Petrochemical site human health risk identification method based on index system and XGBoost

Abstract

The invention discloses a petrochemical site human health risk identification method based on an index system and XGBoost, and the method comprises the steps: constructing a hierarchical architecture of a petrochemical site human health risk identification index system from the perspective of "source-approach-receptor"; by considering that the indexes contain both classification variables and continuous variables, based on the difference of the indexes, dividing the values or attributes of the indexes and resetting the values, wherein the larger the values of the indexes are, the more likely the human health risk is to be caused; when the indexes with the reset values serve as input, and whether petrochemical site pollution exceeds the human health acceptable risk or not serves as output, establishing an XGBoost model and performing grid search, wherein the optimal hyper-parameter is selected with the accuracy rate being as the judgment standard. According to the technical method provided by the invention, the XGBoost identification model is established based on a small number of easily available indexes, the human health risk of the petrochemical site can be rapidly and efficiently identified. Also, the method can be expanded to human health risk identification of other industrial lands.

Description

technical field [0001] The invention relates to the technical field of risk identification of petrochemical sites, in particular to a method for identifying human health risks in petrochemical sites based on an index system and XGBoost. Background technique [0002] Human health risk assessment of contaminated sites is currently a hot topic. According to reports, 30% of industrial sites have soil pollution problems, which pose a threat to human health. Therefore, it is imminent to realize rapid and efficient identification of human health risks in industrial sites. As the pillar industry of our country, the petrochemical industry contributes 20% of the total industrial economy of our country. However, the main raw materials of the petrochemical industry are mostly petroleum fractions, natural gas, etc., and the production process is complicated, and a large amount of pollutants will be generated during the production process, mainly including volatile organic compounds, ph...

AI Technical Summary

Problems solved by technology

However, the main raw materials of the petrochemical industry are mostly petroleum fractions, natural gas, etc., and the production process is complicated, and a large amount of pollutants will be generated during the production process, mainly including volatile organic compounds, phenols, polycyclic aromatic hydrocarbons, petroleum hydrocarbons, sulfides, cyanides, Acid, Hg, Pb, Ni, Cd, etc., are mostly key regulatory pollutants, seriously threatening the soil environment and human health

[0003] At present, my country mainly conducts risk assessment of soil pollution on construction land according to the "Technical Guidelines for Soil Pollution Risk Assessment of Construction Land" (HJ 25.3-2019). The assessment is based on site investigation, and the specific process includes hazard identification, exposure assessment, toxicity assessment, and risk characterization. However, it is difficult to identify the characteristic pollutants of the site. The process of relying on conventional sampling and analysis methods to determine whether the site is polluted and has human health risks is quite complicated, time-consuming, and costly. It is impossible to realize soil pollution investigation and risk control in a short time. The rapid identification method predicts whether the site pollutants exceed the acceptable risk of human health, and provides reference for relevant departments to carry out soil pollution screening and risk control work

[0004] At present, relevant research literature shows that "source-pathway-receptor" is the main perspective to study soil pollution and human health risks, but the existing index system considers many factors, and it is necessary to analyze the production process, However, these data are difficult to obtain and process. The present invention aims to establish a human health risk identification index system for petrochemical sites based on easily obtained site-related data, and quickly identify petrochemical sites through machine learning methods. Does pollution outweigh human health risks?

Images