Advanced Wastewater Analysis: AI-Integrated Flow Injection Analysis (FIA) System for COD Online Monitoring

NCUIR > Engineering College > Executive Master of Environmental Engineering > Electronic Thesis & Dissertation > Item 987654321/93777

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/93777

Title:	Advanced Wastewater Analysis: AI-Integrated Flow Injection Analysis (FIA) System for COD Online Monitoring
Authors:	陳韻天;Chen, Yun-Tien
Contributors:	環境工程研究所在職專班
Keywords:	化學需氧量;化學需氧量線上監測系統;流動注射分析;人工智慧;Environmental Engineering;Chemical Oxygen Demand;Artificial Intelligence;Flow Injection Analysis;Industrial Effluents;Chloride Ion Removal
Date:	2024-01-24
Issue Date:	2024-09-19 17:36:35 (UTC+8)
Publisher:	國立中央大學
Abstract:	在面臨全球暖化、氣候調節逐漸異常，聯合國永續發展目標（Sustainable Development Goals, SDGs）的執行，與水資源永續使用的需求更加突出。透過廢水水質自動監控技術的發展，利用有限的人力達到廢水處理系統的自動化與最佳化，以提升系統之穩定性及處理效率，已然成為發展的趨勢。本研究旨在驗證人工智慧化學需氧量（AICODOMS）線上監測系統，在監測工業廢水化學需氧量方面的有效性及系統可靠度。該系統結合了流動注射分析（FIA）與人工智慧（AI）的科技，用以檢測水體中的化學需氧量。AICODOMS利用AI的影像識別及學習功能，自動判斷滴定終點，以此減少手動滴定及人為操作可能的誤差，也提升檢測的準確性和可靠性。本研究針對高雄市某塑膠製造廠廢水廠放流水之化學需氧量作為實驗基質，進行兩個階段的驗證實驗。首先，針對不同濃度的化學需氧量標準溶液、廢水廠之河放汙水、海放汙水樣本，分別進行八重複的試驗。結果顯示AICODOMS與水樣的已知濃度具有高度的一致性和可靠性。第二階段，將AICODOMS直接安裝於工廠的廢水放流口進行即時線上測試，並將其結果與實驗室的分光光度計與第三方實驗室的重鉻酸鉀滴定法進行比較。AICODOMS在廢水廠河放口中與紫外光分光光度計（R2 = 0.9239）和滴定法（R2 = 0.9446）的檢測結果均顯示高度相關性。另外，針對海放口的驗證，與分光光度計（R2 = 0.9175）和滴定法（R2 = 0.9017）的結果也呈高度相關性，具有一定程度之穩定性及準確性。 ;Environmental engineering evolves to tackle modern challenges, with this study focusing on validating the artificial intelligence chemical oxygen demand online monitoring system (AICODOMS). AICODOMS, combining flow injection analysis (FIA) and artificial intelligence (AI), precisely assesses chemical oxygen demand (COD) in aquatic environments. This research aims to verify AICODOMS′ effectiveness in monitoring COD levels within industrial effluents. Comparative analyses with field datasets highlight its potential as an alternative to traditional COD assessment methods. It demonstrates precision and reliability, reducing human-induced errors and offering applications in industrial wastewater management, emergency response, and innovative AI detection techniques. The study encompasses two experimental phases using wastewater samples from a plastic manufacturing facility. Initially, AICODOMS was validated using treated wastewater and standard COD solutions, including real effluent samples discharged into the river and sea. Strong agreement between AICODOMS and established methods validated its reliability. Additionally, AICODOMS showed strong correlations with spectrophotometric and titrimetric methods for both river (R2 = 0.9239, R2 = 0.9446) and sea discharge (R2 = 0.9175, R2 = 0.9017). In addition, AICODOMS successfully integrates a chloride ion removal technology that can precisely measure COD under high chloride concentrations. The integration of AI-driven color recognition technology with FIA holds promise for efficient and eco-friendly industrial wastewater analysis. This amalgamation could revolutionize assessment methodologies, enhancing efficiency and environmental sustainability in industrial analysis.
Appears in Collections:	[Executive Master of Environmental Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	13	View/Open

社群 sharing

Loading...