A method for extracting OD of truck trips in cities from trajectory data

Abstract

The present invention provides a method for extracting the OD of a truck's trip in a city from trajectory data. The method comprises: determining a speed threshold of a truck in a city, and identifying parking points from the truck's trajectory according to the speed threshold; sorting the truck's parking points in ascending order of parking time, and determining a multi-level time threshold by drawing a Lorenz curve of the parking time; measuring the degree of circuitousness of a truck's single trip path according to the multi-level time threshold, and extracting potential travel OD points from the truck's parking points; eliminating temporary parking points from potential travel OD points, and extracting the truck's real travel OD points. The present invention proposes a method for dynamically identifying truck travel OD points, which can be applied to all cities across the country. The method has strong portability, low complexity, and is easy to implement.

Description

Technical Field

[0001] The present invention relates to the technical field of urban freight management, and in particular to a method for extracting an OD of a truck trip in a city from trajectory data. Background Art

[0002] In the urban freight transportation system, truck transportation plays a key supporting role and is the main form of transportation between large industrial enterprises, logistics warehouses and ports. However, truck transportation can also cause serious social and environmental problems, such as traffic accidents and air pollution, which poses challenges to the sustainable development of cities. In order to eliminate the negative externalities of urban freight and improve the efficiency of the freight system, relevant departments and organizations need to formulate feasible freight policies. Large-scale truck travel OD (Origin to Destination) information is the basic data for formulating these freight policies, providing data support for a deep understanding of the urban freight system.

[0003] Traditionally, truck travel OD information is obtained through traffic surveys. This method is time-consuming and costly, so the amount of data obtained is limited and insufficient for urban freight system analysis. In the era of big data, the development and application of satellite positioning technology makes it possible to obtain large-scale truck trajectory data through vehicle-mounted positioning devices. However, how to extract truck travel OD information from trajectory data is a major problem in practice and has not been well solved. At present, there are few studies on truck travel OD extraction, mainly the following:

[0004] One method for extracting truck travel OD in the prior art is to identify the types of truck stop points with the help of auxiliary information such as truck driver survey data and land use data. The disadvantage of this method is that although it can achieve good recognition results when studying small samples, it is limited by the amount of auxiliary information data, resulting in poor portability of the method.

[0005] Another method for extracting truck travel OD in the prior art is to classify and identify truck parking points based on features such as truck stop time and distance from the city center using SVM (support vector machines). The disadvantage of this method is that the actual truck travel OD information must be known in advance as a test set, and this information is difficult to obtain directly in practice. In addition, this method is highly complex and difficult to apply to large-scale truck data. Summary of the invention

[0006] The embodiment of the present invention provides a method for extracting the travel OD of a truck in a city from trajectory data, so as to realize effective dynamic identification of the travel OD of the truck.

[0007] In order to achieve the above object, the present invention adopts the following technical scheme.

[0008] The method of extracting the OD of truck travel in the city from trajectory data includes:

[0009] determining a speed threshold of a truck in a city, and identifying a stopping point from the truck trajectory according to the speed threshold;

[0010] The parking spots of trucks are sorted in ascending order of parking time, and a multi-level time threshold is determined by drawing a Lorenz curve of parking time;

[0011] Measuring the degree of circuitousness of a single trip path of a truck according to the multi-level time threshold, and extracting potential trip OD points from the truck parking points;

[0012] Temporary parking points among potential travel OD points are eliminated to extract the real travel OD points of the truck.

[0013] Preferably, the method of sorting the parking spots of trucks in ascending order of parking time and determining the multi-level time threshold by drawing a Lorenz curve of the parking time includes:

[0014] The average speed of two consecutive GPS points in the GPS tracks of all trucks in the city is calculated respectively to obtain the distribution of the average speed of all trucks. The distribution of the average speed of the trucks is a mixed distribution of the data drift speed and the normal driving speed of the trucks. The probability distribution of the mixed distribution is as follows:

[0015]

[0016] Among them, Lognorm(x; μ1, σ1) is the probability density function of the lognormal distribution, which is used to fit the speed data of data drift; Norm(x; μ2, σ2) is the probability density function of the normal distribution, which is used to fit the speed data of normal driving of the truck. The maximum likelihood method is used to estimate the parameters of the mixed distribution ω1, ω2, μ1, μ2, σ1 and σ2;

[0017] The speed value corresponding to the lowest point between the two peaks of the mixed distribution is determined as the speed threshold. If the speed of the truck at a certain location is less than the speed threshold, the location is identified as a parking point, and the geographical coordinates of a truck parking point are represented by the average latitude and longitude of all GPS points at the parking location.

