A Combined Model for CODcr Concentration Forecasting Based on Hybrid Taboo Search-Support Vector Machines

Jiayang Wang,Shanguang Wu,Zhuo Wang,Wei Liu,Xiaoli Zhang

2014-01-01

Abstract:A new Combined Model for CODcr concentration forecasting was presented based on Hybrid Taboo search-Support vector machines. In support vector machines, parameters error punishment factor C and nucleus function parameter were not only played a large role in the precision of SVM, but also had a great influence on the performance of machine. So, an optimized Support vector machines by HTS (HTS-SVM) model is provided, in which Taboo Search was improved by combining Immune arithmetic (IA) and Simulated Annealing (SA) to set up a HTS model, and HTS was used to optimize the key parameters of support vector machines (SVM) to make enhancement on the forecasting effect of SVM. HTS-SVM model was used in forecasting the concentration of the chemical composition CODcr. Two forecasting way were adopted, one was by the historical monitoring data, other was by the industrial discharge pollutants. By forecasting the concentration of CODcr in two cases, the results show that this method not only provides a simple and practical method for water quality forecasting, but also has a better forecasting effect, especially for small sample forecasting problem.