Port hinterland division method and device based on heavy-haul train trajectory, equipment and medium

Abstract

The application discloses a port hinterland division method and device based on heavy truck trajectories, equipment and medium, relates to the technical field of port planning, and comprises the following steps: by constructing the functional mapping relationship between the berth space vector data and the loading and unloading class label, cleaning and cutting the heavy truck trajectory, and adopting the berth interest surface space judgment and the loading and unloading residence time verification based on the cargo class attribute dual mechanism, the effective collection and distribution trajectory is accurately extracted, the core cargo source aggregation area is obtained by clustering the land side parking point density, and the multi-level hinterland is divided according to the freight transport contact frequency contribution degree. The application effectively eliminates the noise trajectories such as transit, maintenance and parking, improves the recognition accuracy of hinterland division, and provides reliable support for port resource allocation and competition strategy formulation.

Description

Technical Field

[0001] This invention relates to the field of port planning technology, and in particular to a method, apparatus, equipment and medium for port hinterland delineation based on heavy-duty truck trajectories. Background Technology

[0002] Currently, there are two main approaches to delineating port hinterlands. The first is a theoretical deduction method based on macro-administrative divisions or spatial economic gravity models (such as gravity models and breakpoint models). This method calculates the port's attractiveness to surrounding areas using static statistical data such as regional GDP, population size, and import / export trade volume, and then delineates the hinterland boundaries accordingly. The second approach, which has emerged in recent years with the widespread adoption of positioning technology, utilizes trajectory data analysis. This method collects truck positioning trajectories to create macro-level origin-destination heatmaps or conducts gridded node intensity statistics, thereby reflecting the strength of logistical connections between the port and the landside region.

[0003] However, all of the above approaches have significant shortcomings in practical application. Regarding the first approach, its essence is based on theoretical deduction of potential demand. The static statistical data it relies on cannot reflect the dynamic flow and seasonal fluctuations of real logistics. Furthermore, it treats the entire port as an indiscriminate node, completely ignoring the heterogeneity between different berths and different types of cargo within the port. Therefore, the resulting classification lacks practical operational value in guiding specific berth operations. Regarding the second approach, existing trajectory analysis largely remains at the level of simple heat map drawing and intensity statistics, lacking in-depth matching and verification based on complex spatiotemporal logic. On the one hand, the actions of trucks passing through, repairing, dining, and resting around the port area generate a large number of noise trajectories unrelated to actual loading and unloading at the port, which cannot be eliminated simply by spatial overlap. On the other hand, the loading and unloading processes for different types of cargo vary greatly. Container loading and unloading are fast, while bulk cargo involves cumbersome processes such as weighing, queuing, and leveling. Using a uniform standard to judge loading and unloading behavior will inevitably lead to a large number of misjudgments, seriously interfering with the accuracy of hinterland identification.

[0004] Therefore, how to accurately identify the collection and distribution vehicles that are actually loading and unloading specific types of goods for a specific berth in the trajectory of heavy-duty trucks, which is mixed with a lot of noise, and how to delineate the actual hinterland of the berth level, has become an urgent problem to be solved. Summary of the Invention

[0005] The main purpose of this application is to provide a method, device, equipment and medium for port hinterland division based on heavy-duty truck trajectory, aiming to solve the technical problem of how to achieve accurate quantitative division of port hinterland in two dimensions: berths and subdivided cargo types, by relying on heavy-duty truck operation trajectory data.

[0006] To achieve the above objectives, this application proposes a method for port hinterland delineation based on heavy-duty truck trajectories, including: Acquire berth spatial vector data and cargo loading / unloading category labels for each berth within the target area, and construct the berth function mapping relationship between the berth spatial vector data and the cargo loading / unloading category labels to obtain a berth feature library; Obtain the trajectory dataset of trucks whose load reaches a preset load threshold within the target area, and perform anomaly removal and missing data repair on the trajectory dataset to obtain a cleaned trajectory dataset. Stop points are identified for the continuous trajectories of each truck in the cleaning trajectory dataset, and the continuous trajectories are divided into travel chains based on the identified stop points to obtain a travel chain set; A berth interest surface buffer is constructed based on the berth feature library, and spatial determination is performed on the two ends of each travel chain in the travel chain set and the berth interest surface buffer to obtain the target berth point that falls into the berth interest surface buffer. The actual dwell time is calculated based on the travel chain corresponding to the target stop. When the actual dwell time exceeds the preset loading and unloading time threshold, the travel chain corresponding to the target stop is marked as a valid collection and distribution trajectory with the loading and unloading cargo type label. Extract the landside stops located on the landside from each of the effective collection and distribution trajectories, and perform density clustering on the landside stops with the same loading and unloading cargo type labels. Generate core cargo source agglomeration areas based on the clusters obtained from the clustering. The contribution of freight transport frequency from each core cargo gathering area to the corresponding berth is calculated based on the effective collection and distribution trajectories. The core cargo gathering areas are then divided according to the cumulative proportion of the contribution of freight transport frequency, generating berth-level hinterland division results.

