Research On The Traceability Model Of Industrial Toxic Gases

In recent years,with the rapid development of chemical,pharmaceutical and other industries,sudden explosions,toxic substances and other catastrophic accidents occur frequently.Once the incidents occur,results must be difficult to control,which may bring serious threat to life and property.Therefore,it is necessary to understand the occurrence of such accidents and monitor the source of toxic gases.In the case of emergency,according to the monitored data,the source strength and location of the source are back calculated,and the range and extent of the accident are determined,furthermore,reasonable guidance for the disaster escape route and rescue measures are provided.This paper gives various gas diffusion models and compares them,analyzing their characteristics and application scope,and selecting suitable forward simulation models for the diffusion of toxic gases.Summarize the current various source strength inverse calculation methods,and discuss the advantages and disadvantages of various methods.The differential evolution algorithm is used in the single source inverse calculation,and compared with other optimization algorithms.Differential evolution algorithm has shorter time,but the optimization effect is very significant,and it performs well in both accuracy and running time.It is very effective to apply it to source inverse calculation.In complex multi-source inverse computation,it is difficult to calculate the source strength and location of multiple source points simultaneously.A method based on minimum error iteration is proposed in this paper.Compared with the intelligent optimization algorithm,the minimum error iterative algorithm does not need a given initial range,using the information of the source point more efficiently,the optimization efficiency is improved,and the higher accuracy can be guaranteed.Finally,the minimum error iteration algorithm and differential evolution algorithm are combined to improve their operation accuracy and speed with their respective advantages.In this dissertation,the experiment part uses the MATLAB processing data and establishes the inverse model,uses the Gauss gas diffusion model as the forward simulation data generation model,programming the single source and multi source inverse calculation method,gives each evaluation index,and assesses the efficiency and accuracy of each algorithm.The experiment shows that the result of differential evolution is best in the single source inverse algorithm.In the multi source inverse calculation,the hybrid algorithm combined with the minimum error iterative algorithm and the differential evolution algorithm can most quickly and effectively counter the source strength and location,and provide effective basis for supervision and rescue.