透過您的圖書館登入
IP:3.140.186.241
  • 學位論文

應用小波理論與類神經網路於RC結構內管線洩漏之非破壞檢測

The application of wavelet theory and artificial neural network in the nondestructive testing of leakage within inner pipelines of RC structures

指導教授 : 王安培

摘要


台灣地處環太平洋地震帶,地震頻傳,以及台灣氣候屬於亞熱帶海島型氣候,考慮此兩項環境因子下,為了抵抗地震與潮濕的氣候,建築結構物大多是使用混凝土為主要建材,在加上國內建築管線都採用結構體中暗管施工方式,當在發生地震後或者結構物經過使用後管線劣化,管線因為混凝土的包覆導致使用者無法察覺出異常。直到發生管線堵塞、水流變小、漏水、鋼筋腐蝕、以及壁癌產生時,都已經錯過最佳的維修時間,這時的維修費用與時間相對成本較高。 任何建築物經過使用,都會老化或劣化,甚至在建造時就已經發生損壞,國內目前對於建築內管線檢測,大多都由經驗豐富的工程人員,沿管線檢查不尋常的現象,像是可見的滲漏現象以及異常聲音。這種檢查方式,準確性非常依賴檢測人員的經驗判斷。因此,本研究以聽音法為基礎透過實驗來收集不同類型滲漏聲音,利用小波理論結合多尺度熵與主成分分析來提取各類型滲漏聲音訊號特徵,並使用類神經網路結合各滲漏聲音訊號特徵建構出管線滲漏檢測系統。

並列摘要


Taiwan is loacted in the Circum-Pacific Seismic Zone, where earthquakes happen frequently. In addition, Taiwan’s climate is defined by the classic subtropics island weather characteristics. Taking these two factors into consideration, concrete became the main material for building construction in Taiwan. However, as pipelines are usually placed inside the building structures and wrapped in concrete, their breaking or leaking cannot be easily detected after an earthquake; by the time that other physical symptoms (such as pipe blockage, mildew and corroding iron) begin to appear on the wall, the most economical period for the repair has long passed, causing the owner of the building more money and time to repair. Every building will eventually become old and deteriorated; some will even experience pipeline breakage when construction is not yet complete. Nowadays, a successful inspection of an interior pipeline system is based on the experience of the project personnel. For example, to be able to accurately detect leakage through sound depends much on the skills and expertise of the inspector. This research focuses on auditory methods by collecting different kinds of sounds that inform leakage within the internal pipeline system. Furthermore, it combines signal processing with Artificial Neural Networks technology to promote accuracy and flexibility in the detection of pipeline malfunction within a building structure.

參考文獻


39. 張淑勤,「模糊決策在防洪之應用」, 中原大學土木工程學系,碩士論文, 民國95年7月
38. 陳漢廷,“特徵擷取於橋樑非破壞檢測之應用”,中原大學土木工程研究所,碩士論文,民國93年7月.
41. 許展綸,“應用小波理論與HHT於橋梁之非破壞性檢測”, 中原大學土木工程研究所,碩士論文,民國96年1月.
37. 林建成,“應用類神經網路於二維橋樑非破壞檢測的應用”, 中原大學土木工程研究所,碩士論文,民國92年7月
36. 林昆達,“小波理論與類神經網路在橋樑非破壞檢測的應用”, 中原大學土木工程研究所,碩士論文,民國91年7月

被引用紀錄


劉志豪(2017)。高速鐵路引致地盤振動之自動預測模式評估〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu201700041

延伸閱讀