical plenum using a series of experimental works. An accuracy verification is performed and found to be in good agreement by using the experimental data and the theoretical value. Consequently, the optimization applied in industry corn effectively reduces the design time, lower the manufacturing cost and enhance the product competitiveness.