Research On The Characteristic Portrait Of Truck Drivers Based On Trajectory Data

The national infrastructure investment has been accelerating,and the in-depth implementation of the three major strategies of "One Belt One Road" construction,the coordinated development of Beijing-Tianjin-Hebei and the development of the Yangtze River Economic Belt has brought new development opportunities to the freight market.As the main body of the road freight industry,truck drivers are under great pressure on capital turnover in the freight process,and there is a strong demand for freight financial services such as truck financial leasing,auto insurance installment,and ETC fee advance payment.However,the society has long recognized the integrity of truck drivers on the low side.Therefore,this article focuses on meeting the freight financial needs of truck drivers.By mining its trajectory data,it designs a truck driver’s freight feature portrait mechanism for multi-participants to help improve the level of refined management of truck drivers.In this paper,by sorting out the freight demand of truck drivers,a set of truck driver freight feature portrait mechanism including portrait label design and label extraction method is constructed.Among them,relying on the truck GPS trajectory data of a freight platform,Python programming technology is used to design a set of image tag extraction algorithms such as map matching,POI recognition,trajectory segmentation,and OD recognition,and successfully identify the truck driver’s transportation capacity,transportation times,Transportation efficiency,stable supply of goods,speeding,night driving,fatigue driving,violations,platform click and browse,phone calls,deposit payment and purchase of freight insurance and other portrait label variables.Then,perform factor analysis on the truck driver’s freight feature portrait label extraction data,obtain the common factor score of each truck driver,and use it as the input data of the SOM-K-means two-stage algorithm to obtain the truck driver’s portrait classification result.Each type of truck driver conducts a detailed analysis of freight characteristics to form a word cloud image of each type of truck driver.At the micro-application level,based on the classification results of truck driver portraits,the UBI driving behavior scoring model for truck driver transportation capacity,night driving time,fatigue driving times,speeding times,and violation times is constructed,and the driving behavior score of truck drivers is obtained by using the entropy method;Constructed a UBI rate determination model based on driving behavior scoring and portrait classification with double rate adjustment coefficients.Through calculation example,it is found that under the same basic rate conditions,truck drivers with bad driving habits need to pay up to 15% more For insurance premiums,truck drivers with good driving habits can enjoy a premium discount of up to 25%,and truck drivers with ordinary driving habits can also enjoy a premium discount of up to15% by completing specific freight conditions.Finally,at the macro-strategic level,they are insurance companies and trucks.Dealers,financial institutions,and freight platforms propose personalized product input strategies and customer maintenance strategies for each type of truck driver.