A machine learning-based method for predicting spatial distribution of soil cadmium pollution

Abstract

The application discloses a kind of based on machine learning's soil cadmium pollution spatial distribution prediction method, it is related to soil pollution distribution prediction technical field. Including: carrying out multi-source heterogeneous data acquisition, the area area of the region to be detected is obtained, the regional attribute item is obtained, the detection range is circumscribed based on regional attribute item, and the regional range item is obtained;Cd pollution factor attribute in regional range item is obtained, and the pollution factor set is obtained.The application circumscribes scientific detection range by multi-source heterogeneous data acquisition, integrates industrial, agricultural, life full-dimensional pollution source data to construct pollution factor set, simultaneously adopts unmanned aerial vehicle hyperspectral scanning and in-situ sensing networking, accurately obtains the key physicochemical properties such as soil pH value, organic matter content, trains resistance model based on different physicochemical property grade's cadmium spread test data, can accurately quantify the actual effect that acidic soil accelerates cadmium migration, alkaline soil retards cadmium diffusion.

Images