[0018] Preferably, the method of sorting the parking spots of trucks in ascending order of parking time and determining the multi-level time threshold by drawing a Lorenz curve of the parking time includes:

[0019] All parking spots for trucks are sorted in ascending order of parking time, and a Lorenz curve of the parking time is drawn. The intersection of the tangent at the rightmost endpoint of the Lorenz curve and the x-axis is calculated, and the parking time of the truck corresponding to the intersection is determined as a time threshold. For parking spots whose parking time is less than the time threshold, the Lorenz curve is redrawn and the intersection of the tangent and the x-axis is calculated to determine the time threshold of the next level. The above processing is continuously iterated until the Lorenz curve is a straight line, and a multi-level time threshold is obtained.

[0020] Preferably, the method of measuring the degree of circuitousness of a single trip path of a truck according to the multi-level time threshold and extracting potential trip OD points from truck parking points includes:

[0021] Calculate the first K shortest paths from the departure point to the destination of each trip on the road network, find the nth shortest path closest to the actual travel path of the truck, measure the degree of circuitousness of the truck's single trip path according to the nth shortest path, and dynamically adjust the time threshold based on the circuitousness of the truck's single trip path as a benchmark;

[0022] A maximum time threshold is selected, and parking points with parking time greater than the maximum time threshold are identified as potential travel OD points. The original trajectory is divided into multiple sub-trajectories according to the potential travel OD points, named as first-level sub-trajectories, and the nth shortest path between the departure point and the destination of each first-level sub-trajectory is calculated. If the length of a first-level sub-trajectory is greater than the calculated nth shortest path, it indicates that this first-level sub-trajectory is more circuitous than the single travel path and contains travel OD points with shorter parking time; then, the next level time threshold is selected to identify the travel OD points contained in the circuitous first-level sub-trajectory, and the first-level sub-trajectory is divided into multiple second-level sub-trajectories in one step. The above processing process is continuously iterated until the sub-trajectory cannot be divided, indicating that all potential travel endpoints are extracted.

[0023] Preferably, the step of eliminating temporary parking points from potential travel OD points and extracting the real travel OD points of the truck includes:

[0024] Use road network data to determine whether trucks stay on the road for a long time due to traffic congestion. If a potential travel OD point is located on the road, it indicates that the potential travel OD point is a temporary parking point and needs to be eliminated. Use the point of interest data related to freight in the city to determine whether trucks are loading and unloading in freight companies. If the identified potential travel OD point is not located in the freight company, it indicates that the potential travel OD point is a temporary parking point and needs to be eliminated; otherwise, the potential travel OD point is a real travel OD point.

[0025] After extracting the actual travel OD points of the truck from the trajectory data, the travel path of the truck is extracted in combination with the GPS data, and the travel-related indicators are calculated.

[0026] It can be seen from the technical solutions provided by the above embodiments of the present invention that the present invention proposes a method for dynamically identifying OD points of truck travel, which can be applied to all cities across the country. The method has strong portability, low complexity and is easy to implement.

[0027] Additional aspects and advantages of the present invention will be given in part in the following description, which will become obvious from the following description, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0029] Figure 1 A processing flow chart of a method for extracting the OD of a truck trip in a city from trajectory data provided by an embodiment of the present invention;

[0030] Figure 2 A schematic diagram of determining a speed threshold provided by an embodiment of the present invention;

[0031] Figure 3 A schematic diagram of truck parking spot identification provided by an embodiment of the present invention;

[0032] Figure 4 A schematic diagram of determining a multi-level time threshold provided by an embodiment of the present invention;

[0033] Figure 5 A schematic diagram for determining the nth shortest path provided by an embodiment of the present invention;

[0034] Figure 6 A schematic diagram of extracting potential travel endpoints from a truck parking spot provided by an embodiment of the present invention;

[0035] Figure 7 A schematic diagram of eliminating temporary parking spots provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0037] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or coupling. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0038] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0039] To facilitate understanding of the embodiments of the present invention, several specific embodiments will be further explained below with reference to the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0040] The embodiment of the present invention extracts the travel OD of trucks in a city from large-scale trajectory data, mainly including: determining a speed threshold suitable for a certain city according to the distribution of truck speeds, and identifying parking points from the original trajectory of the truck; determining multiple time thresholds according to the distribution of the truck's stay time at the parking point, and measuring the degree of circuitousness of a single travel path of the truck; based on the degree of circuitousness of the single travel path, dynamically selecting a time threshold of a suitable level to identify potential travel OD points from all parking points; using urban interest point data and road network data to eliminate temporary parking points from potential travel OD points, and then identifying all real travel OD points.