[0007] Furthermore, to achieve the above objectives, this application also proposes a port hinterland delineation device based on heavy-duty truck trajectories, the port hinterland delineation device based on heavy-duty truck trajectories comprising: The berth feature library construction module is used to obtain the berth spatial vector data and loading / unloading category labels of each berth in the target area, and to construct the berth function mapping relationship between the berth spatial vector data and the loading / unloading category labels to obtain the berth feature library; The trajectory cleaning module is used to acquire the trajectory dataset of trucks whose load reaches a preset load threshold within the target area, and to remove anomalies and repair missing data in the trajectory dataset to obtain a cleaned trajectory dataset. The trip chain segmentation module is used to identify the stop points of the continuous trajectories of each truck in the cleaning trajectory dataset, and to segment the continuous trajectories into trip chains based on the identified stop points to obtain a trip chain set. The spatial determination module is used to construct a berth interest surface buffer based on the berth feature library, and to perform spatial determination on the two ends of each travel chain in the travel chain set and the berth interest surface buffer to obtain the target berth that falls into the berth interest surface buffer. The duration verification module is used to calculate the actual dwell time based on the travel chain corresponding to the target stop. When the actual dwell time exceeds the preset loading and unloading time threshold, the travel chain corresponding to the target stop is marked as a valid collection and distribution trajectory with the loading and unloading cargo type label. The clustering module is used to extract the landside stops located on the landside in each of the effective collection and distribution trajectories, and to perform density clustering on the landside stops with the same loading and unloading cargo category labels, and to generate core cargo source agglomeration areas based on the clusters obtained from the clustering. The hinterland division module is used to calculate the freight connection frequency contribution of each core cargo agglomeration area to the corresponding berth based on each effective collection and distribution trajectory, and to divide each core cargo agglomeration area according to the cumulative proportion of the freight connection frequency contribution, thereby generating berth-level hinterland division results.

[0008] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the port hinterland division method based on heavy-duty truck trajectories described above.

[0009] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the port hinterland division method based on heavy-duty truck trajectories described above.

[0010] This application constructs a functional mapping relationship between berth spatial vector data and loading / unloading cargo category labels to clean and segment the trajectories of heavy-duty trucks. It employs a dual mechanism of berth interest surface spatial determination and loading / unloading dwell time verification based on cargo category attributes to accurately extract effective collection and distribution trajectories. Furthermore, it uses landside berth density clustering to obtain core cargo source agglomeration areas and divides the hinterland into multi-level levels based on the contribution of freight connection frequency. This application effectively eliminates noisy trajectories such as transit, maintenance, and parking routes, improving the accuracy of hinterland division and providing reliable support for port resource allocation and competitive strategy formulation. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating the first embodiment of the port hinterland division method based on heavy-duty truck trajectories in this application. Figure 2 This is a flowchart illustrating the second embodiment of the port hinterland division method based on heavy-duty truck trajectories in this application. Figure 3 This is a schematic diagram of the modular structure of the port hinterland division device based on the trajectory of heavy-duty trucks in this application. Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the port hinterland division method based on heavy-duty truck trajectories in the embodiments of this application.

[0013] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0014] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0015] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0016] Existing methods for port hinterland delineation fall into two categories: macroscopic gravity models and simplified truck trajectory analysis. The former relies on static statistical data for extrapolation, and its analysis granularity only covers the entire port, failing to distinguish the radiation differences between different berths and specific cargo types. The latter relies solely on spatial distance to filter vehicle trajectories, lacking a time-based verification mechanism for loading and unloading processes, making it difficult to eliminate invalid trajectories such as transit and rest stops, resulting in significant noise interference in cargo source identification. Furthermore, neither method can quantify the hierarchical structure of the hinterland and the competitive relationship for cargo sources between ports, making it difficult to support refined business needs such as berth construction, precise investment attraction, and coordinated port cluster management.

[0017] Based on the above, this application also provides a method for port hinterland delineation based on heavy-duty truck trajectories, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the port hinterland division method based on heavy-duty truck trajectories in this application.

