Prediction of Building Leak Reasons and Alteration Cost

• Main Authors: Wen-pin Hsien, 謝文彬
• Other Authors: Min-yuan Cheng
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/87207922215137803729

碩士 === 國立臺灣科技大學 === 營建工程系 === 102 === Taiwan belong island climate with rainy and humid season, summer and fall season also suffered from typhoon frequently, in addition to happen frequent earthquakes, these affect the grand waterproof problem of the building. Most of the buildings and construction devices have the leakage problems, as well as the flaws of waterproof projects quality and facilities, therefore these make the leakage problems become common and severe. In order to find an economical and reliable countermeasures for solving the occurred leakage problems, once leakage occurred, not only the consumers’ living quality are affected, but also the contractor’s reputation is also affected from the public. If leakage problem will not be solved, the other subsequent alterations can’t be solved, too. Therefore, the building’s leakage alteration is a critical issue This study collects current leakage alteration cases, and creates the database system. In order to prevent the alteration estimation by personal subjective judgments, this research applies "θ-means cluster analysis algorithm", to estimate the similar cases with leakage reasons and alteration methods. Then, the results can provide similar leakage cases to execute inference analysis as the search to the owners and engineers. The main results include: (1) Collect, filing and analyze current leakage alteration database about historical case which under construction and finished building cases; (2) Collect relevant leakage literature and organize the leakage affection factors with the attribute’s causality, then set up building leakage reasons and affection factors of alteration cost; (3) Cluster 127 alteration cases with (θ-means) method through analysis and calculus, and set up the leakage reasons of alteration methods and estimation mode of the alteration cost. The clustering result of the cases can find out the most similar case, then provide the reference for the alterations decision.