Research On Workplace Fire Risk Assessment Based On Big Data Platform And Bayesian Network

Fire risk assessment has always been a continuously studied subject.Most of the research methods focus on the exploration of the fire process and need to rely on the knowledge of experts.However,with the advent of the period of big data,the field of fire protection in China has developed in the direction of “intelligent fire protection”.Therefore,the method of fire risk assessment has more possibilities.In response to the call of comprehensively promoting “intelligent fire protection”,large data platforms for fire protection are constantly being developed and improved.In Jiangsu Province,the Jiangsu Province Enterprises and Institutions Fire Safety Management Platform has been put into use,and a large number of fire data from workplaces have been collected.Based on the fire safety management system and gathered data,this paper applies Bayesian network to the system to evaluate the fire risk level of units.The main research contents of this paper are as follows:(1)Analyzing the fire safety management status of workplaces and the fire safety management system.By figuring out the historical fire accidents and inquiring the results of spot check of fire agencies,remaining problems of fire safety management are pointed out,for example,the superficial rules and the serious formalization.Then,analyzing the Jiangsu Province Enterprises and Institutions Fire Safety Management Platform,including study its operation mode,data type and the existing scoring method,and indicating the data source of this paper.(2)Exploring the method of Bayesian network construction for workplace fire risk assessment based on the statistics in the system.Referring to the existing manual assessment method Guidance for Assessment of Fire Prevention and Control Ability in Enterprises and Institutions,in this paper,the focus of fire risk assessment is on the estimation of fire prevention and control ability.The Bayesian network of workplace fire risk assessment construction method is as followed: Firstly,setting the inspection items in the system as the edge node of Bayesian network.Secondly,constructing the Bayesian network structure referring to the assess index in the Guidance.Thirdly,processing the data according to the situation of the workplace.Finally,using computer aided software to do the parameter learning and establish the entire Bayesian network of workplace fire risk assessment.Taking an actual enterprise or institution as example,constructing a Bayesian network of fire risk assessment by using the method proposed in this paper,and analyzing based on the results.(3)Visualization design of Bayesian network in Fire Safety Management System.Doing user interface design by studying the design style of the system,to facilitate the administrator’s management.In this paper,Bayesian network and large data platform are combined to evaluate the fire risk of workplace,which improves the module of fire risk assessment in the system and contributes to the fire safety management of workplace.Provides a new idea for the fire risk assessment under the background of “intelligent fire protection”.