[0018] In this embodiment, the port hinterland delineation method based on heavy-duty truck trajectories includes steps S10 to S70: Step S10: Obtain the berth space vector data and loading / unloading category labels for each berth within the target area, and construct the berth function mapping relationship between the berth space vector data and the loading / unloading category labels to obtain the berth feature library.

[0019] It should be noted that berth spatial vector data refers to the planar vector data describing the geographic boundaries of the berth's operating face and the entrance / exit of the rear storage yard, which can be represented using polygon coordinate strings commonly used in geographic information systems; cargo category labels can be broad categories such as liquid bulk cargo, dry bulk cargo, general cargo, and containers, or can be refined to specific cargo types such as iron ore, coal, and grain; berth function mapping relationship refers to the one-to-one correspondence between berth spatial vector data and cargo category labels; berth feature database refers to a structured database that stores the spatial vector data of each berth, cargo category labels, and berth function mapping relationship.

[0020] Further, step S10 includes: First, acquiring the polygonal boundary data of the working face and the polygonal boundary data of the rear yard entrance / exit for each berth within the target area, and merging the polygonal boundary data of the working face and the polygonal boundary data of the rear yard entrance / exit into berth spatial vector data. It should be understood that the polygonal boundary data of the working face refers to the area vector data describing the geographic boundary of the loading and unloading operation area in front of the berth, and the polygonal boundary data of the rear yard entrance / exit refers to the area vector data describing the geographic boundary of the access area for trucks entering and exiting the rear yard. Both can be represented using polygonal coordinate strings in a geographic information system. Specifically, the polygonal boundary data of the working face for each berth within the target area is acquired from the port geographic information management platform or waterway surveying results. The working face boundaries cover the shoreline area actually served by loading and unloading equipment such as quay cranes and gantry cranes. Then, the polygonal boundary data of the rear yard entrance / exit for each berth is acquired from the same data source. The entrance / exit boundaries cover the gate and access area for trucks entering and exiting the yard. Since trucks may park at the berth working face or enter and exit through the yard entrance after entering the port area, using only the working face boundary would miss trucks entering and exiting from the yard side. Therefore, the polygonal boundary data of the working face and the polygonal boundary data of the yard entrance are associated and merged according to the berth number. The coordinate strings of the two polygons are then spliced ​​together to form the berth spatial vector data, so that the berth spatial vector data can completely cover all areas where trucks may park in the port area.

[0021] Secondly, acquire vessel reporting data and port operation ledgers for the target area, and extract cargo loading / unloading tags for each berth from these data. It's important to understand that vessel reporting data refers to the data submitted by vessels to the shipping management department before berthing, which may include information such as vessel name, berth, cargo type, and loading / unloading quantity. Port operation ledgers refer to the actual berth operation information recorded by the port operation system, which may include berth number, cargo type, operation volume, and operation time. Specifically, retrieve vessel reporting data for the target area within a historical time period from the shipping management department's vessel reporting system, filtering out the berth and cargo type fields. Simultaneously, retrieve the port operation ledgers for the same period from the port operation system, filtering out the berth number and cargo type fields. Since a single source may result in omissions or incomplete records, the ship port declaration data is compared and linked with the port production operation ledger by berth number. The cargo types that have appeared at the same berth are merged to generate loading and unloading cargo category labels for each berth. For example, if a berth has been engaged in iron ore unloading for a long time, then the loading and unloading cargo category label for that berth is iron ore under the dry bulk cargo category.

[0022] Then, a one-to-one mapping relationship is established between the spatial vector data of each berth and the cargo category labels, thus obtaining the berth function mapping relationship. Specifically, using the berth number as the association key, the spatial vector data of each berth is mapped one-to-one with the extracted cargo category labels, creating a mapping entry for each berth. All mapping entries constitute the berth function mapping relationship, thereby giving each berth area on the map the corresponding cargo category attribute.

[0023] Finally, the spatial vector data of each berth, the cargo type labels, and the berth function mapping relationships are stored in a pre-set database to obtain the berth feature library. Specifically, the pre-set database refers to a pre-established database for storing berth-related information; it can be a relational database or a geographic database that supports spatial data storage. It's important to understand that, specifically, the spatial vector data of each berth, the cargo type labels, and the berth function mapping relationships are organized into a data table according to a unified berth number and written into the pre-set database to obtain the berth feature library. Subsequent steps can directly retrieve the spatial boundaries and cargo type attributes of any berth from the berth feature library using the berth number.

[0024] Step S20: Obtain the trajectory dataset of trucks with loads reaching the preset load threshold within the target area, and perform anomaly removal and missing data repair on the trajectory dataset to obtain a cleaned trajectory dataset.

