Research On Safety Control Method Of Intelligent Coke Oven

In recent years,due to the continuous expansion of infrastructure construction,the demand for iron and steel increases,which leads to the strengthening of iron and steel production,coking and other important links in iron and steel production are also constantly reforming and expanding.Due to the complex production process,bad working environment,uncertain factors and numerous risks and hidden dangers in the process of coking process,safety problems in the coking industry have been difficult to be effectively solved.In addition,the government’s requirements for industrial safety continue to improve,but also make enterprises become more attention to safety production,and then promote the relevant industry personnel on industrial safety issues research.Risk management and safety prevention and control are important guarantees to achieve safe production.Taking coke ovens as the core and aiming at the existing safety problems in coke ovens,the design of appropriate safety control methods is a prerequisite to ensure the safety of coking process.In this paper,according to the statistical information of the safety accidents occurred in the coking process in the past,the accident causes and accident injury evaluation are carried out,and the appropriate safety control methods are put forward to prevent and control the accidents.In this paper,the collision accident of four trucks in coke oven with high probability,gas poisoning accident with great harm and explosion accident with high complexity of dust removal are selected as the main research objects.The main research contents are summarized as follows:(1)Aiming at the collision accident of four trucks in coke ovens,this paper proposes to use multi-agent system to reduce the probability of truck collision accident.Combined with the multi-agent system,the multi-agent structure system suitable for the safety control of four large vehicle crashes is completed.According to the function,each single agent in the system is designed,and the multi-agent cooperative evolutionary algorithm is selected to help complete the coordination work among the various intelligences in the system.(2)Aiming at the gas poisoning accident,which is the most harmful accident in the coking industry,the causes of the gas accident are analyzed in detail,and the CO concentration in the gas existing in the monitoring area is collected by using the established on-site real-time monitoring system.Based on the analysis of the existing industrial gas data,the gas leakage source is inferred.Meanwhile,a modified MMo E prediction model based on Jaya algorithm is proposed to carry out the follow-up simulation and prediction processing of gas accidents.Based on the original MMo E model,this model optimizes the output of each expert network through Jaya,which reduces the negative influence of gating weight,retains the promoting effect of gating weight,and improves the predictive ability of the model.(3)In view of dust removal explosion accident,multi-agent technology is used to strengthen the safety monitoring of dust removal environment area.Meanwhile,in order to avoid strong interference caused by dust,MADDPG model based on attention mechanism is used to optimize the system,which greatly reduces the negative impact brought by dust interference and guarantees the anti-interference of the system.