The Study Of Early Warning Model Of Heavy Metal Pollution In Soil

It is helpful to do the early warning job for the government to make the protection decision of the soil.It also helps a lot for people to live a better life.But since there are a lot of problems the research is made hard to go forward. The monitoring job takes a long time and costs much.And the technology of monitoring is not matured enough.By analyzing the correlation between soil pollution and the indicators from the industry, the i-ndicators that can reflect the soil pollution degree are been chosen,which can solve the data deficient problem.The results are as follows.(1) Based on the data of Huangshi City from 1984 to 2004, a universal indicator system for the early warning system is established.The experiment shows that SO2 in the air from the industry, heavy metals from wastewater, the output of the industry and the deal of the useless water all have strong correlation with the heavy metal pollution,which helps find out the early waring indicators. To make the early warning model versatile,the concept of "regional difference index" is proposed,which makes the model be fit for different regions.(2) The concept of " characteristic pollution year" is brought up which supports the establishment of the evaluation system of standards, which may be helpful for others.(3) By using LDIW-PSO algorithm for optimization of BP network the forecast data of indicators is got. And the experiment proves that it can improve the convergence of the network.(4) Using Multi-Agent technology builds the model.By the JADE development platform the following agents are built : BP Agent, Indicator Proceeding Agent and Forecast Agent.which make up the multi-agent system.(5) The test for Beijing proves the reliability of early warning model.The model can be assisted for the departments of the soil pollution.