Research On Information Entropy-Based Pollution Source Localization In Obstacle Scenes

Pollution sources usually refer to places,equipment or devices that discharge harmful substances or have harmful effects on the environment.The pollutants released by them enter the environmental system at an inappropriate concentration or speed,which will destroy the environment and cause huge human and financial losses,and even endanger human life seriously.Therefore,locating pollution sources has a wide range of application scenarios and certain research value.This article has conducted a study on the location of pollution sources,taking the possible leakage of gas sources in chemical plants as the background to locate the pollution sources.Compared with other pollutants,gas pollutants have unstable gradient information.Gas source localization is a goal that needs to be accomplished in many common scenarios.In order to complete the source location task more efficiently,the main research points of this topic are divided into the following parts:(1)The diffusion of leaked gas can be described by partial differential equations(PDE).The solution of PDE can provide information such as the location of pollution sources and the distribution of pollutant concentration.However,the traditional method has strict requirements for calculation.Therefore,this paper uses the lattice boltzmann method to describe the leakage process of pollutants,so as to obtain data sets.The Extreme Learning Machine(ELM)is used to predict the concentration of the environmental field.The purpose of this is to achieve a more accurate fitting of the environmental field with less computational effort,facilitating subsequent treatment of pollution sources.(2)In the traditional information entropy source searching algorithm,the motion set of this method mostly contains only four directions.The robot can only move in these directions,and its walking step size is fixed.This can affect the efficiency of the robot.In view of the limitations of fixed step size and motion set,the motion set is expanded and an adaptive step size mechanism is added to make the step size of the robot change flexibly with the search process.In this case,it can effectively reduce the path of the robot,shorten the time for the robot to find the source,reach the source location faster,and improve the efficiency of pollution source location.(3)Obstacles will affect the gas diffusion,causing leakage materials to accumulate and form a pseudo source,thus affecting the robot’s judgment.In order to identify the pseudo sources,this paper adopted the source identification method based on statistics.To collect more comprehensive and accurate information,the author links the source identification method based on statistics with Bug1 to avoid the limitation of confirming the boundary setting of the area.Enables the robot to collect more directional concentration information,thereby preventing the robot from misjudging and realizing pollution source localization in obstacle scenarios,so as to improve the source location accuracy.To sum up,this paper studies the pollution source location in the obstacle scene,which is carried out through three parts: pollution source distribution prediction,source location using information entropy algorithm,and pseudo source recognition in the obstacle scene,so as to improve the integrity and accuracy of the pollution source location process.