臺中都會區環境大數據AI智能勾稽運算評估查核驗證計畫
| 中文摘要 | 近年來行政院環境保護署大力推動空品物聯網布建，本計畫特針對中區轄區四工業區加裝微氣象感測器，並以空拍取得高解析度工廠座標，運用AI環境大數據智慧分析平台，分三季進行污染風險分析，有效限縮稽查對象；接著針對這些工廠申報資料間異常勾稽，最後決定其中部分工廠，施以科技儀器蒐證採樣與深度查核驗證作業，此做法不僅提升稽查成效，發揮智慧稽查效果，其結果進一步對照同時段平台的分析結果進行比對確認，兩者間的確具備相當的一致性。另一方面，基於目前CEMS查核分析數據需以人工方式進行，資料處理量相當龐大，需耗費大量人力時間，無法有效篩選異常數據，可能導致查核前分析CEMS監測數據資料失真，或誤判為違反規定。本計畫發展CEMS資料自動化分析技術，包括資料前處理模組、異常偵測模組、資料視覺化模組，並提供網頁供即時操作查詢，最終在臺北市內湖資源回收廠進行一根次測試，其結果的確增加人員對CEMS監測數據判讀能力，能減少人員判讀時間，降低人為作業失誤，有效提升CEMS查核量能。這些執行成果均都彙整於各項推廣活動中，包括專諮會與研討會、推廣影片、新聞稿等，期待提升我國環保單位科技執法的能力。 | ||
|---|---|---|---|
| 中文關鍵字 | 空氣污染、AI智慧稽查、排放管道連續自動監測系統 | ||
基本資訊
| 專案計畫編號 | 經費年度 | 110 | 計畫經費 | 7940 千元 | |
|---|---|---|---|---|---|
| 專案開始日期 | 2021/01/21 | 專案結束日期 | 2021/11/30 | 專案主持人 | 曹志成 |
| 主辦單位 | 中區環境督察大隊 | 承辦人 | 蘇純平 | 執行單位 | 銓崴國際有限公司 |
成果下載
| 類型 | 檔名 | 檔案大小 | 說明 |
|---|---|---|---|
| 期末報告 | 摘要詳細版.pdf | 0MB |
Using Big Data Algorithms to Enhance the Current Auditing and Inspecting Operation in the Urban Area of Taichung City
| 英文摘要 | With the EPA’s recent development of implementing air quality IOT projects widely in Taiwan, the project was required to choose four industrial parks from central Taiwan to install micro-scaled wind speed and wind direction sensors and use high-definition drone cameras to acquire higher resolution geographical coordinates of the plants in the parks, which would be pre-set in the TrackAIR big data platform to automatically identify the pollution emitting events and their probable polluting sources accordingly. Then a discrepancies investigation among the various environmental permitting documents would be followed for the previously identified plants, which would provide facts for the inspectors to make their plans to either conduct a surrounding survey or enter the target plant for source testing with appropriate scientific tools. Such approach will better the effectiveness and efficiencies of the inspection in a smarter way. Comparing the scientific tools analysis and TrackAIR’s findings for the same time period, it suggested a pretty good consistency. Another issue that CEMS auditing is quite labor intensive due to the large amount of data and information needing manual processing, which therefore sometimes leads to mis-interpretate the CEMS monitoring data, or falsely judge their conformity to legitimate requirements. An automatic assessment technology has been developed as a computer program including the Data Pre-processing Module, Abnormality Detection Module, Data Visualization Module, and a Web Page to access all processed information. The program was tested at the Taipei Neihu Recycling Plant, which demonstrated its benefits in terms of increasing the inspector’s data handling capabilities, reducing auditing time consumed, minimizing human errors, and enlarge auditing capacities. All the above results have been summarized in the promoting activities including an expert consultation seminar, two workshops, promotion films and press release drafts, which were aimed to expedite the adoption of advanced technologies on environmental law enforcements. | ||
|---|---|---|---|
| 英文關鍵字 | Air pollution,，AI smart auditing， CEMS | ||