Freight Vehicle Operation Pattern Recognition And Predictability Based On Trajectory Data

Freight transportation is one of the central aspects of logistics and is closely related to all aspects of logistics operations,playing a vital role in linking all parts of social production.Road transport is one of the main modes of cargo transportation,undertaking more than 70% of freight transport tasks.With the development of network communication technologies and freight matching platforms,road truckers have a significant increase in the temporal and spatial scope of information access,and truckers can operate individually without long-term planning and overall freight information.In this context,the operating behavior of a large number of trucks depends entirely on the individual truckers’ operating strategies and preferences.In order to understand the increasing complexity of road freight travel,there is a growing need to more effectively identify freight vehicle travel characteristics and operating pattern characteristics.In view of the above problems,this paper analyzes the freight activity characteristics based on the GPS track data of highway freight vehicles to explore the potential operation mode of trucks.In this paper,the operation characteristics of trucks are divided into spatial operation characteristics and travel operation characteristics,and the pattern analysis is carried out from these two aspects.In order to explore the freight activity pattern,this paper first analyzes the spatial and temporal characteristics of truck activities and finds that freight activities are limited by spatial scope and driver’s ability.The average daily operating mileage of trucks is usually within 200 km,and the activity range is usually within 600 km in diameter,while the number of consecutive working days of trucks is generally limited to five days.Based on the activity characteristics of trucks,the concept of network motif is invoked to identify the spatial operation patterns of trucks,and it is found that up to 90% of the spatial location relationship of truck operation activities can be described by 20 patterns.In addition,this paper establishes ten indicators systems for evaluating truck travel operation characteristics from three dimensions of spatial scope,operation intensity and operation stability,and identifies truck travel operation patterns by FCM clustering algorithm.The analysis results show that there are four different types of stability patterns for trucks,and four different types of travel chain patterns.Predictability is an important information to understand the characteristics of different groups of trucks as well as an important element of travel characteristics analysis.Therefore,this paper conducts statistical analysis on the predictability of different truck groups,and the results show that the truck groups with high stability,low operating intensity and small spatial activity range have the highest location prediction accuracy,and the prediction accuracy of 80%of the vehicles in the group reaches 80%.The results of this paper reveal the spatial movement and activity intensity patterns of trucks,identify different operation patterns of trucks and their predictability differences,help to understand truck movement characteristics,help to evaluate the impact of policies on freight flow and formulate relevant policies to optimize the layout of road freight network,and facilitate freight system optimization,urban planning and movement prediction modeling.