Abstract:Based on the traditional water quality (WQ) monitoring section setting of a lake in Beijing, the paper used multi-Agent modeling to achieve the section optimization through PCA and CA. Firstly, the method established WQ Agent model, then taking the comprehensive evaluation score (CES) of principle and non-principle components as the evaluation factors in coarse and fine segmentation respectively, established multi-WQ-Agent model with three layer structure through the interaction. In coarse segmentation model, the adjacent WQ Agent would merge into Agent’ if thier CES of principle components had high similarity; in fine segmentation model, the WQ Agent’ would split into Agent* if the CES of non-principle components had high non-similarity. Finally, the best WQ Agent collection could be obtained and the center point of each WQ Agent* were the optimized monitoring point. Through verification, the result was the same with hierarchical clustering algorithm, showing the method has certain accuracy and provides a new effective way for the optimization of water quality monitoring section layout.