During the monitoring process,due to the need for a large area of monitoring points,a large amount of manpower and material resources are required.At the same time,the existing soil monitoring points have redundancy and low representativeness,which has an impact on the accuracy and comprehensiveness of the monitoring results.Therefore,this article uses a soil point optimization method based on autoencoder dimensionality reduction and graph convolutional network(GCN)inference to achieve the goal of comprehensively representing regional pollution with as few monitoring points as possible by dividing monitoring points into different clusters.Specifically,this article uses a soil point optimization method that combines graph convolutional network inference and autoencoder dimensionality reduction.This method utilizes the structural information and higher-order feature representations of soil monitoring point data to cluster monitoring points into different clusters.By analyzing the correlation,it was found that the clustered center points can serve as representatives of the original cluster,reducing the redundancy of soil monitoring points and achieving optimization of monitoring points.Finally,experiments have shown that the GCN method can effectively reduce the number of monitoring points while maintaining the accuracy and comprehensiveness of monitoring results.
万方数据
维普