[0041] The processing flow chart of a method for extracting the OD of a truck trip in a city from trajectory data provided by an embodiment of the present invention is as follows: Figure 1As shown, the processing steps include the following:

[0042] Step S10: Determine a speed threshold of a truck in the city and identify a parking point from the truck trajectory.

[0043] Figure 2 A schematic diagram of determining a speed threshold provided by an embodiment of the present invention, for each truck's GPS (Global Positioning System) trajectory, the average speed of two consecutive GPS points in the GPS trajectory is calculated to obtain the distribution of the average speeds of all trucks in a city. The distribution of the average speed of the truck is a mixed distribution of the data drift speed and the normal driving speed of the truck, and the average speed distribution is obtained by fitting the mixed distribution using formula (1).

[0044] The data drift speed refers to the possible deviation in positioning of the GPS device, which may cause the GPS point of the stationary vehicle to drift, which is reflected in the data that the average speed of two consecutive GPS points is not 0. Therefore, the present invention sets a speed threshold to distinguish the data drift speed from the normal driving speed of the truck, and then identify the real parking point of the truck. If the average speed of two consecutive GPS points is less than the set speed threshold, the truck is stationary and a parking point can be identified.

[0045] The probability distribution of the mixed distribution is as follows:

[0046]

[0047] Among them, Lognorm(x; μ1, σ1) is the probability density function of the lognormal distribution, which is mainly used to fit the speed data of data drift; Norm(x; μ2, σ2) is the probability density function of the normal distribution, which is mainly used to fit the speed data of normal truck driving. The maximum likelihood method is used to estimate the parameters of the mixed distribution ω1, ω2, μ1, μ2, σ1 and σ2.

[0048] Figure 3 A schematic diagram of truck parking point recognition provided by an embodiment of the present invention is provided, where the saddle point (i.e. Figure 3 The speed value corresponding to the lowest point between the two peaks is determined as the speed threshold. If the speed of the truck at a certain location is less than the determined speed threshold, the location is identified as a parking point. The geographic coordinates of a truck parking point are represented by the average latitude and longitude of all GPS points at the parking location.

[0049] Step S20: sorting the parking spots of trucks in ascending order of parking time, and determining a multi-level time threshold by drawing a Lorenz curve of the parking time.

[0050] Figure 4A schematic diagram of determining a multi-level time threshold provided by an embodiment of the present invention, the method uses a non-parametric iterative method to determine the multi-level time threshold. First, the truck parking spots are sorted in ascending order of parking time, and the Lorenz curve of parking time is drawn; then, the intersection of the tangent line at the rightmost end point of the Lorenz curve and the x-axis is calculated, and the truck parking time corresponding to this intersection is determined as a time threshold, such as Figure 4 Then, for parking points whose parking time is less than this time threshold, the Lorenz curve is redrawn and the intersection of the tangent line and the x-axis is calculated to determine the time threshold of the next level, as shown in FIG. Figure 4 b; this process is iterated continuously until the Lorenz curve is a straight line, as shown in Figure 4 After the iteration process is completed, a multi-level time threshold can be obtained, which is used to identify potential travel OD points in step S30.

[0051] In the process of identifying potential travel OD points, the time threshold needs to be dynamically adjusted. The degree of tortuosity of a truck's single trip path is the benchmark for the dynamic adjustment of the time threshold. In order to measure the degree of tortuosity of a truck's single trip path, first, satellite images and city POIs are used to extract some truck OD points from the original GPS data as a sample data set. The OD points in the sample data set of the truck's actual travel are manually extracted from the GPS data using satellite images and city POIs, and used as sample data for determining the algorithm parameters. These manually extracted OD points are a small part of the real travel OD points extracted in step S40 or can be understood as a subset. Places related to truck activities usually have significant architectural features that can be identified from satellite images, so a small part of truck OD points can be extracted from GPS data using manual methods. However, the geographical features of most truck OD points are not obvious, and the manual method is time-consuming and therefore not suitable for large-scale GPS data. This is also the actual background considered by the present invention.

