Soil is the carrier of life,but the development of chemical enterprises will inevitably cause damage to the soil,and the accumulation of pollutants will seriously harm the safety of all kinds of life.Therefore,it is urgent to repair and treat the site pollution.However,the investigation,assessment and remediation of site pollution are faced with many problems,such as multi-source heterogeneous data,inability to quickly extract pollution characteristics from large-scale site data and difficulty in selecting drilling sampling sites.To solve these problems,this paper proposes to use Hadoop platform to manage multi-source heterogeneous data,use parallel computing technology to quickly extract pollution characteristics,and build site pollution situation assessment method based on Cat Boost algorithm,so as to provide reference for drilling sampling site selection.Finally,relevant results and technologies are integrated into the visual system.In order to provide data support for relevant departments to control or repair site pollution.The main research contents of this paper are as follows:(1)Study the storage and calculation of multi-source heterogeneous contaminated site data based on Hadoop platform.In order to effectively organize and manage contaminated site data,Hadoop platform was built through Linux system,and based on Map Reduce computing framework,parallel computing technology was studied to extract descriptive statistical features,spatial distribution features and risk features of contaminated sites,and thematic maps of site pollution distribution were drawn by Arc GIS.(2)Study the pollution situation assessment method based on Cat Boost.Based on the standards and regulations related to environmental investigation,the pollution situation assessment index was established,and the pollution situation assessment method was proposed with Cat Boost algorithm as the core.The accuracy evaluation results show that this method can realize the assessment of site pollution situation with high precision without using the later survey data,and the identification accuracy of medium and high risk plots on the demonstration site is 73.3% and 81.4% respectively,which has high practical value.(3)Design and develop visual system for site pollution situation assessment.Based on the field all party and government management department and the field repair demand analysis,system architecture design with portal as the core,through the Vue login page framework development,home page,data management,data visualization,risk assessment,pollution situation assessment six big modules,so that the relevant departments more intuitive understanding of the field of pollution,at the same time,provide data support. |