논문 국내 국내전문학술지(KCI급) 하수처리시설 수질인자의 머신러닝 예측 모델
- 학술지 구분 국내전문학술지(KCI급)
- 게재년월 2023-02
- 저자명 임준묵공동(교신),주형구
- 학술지명 대한산업공학회지
- 발행처명 대한산업공학회
- 발행국가 국내
- 논문언어 외국어
- 전체저자수 2
홈페이지논문 초록 (Abstract)
The purpose of this study is to develop a machine learning-based prediction model for the value of COD, a key factor that measures the quality of sewage flowing from a sewage treatment facility. Considering that the inflowing sewage water quality data has a time-series characteristic, a machine learning model using ARIMAX, RNN, and LSTM was developed as a predictive model for COD. For each model, after learning based on big data collected from domestic J sewage treatment facility, the prediction performance was evaluated by RMSE. ARIMAX model showed an accuracy of 7.83% compared to the average, RNN was 3.62%, and LSTM was 3.56%. Overall, the LSTM model was evaluated as the best performing predictive model. If our model is used in a sewage treatment facility, it is possible to predict the sewage water quality in real time, which is expected to greatly contribute to improving the efficiency of sewage treatment.