Research On The Application Of Evaluation Method Of Fleet Operation Safety Level Based On Internet Of Vehicles Data

With the progress of people’s life,the car ownership is gradually increasing.However,in recent years,the number of traffic accidents has also increased significantly.According to the data released by the Ministry of transport,more than 70%of traffic accidents are caused by drivers,and the standard driving behavior directly affects the road traffic safety.At the same time,as a professional transportation unit of automobile,the safety of transportation fleet should be paid more attention because of its long driving distance and important transportation of goods.Most of the fleet management still has the problems of extensive and backward management theories and measures,which can not meet the needs of refinement and intelligence in the process of vehicle safety management.Therefore,it is necessary to study the safety management of transport fleet.At present,the research content of safety management evaluation for motorcades at home and abroad is relatively scattered.In the process of specific implementation of safety management,some motorcades often only focus on driving,emphasize oral deployment,unable to grasp the actual driving state of drivers in the motorcade,unable to guide them correctly,unable to standardize driving,and may have potential safety hazards.Therefore,vehicle management departments and major insurance companies began to study in order to reduce the risk of the fleet and improve the safety management level of the fleet.In this dissertation,the combined evaluation method is used to evaluate the driving risk of motorcade,and the safety management of motorcade is carried out according to the results of the model.First of all,the data of driving behavior is preprocessed,and the original data of modeling is formed through index screening,data cleaning and other operations,and the comprehensive evaluation method is selected for modeling.First,determine the indicators of the criteria layer,then use entropy method to calculate the weight of these indicators as the scheme layer.Then determine the weight of each index in the criteria layer,which uses the analytic hierarchy process.Then,the entropy method is used to optimize the judgment matrix of AHP,and then the driving risk score of each team can be calculated.Finally,the association rule method is used to verify the results of the model,the data of important indicators are discretized and classified,and the association degree between important indicators and risk score is verified by association rules.The results show that the model has a high correlation with the results of fleet risk score,and the model is considered to be effective.Finally,the safety management of motorcade is carried out.Firstly,the safety management of all motorcades is carried out.According to the model results,all motorcades are divided into three categories:extremely irregular motorcade,irregular motorcade and relatively standard motorcade.Different safety management is carried out for motorcades with different driving styles to realize the diversity and uniqueness of safety management.Next,the internal safety management of the team is carried out.The SWOT analysis method is used to analyze the advantages of the internal safety management of the enterprise to the team itself,and then to analyze the disadvantages of the team,then to analyze the opportunities and threats faced by the team,and to put forward more detailed suggestions for the safety management of the team.In this dissertation,the real-time alarm data of vehicle driving is used to establish a fleet risk assessment model combining with various methods.The traditional static assessment is transformed into dynamic assessment based on risk factors,and the traditional fleet risk post analysis is transformed into pre analysis and prevention.This model is conducive to the timely monitoring of vehicles by the fleet safety management system,the differentiation of insurance pricing for different fleets based on the results of risk assessment by insurance companies,and the standardization of drivers’ own driving behavior,which provides a new way for the development of public transport safety and vehicle insurance industry.There are still some deficiencies in this dissertation,mainly in the following two aspects:first,it is impossible to get the specific accident data of each fleet,so the model results only apply to the horizontal comparison of driving risks between fleets.Second,from the data set point of view,the indicators are still not comprehensive.The objective factors such as road conditions,weather and drivers are not taken into account in the assessment of driving risks;and the indicators related to the business operation are not taken into account in the assessment of the safety management level of motorcade.