Research On Regional Risk Based On Big Data Analysis Of Accidents In Metallurgical Industry

With the increasing progress of computer science and technology and the rapid development of information technology,information security management has penetrated into all kinds of production enterprises.Metallurgical industry is not only an important basic industry of China's national economy,but also an important support in traditional industries,and plays an important role in the national economy.However,its traditional passive safety management and post safety management mode can not adapt to the development of metallurgical industry.Combining the information management technology of metallurgical safety production accidents with regional risk analysis is an effective means to solve the safety management problems faced by the metallurgical industry.Firstly,this paper expounds the related concepts of big data of metallurgical accidents,and briefly introduces the establishment process of My SQL database.By introducing the related data acquisition technology such as web crawler,this paper analyzes the metallurgical process,main equipment and facilities and layout,and designs the acquisition method of metallurgical accident big data.Then,through the analysis of the obtained big data of metallurgical accidents,combined with the accident description index of metallurgical industry,the time-space and overall analysis of the accident data of metallurgical industry are carried out,and the relevant characteristics of metallurgical accident data are obtained.The obtained nonparametric data and accident severity are tested by Pearson chi square test,and the accident index associated with accident severity is obtained.Then,in order to accurately select the accident prediction index,fully mine and analyze the obtained accident data information,and determine the number of metallurgical accidents,accident casualty linkage coefficient and monthly average accident rate as the evaluation index of the prediction model.According to the characteristics of gentle accident data,BP neural network prediction model,grey prediction model and Gaussian fitting prediction model are selected for comparative research.The optimal prediction model is determined by obtaining the goodness of fit R~2of each model.Based on the optimal prediction model,the accident probability of eight major regions in metallurgical enterprises in 2021 is predicted.Finally,based on the four accident indicators associated with the severity of the accident,the index system of the regional risk model is constructed.The eight regional risk values are calculated through the improved regional risk assessment method,and are superimposed with the eight regional accident prediction results to obtain the regional risk analysis results.The research on big data of metallurgical accidents in this paper effectively improves the ability of enterprise information security management.Based on accident big data,the combination of accident data prediction and regional risk analysis model can help enterprises predict the occurrence law of accidents in advance,help enterprises formulate corresponding accident prevention measures in advance,provide strong support and help for enterprises to scientifically and effectively implement regional safety risk management and control,and also provide reference for the government's safety management and control of metallurgical industry.