다변량 통계기법을 활용한 하수처리장 에너지 소비상태의 진단 및 평가 = Diagnosis and assessment of energy consumption using multivariate statistical techniques in wastewater treatment plant

다국어 초록 (Multilingual Abstract)

The meanwhile, wastewater treatment plant(WWTP) has been recognized as a facility that consumes a lot of energy. Besides, most of the WWTPs have been operated in the excessive operation conditions in order to maintain stable wastewater treatment. Therefore in this study, energy consumption conditions diagnosis and prediction technique were developed using multivariate statistical techniques for efficient energy consumption in the WWTP. First, diagnose qualitative of energy consumption in the WWTP was applied by multivariate statistical techniques such as PCA(Principal Component Analysis), K-means clustering analysis and discriminant function. The diagnosis results of energy consumption conditions were classified into three groups:　High energy consumption; Medium energy consumption; and Low energy consumption.　In case of high-energy consumption groups(High, Medium), new operating conditions was applied for energy saving. New operation conditions was derived from simulation within effluent water criterion. Secondarily, prediction model was developed using multiple regression analysis to predict the energy consumption of the WWTP. The R2 value in the regression analysis appeared 67%, and performance of the electric power prediction model had less than ±5% error. Finally, Electric power in new operation conditions at WWTP and electric power in traditional operation conditions at WWTP were compared by predicted model. As a result, high-energy consumption groups result in more energy saving-efficient than low-energy consumption groups and can be economized 1,077 kwh/day averagely.
For energy saving in WWTP, diagnosis and prediction models of energy consumption conditions was developed using multivariate statistical analysis techniques. To utilize these techniques is expected to contribute to the efficient operation of the WWTP.

목차 (Table of Contents)

• 1. 서론 1
• 1.1 연구 배경 및 목적 1
• 1.2 연구 범위 3
• 2. 이론적 배경 4
• 2.1 하수처리장의 에너지 소비현황 4
• 2.2 다변량 통계분석 기법 7
• 2.2.1 주성분분석 7
• 2.2.2 군집분석 8
• 2.2.3 판별분석 11
• 2.2.4 회귀분석 13
• 2.3 활성슬러지 모델 15
• 3. 재료 및 방법 18
• 3.1 연구 대상 18
• 3.2 에너지 소비상태의 정성적 진단 21
• 3.3 최적의 운전변수 도출 25
• 3.4 수학적 모델 및 최적화 26
• 3.5 새로운 운전조건의 범위 설정 28
• 3.6 에너지 소비상태의 정량적 예측 30
• 4. 결과 및 고찰 32
• 4.1 에너지 소비상태의 정성적 진단 32
• 4.1.1 주성분분석 32
• 4.1.2 K-평균 군집분석 36
• 4.1.3 판별 함수의 적용 40
• 4.2 수학적 모델 및 최적화 결과 43
• 4.3 최적의 운전조건 도출 53
• 4.4 에너지 소비상태의 정량적 예측 56
• 4.4.1 다중회귀분석 적용 56
• 4.4.2 통계적 가설 검정 62
• 4.4.3 회귀식의 유효성 판단 64
• 4.4.4 다중공선성 검토 67
• 4.5 에너지 절감량 평가 69
• 5. 결론 73
• 참고문헌 75