[0052] Then, the first K shortest paths from the departure point to the destination of each trip are calculated on the road network; then, the nth (n≤K) shortest path closest to the actual travel path of the truck is found to measure the degree of circuitousness of the truck's single travel path. Figure 5 A schematic diagram of determining the nth shortest path provided by an embodiment of the present invention.

[0053] Step S30: extracting potential travel OD points from the truck parking spots according to the multi-level time thresholds and the nth shortest path.

[0054] Figure 6A schematic diagram of extracting potential travel endpoints from a truck parking point provided by an embodiment of the present invention. The present invention uses a time threshold dynamic adjustment method to extract potential travel OD points from the parking points identified in step S10. First, the maximum time threshold determined in step S20 is selected to identify parking points with a parking time greater than this time threshold as potential travel OD points, and the original trajectory is divided into multiple sub-trajectories according to these travel endpoints, named as first-level sub-trajectories, such as Figure 6 As shown in the second layer; then, the nth shortest path between the departure point and the destination of each first-level sub-trajectory is calculated. If the length of a first-level sub-trajectory is greater than the calculated nth shortest path, it means that this first-level sub-trajectory is more circuitous than a single trip path, that is, it may contain a travel OD point with a shorter parking time; then, the next level time threshold is selected to identify the travel OD points contained in the circuitous first-level sub-trajectory, and the first-level sub-trajectory is divided into multiple second-level sub-trajectories in one step, as shown in Figure 6 As shown in the third layer; the above process is iterated continuously until the sub-trajectory cannot be divided, indicating that all potential travel endpoints are extracted, such as Figure 6 The fourth floor is shown.

[0055] Step S40: Eliminate temporary parking points from potential travel OD points and extract real travel OD points.

[0056] Figure 7 A schematic diagram of eliminating temporary parking points provided in an embodiment of the present invention. The potential travel endpoints identified in step S30 may include long-term temporary parking points, such as long-term traffic congestion points and driver rest points. These temporary parking points need to be eliminated to improve the accuracy of the method. First, use the road network data to determine whether the truck stays on the road for a long time due to traffic congestion. If a potential travel OD point is located on the road, it indicates that the potential travel OD point is a temporary parking point and needs to be eliminated. Figure 7 Then, the POI data related to freight in the city is used to determine whether the truck is loading and unloading in the freight enterprise. If the identified potential travel OD point is not located in the freight enterprise, it indicates that the potential travel OD point is a temporary parking point and needs to be eliminated, such as Figure 7 b; otherwise, the potential travel OD point is the real travel OD point, such as Figure 7 As shown in c.

[0057] After extracting the actual travel OD points of the truck from the trajectory data, the travel path of the truck can be extracted in combination with the GPS data, and travel-related indicators such as travel distance, travel time, loading and unloading time, etc. can be calculated.

[0058] Table 1 shows the comparison between the method of the present invention and the existing methods. Method accuracy M acc The calculation is as follows:

[0059] M acc =NA / (NA+NM+NU) (2)

[0060] NA represents the number of accurately identified travel OD points, NM represents the number of incorrectly identified travel OD points, and NU represents the number of unidentified travel OD points. The results show that the accuracy of the method proposed in the present invention is significantly higher than that of the existing technology.

[0061] In summary, the embodiment of the present invention provides a data-driven speed threshold determination method, which is more accurate, objective, and universal, and greatly reduces the possibility of parking spots being misidentified.

[0062] The present invention proposes a data-driven multi-level time threshold determination method, which is accurate, objective and universal, avoids temporary parking spots from being mistakenly identified as parking spots, and improves the accuracy of the method.

[0063] The present invention proposes a method for processing temporary parking points using widely available urban point of interest data and road network data, thereby improving the accuracy of the method.

[0064] The present invention proposes a dynamic identification method for truck travel OD points, which can be applied to all cities across the country. The method has strong transferability, low complexity and is easy to implement.

