Research Of Accident Analysis And Prediction Of Chemical Manufacturers

China is a big country of producting and using chemicals.This paper collects the dataabout industrial production accidents happened in chemical plants in China over the lastdecade, found that chemical production accidents are not a declining trend, but fluctuatingtrend shows that China’s chemical safety needs to be further improved. China has successivelyestablished a variety of chemical park in recent years, once the accident happen, theconsequences are very serious, therefore, accident prediction and prevention efforts need todo well. But there is no studies about prediction of chemical accidents so far, so this paperresearch the appropriate forecasting method based on the data characteristics. For nationalmacroeconomic policy-making and national authorities find the exact accident controlpriorities, develop safety management objectives provide an important scientific basis,Guidance the enterprise make effective security management and decision-making.These data are analyzed from several aspects, mainly about the types of the accidentsoccurred, the seasons that accidents happened, the geographic distributions of occurrence andthe causes that the accidents arose. Then the paper summarizes the accident predictionmethods currently used, including their applications, advantages and disadvantages. Finallythe appropriate method that can be used in predicting accidents of chemical production isselected. In order to make more accurate predictions, the selected method is improved, and theapplicability of the improved method is also proved in this paper.In this paper, the statistical data of chemical accidents occurred are collected from2005to2014. In order to find the patterns and characteristics of the accidents, the data are analyzedfrom following aspects: the level of accidents, the types of accidents, happening seasons,geographic distribution, happening causes. Finally the conclusion is made that chemical accidents most likely to occur in the spring and summer; from the point of geographicdistribution, most chemical accidents happened in Jiangsu, Shandong, Zhejiang and Hubeiprovince due to the large number of chemical companies and the lack of supervision in theseprovinces. Therefore, all regions are required to enforce and strengthen the supervision andinspection of production safety. From the view of happening causes, chemical companieswere dealing with some common issues during2014, Firstly, the production safetymanagement, such as fire issue, working in a limited space, improper operations, dustproblem and other issues. Secondly, the lack of supervision from authorities, like illegalbusiness, the dilapidated equipment which cannot meet the safety requirements, and soon.Finally, the paper summarizes the causes and the characteristics of chemical productionaccidents that happen frequently. Then combined with the3E principles of accidentprevention, some measures of accident prevention are suggested according to the aspects ofengineering technology, education and enforcement, etc.According to the characteristics of chemical production data, Grey Markov predictionmethod is selected to predict the chemical production accident. In order to make moreaccurate predictions, Gray Markov prediction method is improved in this paper. A higheraccuracy of the improved method is proved, which makes it applicable for the prediction ofchemical accidents.