Path Optimization Of Safety Inspection Based On Fire Risk Level In Chemical Industry Park

With the rapid development of China’s chemical industry,chemical enterprises continue to focus on the park,the harm caused by fire accidents in the park is becoming more and more serious.If the potential safety hazard is found and eliminated in time before the accident,the accident can be prevented effectively.If effective measures can be taken in time at the beginning of the accident,the accident loss can be reduced.Safety inspection is an effective means to investigate hidden dangers and find early accidents,which can reduce accident losses and even prevent fire accidents.In view of the low efficiency of common inspection methods and the fact that the inspection path cannot be updated independently according to the fire risk level,this paper combines the intelligent dynamic fire risk assessment of the inspection location with the path optimization of the safety inspection robot to carry out the path optimization of the safety inspection robot based on the fire risk level of the chemical park.Firstly,taking the storage area and production area with high risk as an example,the class A hazardous chemicals warehouse and methanol distillation process device were selected as the research objects to establish the evaluation index system,and the evaluation index data were collected to provide data support for the subsequent real-time risk assessment.Then,the dynamic intelligent fire risk assessment model is established by using the improved support vector machine.The index data corresponding to the evaluation index system are used as the input of the evaluation model,and the fire risk level is used as the output to verify the effectiveness of the evaluation model.Finally,the improved discrete electrostatic discharge(IDESDA)algorithm combined with the fire risk level of each inspection position was used to realize the path optimization of the safety inspection robot based on the fire risk level.The following three conclusions are obtained:(1)The inspection location of chemical industry park is divided into reservoir area and production area.Class A dangerous chemical warehouse and methanol distillation purification process device are selected as fire risk assessment objects in two types of inspection locations.The evaluation index of class A hazardous chemicals warehouse and the fire risk assessment system of methanol distillation purification process device were established,which laid the foundation for the establishment of fire risk assessment model.(2)Traditional Support Vector Machine(SVM)has some disadvantages on parameter selection.We use electrostatic discharge algorithm(ESDA)to select SVM parameters,and the effectiveness of the way is verified by public data sets.For the shortcomings of poor global ability of RBF kernel function,NP kernel function is introduced to ensure that the kernel function has both learning ability and generalization ability.In this paper,the results of different models are compared and sample set is A class a library sample set.The results prove that the ESDA-NPSVM dynamic risk intelligent evaluation model is suitable for fire risk assessment.(3)In order to better solve the path optimization of safety inspection robot,discrete electrostatic discharge method(DESDA)is proposed for the first time in this paper.Aiming at the shortcomings of long convergence time and easy to fall into local optimum,IDESDA is proposed.The algorithm performance comparison shows that the proposed IDESDA has better optimization accuracy and stability in solving TSP.Experimental results show that the algorithm is effective in path optimization of inspection robot based on fire risk level.