Method for predicting sulfur cycle coupling denitrifying phosphorus removal process performance in two stages

• Main Authors: XIANG XIANG, ZHOU QINHAO, BI XINQI, JIANG JINQI, XIANG PENGYU, ZHANG JUN, ZHOU LICHANG, GUO GANG
• Format: Patent
• Language: Chinese; English
• Published: 10.10.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a two-stage prediction and optimization method for sulfur cycle coupling denitrification phosphorus removal process performance. The method comprises the following steps: firstly, respectively compiling data sets for sulfate reduction amount and phosphorus removal rate; in the first stage, anaerobic stage process parameters are used as input, and a machine learning algorithm, a cross validation method and a search method are selected to predict the sulfate reduction amount; in the second stage, the technological parameters of the anoxic stage and the predicted value of the sulfuric acid reduction amount serve as input, and a machine learning algorithm, a cross validation method and a search method are selected to predict the phosphorus removal rate; in the model training process, the model is evaluated through the decision coefficient R2 and the root mean square error RMSE, and the optimal model is selected according to the evaluation result. And building a scheduling platform accordin