신경망과 유동전류계를 이용한 정수장 응집제 주입제어에 관한 연구 = A Study on the Coagulant Dosing Control Based on Neural Network and Streaming Current Detector for Water Treatment Plant

다국어 초록 (Multilingual Abstract)

Coagulation process is one of the most important processes in water treatment procedures for stable and economical operation, and coagulant dosing of this process for most plants is generally determined by the jar test. However, this method does not only take a long time to analyze and get the result but also has difficulties in applying to automatic control. This paper shows the feasibility of applying neural network to control the coagulant dosing automatically in water treatment plant. To be specific, the predicted results of the neural network model is shown to be similar to that of jar test. The input variables for learning the neural network are turbidity, water temperature, pH, and alkalinity. Combining the neural network and SCD(Streaming Current Detector) for feedforward and feedback control of injecting coagulant, a rapid change of the raw water quality can be accommodated.

참고문헌 (Reference)

1 정우섭, "지능형 응집제 투입 시스템의 개발" 3 (3): 652-658, 1997

2 주대성, "인공신경회로망을 이용한응집제 최적주입률 결정" 15 (15): 23-30, 1999.

3 김용열, "유전-퍼지를 이용한 정수장 응집제 주입률 결정에 관한 연구" 제7호 : 876-882, 1999

4 남의석, "신경회로망을 이용한 상수처리시스템의 응집제 주입 공정 최적화" 3 (3): 644-651, 1997.

5 "상수도시설기준" 한국수도협회 1997.

6 E. L. Bean,S. J. Campbell,F. R. Anspach, "Zeta potential measurements in the control of coagulation chemical doses" 214-227, 1964.

7 S. K. Dental, "Using streaming detectors in water treatment" 81 (81): 85 1989-94,

8 "Operational Control of Coagulation and Filtration Processes" 42-44, 1992.