[0025] It should be noted that the preset load threshold refers to the lower limit of the approved load for screening target trucks. In this embodiment, 12 tons is chosen because heavy-duty general cargo vehicles and semi-trailer tractors with a total mass of 12 tons or more are the main vehicle types for long-distance collection and distribution at the port, while light trucks mainly undertake short-distance delivery within the city and have a weaker connection with the port's hinterland transportation. The road freight vehicle monitoring platform covers the operation data of heavy-duty general cargo vehicles and semi-trailer tractors with a total mass of 12 tons or more, and collects the positioning trajectory records reported by vehicle positioning terminals at fixed intervals. Each positioning trajectory record includes vehicle identification, recording time, longitude, latitude, instantaneous speed, and azimuth angle. The cleaned trajectory dataset refers to the data set obtained after removing abnormal trajectory points and repairing missing trajectory segments from the trajectory dataset.

[0026] Further, step S20 includes: First, obtaining the location trajectory records and vehicle certified loads of each truck within the target area from the road freight vehicle monitoring platform, and filtering out the location trajectory records corresponding to trucks whose certified loads are less than a preset load threshold, to obtain a trajectory dataset. The location trajectory records include the recording time, longitude, latitude, instantaneous speed, and azimuth. It should be understood that the road freight vehicle monitoring platform refers to a dynamic monitoring system for freight vehicles established by the transportation authority. This can be a national public monitoring and service platform for road freight vehicles or a provincial transportation and logistics information platform. The road freight vehicle monitoring platform collects operational data reported by freight vehicle positioning terminals at fixed intervals.

[0027] Secondly, the estimated speed and estimated acceleration are calculated based on the longitude, latitude, and recording time of two adjacent trajectory points in the trajectory dataset. Trajectory points with estimated speeds exceeding a preset speed limit threshold or estimated accelerations exceeding a preset acceleration range are removed from the trajectory dataset, resulting in the first trajectory dataset. It's important to understand that estimated speed refers to the driving speed calculated from the spatial displacement and time difference between two adjacent trajectory points, and estimated acceleration refers to the acceleration calculated from the speed change and time difference between adjacent trajectory points. The preset speed limit threshold is the upper speed limit used to determine drift; in this example, it is set to 120 km / h, based on the fact that the legal maximum speed limit on domestic highways is 120 km / h, and data exceeding this value is physically impossible. The preset acceleration range refers to the reasonable range of acceleration and deceleration capabilities of the truck, which is determined with reference to the braking and power performance of heavy-duty trucks. Specifically, for each truck in the trajectory dataset, the sequence of trajectory points is sorted by recording time. Two adjacent trajectory points are taken, and the spatial displacement is calculated using the spherical distance formula based on the longitude and latitude of the two points. The estimated velocity is obtained by combining the time difference between the recording times of the two points. The estimated acceleration is obtained by the velocity change and time difference of the adjacent trajectory points. Trajectory points whose estimated velocity is greater than the preset upper limit threshold or whose estimated acceleration exceeds the preset acceleration range are identified as drift points and removed from the trajectory dataset. The remaining data constitute the first trajectory dataset. The estimated velocity exceeding the limit corresponds to vehicle instantaneous distortion, and the estimated acceleration exceeding the limit corresponds to speed change caused by signal jump.

[0028] Next, the absolute value of the azimuth change of two adjacent trajectory points in the first trajectory dataset with a sampling interval less than a preset sampling interval threshold is calculated. The trajectory point whose absolute azimuth change exceeds the preset azimuth jump threshold is then removed from the first trajectory dataset, resulting in the second trajectory dataset. It's important to understand that the preset sampling interval threshold refers to the upper limit of the sampling interval used to determine azimuth jumps; in this example, it's set to 15 seconds. Abrupt azimuth jumps are only considered abnormal when the sampling interval is less than this value. The preset azimuth jump threshold refers to the upper limit of the azimuth change used to determine positional abrupt changes; in this example, it's set to 90 degrees, based on the fact that a truck's trajectory is smooth when driving normally on a road network, and it won't experience right-angle turns in a short period. Specifically, for the first trajectory dataset, two adjacent trajectory points with a sampling interval less than a preset sampling interval threshold are selected, and the absolute value of the difference in azimuth angle between the two points is calculated as the absolute value of the azimuth angle change. When this value is greater than the preset azimuth angle jump threshold, it indicates that the latter trajectory point has undergone a turn or abrupt change that does not conform to the road driving pattern relative to the former trajectory point. The latter trajectory point is removed from the first trajectory dataset to obtain the second trajectory dataset. This judgment is not made for trajectory points with a large sampling interval to avoid misjudging normal turns as abnormal.

