Research on Soft Sensor Modeling of Support Vector Machine for Wastewater Treatment

Abstract

To improve the prediction accuracy of water quality indexes such as BOD (Biochemical Oxygen Demand) in wastewater treatment process, a novel soft sensor modeling method based on support vector machine (SVM) is designed. The Gaussian kernel function is configured for the proposed method, and the grid search method is combined with K-fold cross-validation to search the optimal values of Gamma and C parameters, thereby improving the prediction accuracy of the proposed model. Finally, the method is tested by using the production data of wastewater treatment. The experimental results show that the proposed model has high prediction accuracy, which provides an effective method for guiding practical production.