Research On Risk Analysis And Prediction Of Urban Logistics From The Perspective Of Public Safety

Urbanization in China has a population of more than 60%,and the city become the main hub of population,the demand of the urban flow of various elements in the high frequency drive the vigorous development of the city logistics,at the same time,public security event of urban logistics are characterized by diversity,complexity,urgency and uncertainty.Urban logistics poses a serious threat to urban public security,economic and social operation order and people's life and property safety.For realizing the effective governance of the urban logistics public security risks,on the one hand,it is necessary to recognize the risk factors that urban logistics operation has a great impact on public security,and improve the governance scheme.On the other hand,it is necessary to further analyze the realistic dilemma,summarize the occurrence characteristics and rules of the current urban logistics public security events,and then build an efficient and coordinated and refined urban logistics public security risk management scheme.In order to ensure the safety of operation,logistics enterprises generally adopt the traditional means such as strengthening supervision and safety publicity.However,continuous input of manpower and material resources has brought huge economic burden to logistics enterprises.At present,logistics risk research has formed a wealth of risk management theories and methods,but there are still a lot of practical problems to be solved.On the one hand,the existing research on logistics risk management relies too much on empirical judgment or expert knowledge,and the theoretical research on risk analysis methods is still lacking.On the other hand,due to the lack of uniform specifications for accident data records,the degree of data structure is different.How to carry out risk analysis on unstructured and semi-structured data has become one of the hot issues in current research.Based on the above questions,this paper follows the risk management process,mainly including:(1)Statistics and analysis of urban logistics public safety accidents from 2008 to2020 are conducted.Statistics are carried out from the aspects of accident type,accident time,accident source direction,accident consequence,etc.The accident law was analyzed from the following aspects: the year of the accident and the number of deaths,the types of accident and the number of deaths,the types of accident and the level of accident,the types of accident and the direction of risk sources.The causes and internal laws of the accident are deeply analyzed to provide reference for the establishment of the subsequent risk index system.(2)The literature method and the business process analysis method are used to identify the risk factors,and the candidate set of risk factors are established.The NASA-TXL scale method is adopted to get the weights of factors.Under the two attributes of risk probability and risk loss rate,using the technique for order preference by similarity to ideal solution to screen and sort out the importance degree of candidate risk factors,and establish the index system of risk factors.The independence of risk factors is tested,which laid a foundation for the establishment of Bayesian network evaluation model.(3)Put forward the improved Apriori algorithm for fast mining frequent itemsets,design the standardized process of mining association rules of urban logistics risk factors from the perspective of public security.Using the data analysis and processing method,analyze the statistics of 235 accident,received 374 association rules of risk factors.The visualization shows the high support association rules,high confidence association rules and all association rules.Through the visualization results,it can be concluded that the association rules of urban logistics risk factors show significant aggregation characteristics.(4)Based on the interpretative structural model,the initial structure of Bayesian network is established.And through the causal mapping method,the final dynamic risk assessment model of Bayesian network is established.By using Bayesian network inference function to evaluate the results of risk and causes of reasoning,through sensitivity analysis,reveals the "people-car-goods" is the important factors of causing accidents,the results show that the Bayesian network in improving the capacity of city express logistics operations,avoid public security risk is effective.(5)A risk level prediction model optimization method combining particle swarm optimization with generalized regression neural network algorithm is proposed.By comparing the prediction results of the model with those of BP neural network algorithm,the results show that the PSO-GRNN model has the advantages of high prediction accuracy,high stability and small error,and the corresponding risk response mechanism can be adopted in advance to reduce the loss of people's life and property caused by the accident.In this paper,there are 53 graphs,49 tables and 144 references.