改良式A2O系統好氧槽與膜濾池系統動力模式之發展與建立;Development and establishment of system dynamics model of aerobic tank and membrane tank in modified A2O system

NCUIR > Engineering College > Graduate Institute of Environmental Engineering > Electronic Thesis & Dissertation > Item 987654321/78041

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/78041

Title:	改良式A2O系統好氧槽與膜濾池系統動力模式之發展與建立;Development and establishment of system dynamics model of aerobic tank and membrane tank in modified A2O system
Authors:	李宗霖;Li, Zong-Lin
Contributors:	環境工程研究所
Keywords:	改良式A2O系統;系統動力模式;類神經網路;Modified A2O system;System dynamic model;Artificial neural network
Date:	2018-08-24
Issue Date:	2018-08-31 15:09:48 (UTC+8)
Publisher:	國立中央大學
Abstract:	改良式A2O活性污泥處理系統是透過微生物作用將水中碳、氮、磷去除以達到滿足流水標準之目的，而一般A2O廢水處理系統屬半開放性的系統，外界環境變化皆會影響污水水質特性，進而影響處理效能。因此若無法及時因應水質變化而做出合適的操作控制，將可能導致出流水質惡化。本研究以系統動力學概念分析系統內各項物質與微生物間的關聯與作用機制，並以此為基礎發展與建立改良式A2O系統好氧槽與膜濾池系統動力模式，同時利用類經網路進行模式內微生物反應速率推估，以作為系統動力模式運作之輸入參數並計算系統出流水質。將模擬結果與實際出流水質數據比較，MLSS、氨氮、硝酸鹽氮及正磷酸鹽的模擬結果與實際濃度的相對誤差為3.57%~196.65%之間，但其絕對誤差約小於8.1919mg/L，為可容許的誤差範圍，表示此系統動力模式具備有效推估該物質濃度的能力，而SCOD模擬濃度與實際濃度之絕對誤差最大值為24.07mg/L，其模擬結果並不如其他水質項目推估結果良好，可能是因為單一槽時SCOD推估結果的誤差逐步累積至整體系統誤差。而利用第二次實驗的數據進行模擬時，模擬結果與實際數據誤差較第一次增加，比較兩次採樣進流水質雖有變化但皆在第一批次實驗數據範圍附近震盪，因此判定是因為系統菌相改變導致。;The modified A2O activated sludge treatment system removes carbon, nitrogen and phosphorus from water by microbial action to meet the flow standard. Generally, the A2O wastewater treatment system is a semi-open system, and changes in the external environment will affect the water quality characteristics of the sewage, thereby affecting the treatment efficiency. Failure to properly control the water quality in a timely manner may result in deterioration of the outflow water quality. This study uses the concept of system dynamics to analyze the reaction mechanism between various substances and microorganisms in the system, and based on this, develops and establishes the dynamic mode of the aerobic tank and membrane filter system of the modified A2O system. At the same time, the rate of microbial reaction in the model is estimated by using artificial neural network. Compare the simulation results with the actual outflow water quality data, the relative error between the simulation results of MLSS, ammonia nitrogen, nitrate nitrogen and orthophosphate is 3.57%~196.65%, and the absolute error is less than 8.1919mg/L, which is an allowable error range. This result indicates that the model can effectively estimate the concentration of the substance. The absolute error of the simulated concentration and actual concentration of SCOD is 24.07 mg/L, and the simulation results are not as good as other water quality project estimation results, which may be caused by the gradual accumulation of errors in the SCOD estimation results in a single tank. When the data of the second experiment is used for simulation, the error between the simulation result and the actual data increases. Comparing the two experimental data, although the influent water quality has changed, it is oscillated around the first batch of experimental data. Therefore, the judgment caused the error because the system phase changes.
Appears in Collections:	[Graduate Institute of Environmental Engineering ] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	290	View/Open

社群 sharing

Loading...