A Water Quality Evaluation Method of Projection Pursuit Regression Based on Ant Colony Algorithm in Nanning Urban River

Pollution problem is getting worse in China´s urban rivers at present. This paper, after analyzing the factors affecting urban river water quality and choosing such six principal evaluation factors as BOD₅, CODcr, petroleum, volatile phenol, NH₃-N and total phosphorus, establishes a Projection Pursuit analysis model for evaluating water quality in urban rivers, optimizes the evaluation model with the help of Ant Colony Algorithm, and applies the model in the evaluation and ordering of the water quality in ten urban rivers in Nanning. The research shows that Projection Pursuit Regression Analysis in water quality evaluation helps avoid the errors caused by the subjective reasons as manifested in traditional evaluation methods, and, as the evaluation results are accurate and the method is simple, it can serve as a new way to evaluate the water quality in China´s urban rivers.