透過您的圖書館登入
IP:3.141.244.201
  • 學位論文

利用類神經網路預測建築物在干擾效應下之設計風載重

Application of neural network on the predications of interference effects on buildings design wind load

指導教授 : 鄭啟明

摘要


台灣地區屬於亞熱帶島嶼,冬天有東北季風,夏秋時節又常受到颱風侵襲,由於高層建築物屬於柔性建物,對風的影響較為敏感,因此在設計時必須考慮風力對高層建築物的影響,其中鄰近的高層建築物間的相互影響也是在設計時必須考慮的重要因素。高層建築物間的干擾效應其影響的參數很多,如:兩建築物間的距離、建築物的幾何形狀及上游流場地況等。本研究中,利用類神經網路對已存在有限的高層建築物之干擾係數資料庫進行預測。研究所使用的為倒傳遞類神經網路屬於類神經網路中的一種網路模式,主要架構為輸入層、輸出層及單層隱藏層,並透過改變輸入層及隱藏層神經元個數進而找尋較佳之網路模式。研究結果指出,在A地況中(α=0.32),倒傳遞類神經網路具有較佳之預測結果,其誤差較合理,尚可符合高層建築物進行初步耐風設計的需求;C地況中(α=0.15),其預測誤差較大,特別是在橫風向背景部分的類神經網路預測值與風洞試驗值之間存在很大的差異,此部分差異極可能源自實驗數據,有待進一步釐清。

並列摘要


High-rise building is sensitive to wind. In Taiwan, typhoon is a problem, it can bring large wind load on buildings. Wind force is an important lateral design load for high-rise buildings, especially in interference effects of tall buildings. The interference phenomenon between two adjacent tall buildings involves many parameters, such as distance of the two adjacent buildings, buildings’ geometry shape and the upstream terrain conditions. In this study the Artificial Neural Network technique was applied on an existing limited aerodynamic database to predict tall buildings’ interference effects. The ANN framework includes input layer, output layer and a single hidden layer. The optimal number of neuron was selected through detailed parameter studies. The results indicate that the back-propagation ANN model can predict interference effects of tall buildings up to a reasonable margin of error in terrain A (α=0.32), at least for the initial building design stage. In terrain C (α=0.15), however, there exist significant deviations between ANN predictions and wind tunnel measurements especially in the acrosswind background part. This deviation is likely due to some unidentified errors during wind tunnel measurements.

參考文獻


[2] 林倚仲, 2005, “干擾效應對高層建築設計鋒利的影響”, 淡江大學土木工程學系碩士班.
[1] 王嘉國, 2006, “干擾效應對高層建築設計鋒利的影響(II)”, 淡江大學土木工程學系碩士班.
[3] Armitt, J. & Counihan, J., 1968, " The Simulation of the Atmospheric Environment ", Vol.2.﹐pp.49-71.
[4] Counihan﹐J.﹐1970﹐" An Improved Method of Simulation Atmospheric Boundary Layer "﹐Atmospheric Environment﹐Vol.4﹐pp.159-275.
[5] Counihan﹐J.﹐1970﹐" Further Measurements in a Simulated Atmospheric Bounday Layer "﹐Atmospheric Environment﹐Vol.4﹐pp.159-275.

被引用紀錄


陳品鈞(2017)。以類神經網路建構CFD數值模擬之風力頻譜修正模式〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846/TKU.2017.00582
劉博溢(2015)。以類神經網路建立半圓頂型屋蓋結構子午線上風壓頻譜之估算模式〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846/TKU.2015.01001
林昶志(2014)。氣動力資料庫建築抗風設計之風力係數與風力頻譜估算模式〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846/TKU.2014.01163
吳建緯(2013)。以類神經網路建立建築物風力係數與風力頻譜之估算模式〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846/TKU.2013.00204
杜富元(2012)。對於不同幾何建築物在干擾效應下之設計風載重的影響〔碩士論文,淡江大學〕。華藝線上圖書館。https://doi.org/10.6846/TKU.2012.01021

延伸閱讀