[0065] Table 1 Comparison of the accuracy of the method of the present invention and the prior art

[0066]

[0067] Those skilled in the art can understand that the accompanying drawings are only schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0068] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0069] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0070] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for extracting the OD of truck travel in a city from trajectory data, characterized in that: include: determining a speed threshold of a truck in a city, and identifying a stopping point from the truck trajectory according to the speed threshold; The parking spots of trucks are sorted in ascending order of parking time, and a multi-level time threshold is determined by drawing a Lorenz curve of parking time; Measuring the degree of circuitousness of a single trip path of a truck according to the multi-level time threshold, and extracting potential trip OD points from the truck parking points; Eliminate temporary parking points from potential travel OD points and extract the real travel OD points of the truck; The method of sorting the parking spots of trucks in ascending order of parking time and determining the multi-level time threshold by drawing the Lorenz curve of parking time includes: The average speed of two consecutive GPS points in the GPS tracks of all trucks in the city is calculated respectively to obtain the distribution of the average speed of all trucks. The distribution of the average speed of the trucks is a mixed distribution of the data drift speed and the normal driving speed of the trucks. The probability distribution of the mixed distribution is as follows: Among them, Lognorm(x; μ1, σ1) is the probability density function of the lognormal distribution, which is used to fit the speed data of data drift; Norm(x; μ2, σ2) is the probability density function of the normal distribution, which is used to fit the speed data of normal driving of the truck. The maximum likelihood method is used to estimate the parameters of the mixed distribution ω1, ω2, μ1, μ2, σ1 and σ2; The speed value corresponding to the lowest point between the two peaks of the mixed distribution is determined as a speed threshold. If the speed of the truck at a certain location is less than the speed threshold, the location is identified as a parking point. The geographical coordinates of a truck parking point are represented by the average latitude and longitude of all GPS points at the parking location. The method of sorting the parking spots of trucks in ascending order of parking time and determining the multi-level time threshold by drawing the Lorenz curve of parking time includes: All parking spots of trucks are sorted in ascending order of parking time, and a Lorenz curve of parking time is drawn. The intersection of the tangent line at the rightmost end point of the Lorenz curve and the x-axis is calculated, and the parking time of the truck corresponding to the intersection is determined as a time threshold. For parking spots whose parking time is less than the time threshold, the Lorenz curve is redrawn and the intersection of the tangent line and the x-axis is calculated to determine the time threshold of the next level. The above processing is continuously iterated until the Lorenz curve is a straight line, thereby obtaining a multi-level time threshold. The method of measuring the degree of circuitousness of a single trip path of a truck according to the multi-level time threshold and extracting potential trip OD points from truck parking points includes: Calculate the first K shortest paths from the departure point to the destination of each trip on the road network, find the nth shortest path closest to the actual travel path of the truck, measure the degree of circuitousness of the truck's single trip path according to the nth shortest path, and dynamically adjust the time threshold based on the circuitousness of the truck's single trip path as a benchmark; Select a maximum time threshold, identify parking points with parking time greater than the maximum time threshold as potential travel OD points, divide the original trajectory into multiple sub-trajectories according to the potential travel OD points, and name them as first-level sub-trajectories. Calculate the nth shortest path between the departure point and the destination of each first-level sub-trajectory. If the length of a first-level sub-trajectory is greater than the calculated nth shortest path, it indicates that this first-level sub-trajectory is more circuitous than the single travel path and contains travel OD points with shorter parking time. Then, select the next level time threshold to identify the travel OD points contained in the circuitous first-level sub-trajectory, and divide the first-level sub-trajectory into multiple second-level sub-trajectories in one step. Continuously iterate the above process until the sub-trajectory cannot be divided, indicating that all potential travel endpoints are extracted. The method of eliminating temporary parking points from potential travel OD points and extracting the real travel OD points of the truck includes: Use road network data to determine whether trucks stay on the road for a long time due to traffic congestion. If a potential travel OD point is located on the road, it indicates that the potential travel OD point is a temporary parking point and needs to be eliminated. Use the point of interest data related to freight in the city to determine whether trucks are loading and unloading in freight companies. If the identified potential travel OD point is not located in the freight company, it indicates that the potential travel OD point is a temporary parking point and needs to be eliminated; otherwise, the potential travel OD point is a real travel OD point. After extracting the actual travel OD points of the truck from the trajectory data, the travel path of the truck is extracted in combination with the GPS data, and the travel-related indicators are calculated.

Images