[0029] Then, spatial constraint verification is performed on the second trajectory dataset based on the road network buffer zone corresponding to the preset road level. Trajectory points deviating from the road network buffer zone are removed from the second trajectory dataset to obtain the third trajectory dataset. It's important to understand that the road network buffer zone refers to a strip-shaped area extending to both sides of a predetermined width from the centerline of each level of highway. The width is set according to the road level; for expressways, it can be 50 meters, and for ordinary roads, it can be 20 to 30 meters. The value is chosen to cover the road surface width while tolerating conventional positioning errors. Specifically, based on the road network data of the target area, road network buffer zones of corresponding widths are generated for expressways and ordinary roads of each level according to the preset road level. Each trajectory point in the second trajectory dataset is superimposed and verified against the road network buffer zone of its corresponding road. Trajectory points falling outside the road network buffer zone are determined to be spatial positioning failure points and are removed from the second trajectory dataset to obtain the third trajectory dataset.

[0030] Finally, interpolation is performed on trajectory segments in the third trajectory dataset where the sampling interval is greater than the preset missing interval threshold to complete the cleaned trajectory dataset. It's important to understand that the preset missing interval threshold refers to the lower limit of the time interval for determining missing trajectories; in this example, it's set to 60 seconds. If this interval is exceeded, a missing segment is considered to exist between consecutive trajectory points. Specifically, the trajectory point sequence of each truck in the third trajectory dataset is traversed, and positions where the sampling interval between adjacent trajectory points is greater than the preset missing interval threshold are found. These positions are the missing segments. Linear interpolation is used to fill in trajectory points within the missing segments at fixed time intervals. The longitude and latitude of the filled points are obtained by interpolating the trajectory points at both ends of the missing segment according to the time ratio. The completed data constitutes the cleaned trajectory dataset.

[0031] Step S30: Identify the stopping points of the continuous trajectories of each truck in the cleaning trajectory dataset, and divide the continuous trajectories into travel chains based on the identified stopping points to obtain a travel chain set.

[0032] It should be noted that a stop point refers to a spatiotemporal convergence point formed when a truck stays in one location for more than a set time, corresponding to actual transportation behaviors such as loading, unloading, and waiting for dispatch; a travel chain refers to a continuous trajectory segment between two adjacent stop points of the same truck, which includes a starting point and an ending point.

[0033] Further, step S30 includes: First, traversing the cleaning trajectory dataset to obtain the continuous trajectories of each truck. It should be understood that a continuous trajectory refers to a complete sequence of trajectory points formed by the same truck in the cleaning trajectory dataset, ordered by recording time.

[0034] Secondly, based on the continuous trajectory, the set of trajectory points that continuously fall into a circular area with a preset dwell radius threshold and whose cumulative dwell time is not less than the preset dwell time threshold are converged into a stop point. It is important to understand that the preset dwell radius threshold refers to the spatial tolerance for determining that the truck is in a stationary state. In this example, it is set to 50 meters, based on the fact that the positioning signal still has a random drift of tens of meters when the truck is stationary. This value can tolerate such drift without mixing adjacent parking positions together, and can be adjusted between 30 meters and 100 meters. The preset dwell time threshold refers to the lower limit of the cumulative dwell time for determining that the truck forms a stop. In this example, it is set to 30 minutes, based on the fact that the dwell time of actual transportation behaviors such as loading and unloading and waiting for dispatch usually exceeds this time, while the dwell time of transit and short-term waiting at red lights is significantly shorter than this time. The stop point sequence refers to the sequence formed by arranging all the stop points of the same truck in chronological order of recording time. Specifically, for the continuous trajectory of each truck, scanning starts from the first trajectory point and proceeds backward. A circular area is constructed with the current trajectory point as the center and a preset dwell radius threshold as the radius. It is determined whether subsequent trajectory points continuously fall into this circular area. If so, the dwell time of these trajectory points within the circular area is accumulated. When the accumulated dwell time is not less than the preset dwell time threshold, all trajectory points falling into the circular area are converged into a single stopping point. The position of the stopping point is taken as the center position of the trajectory point set, and the time of the stopping point is taken as the recording time of the first trajectory point in the trajectory point set. Then, scanning continues from the next trajectory point outside the circular area until the continuous trajectory is processed. If the accumulated dwell time does not reach the preset dwell time threshold, convergence is not performed, and the scanning position is moved backward directly. Therefore, trajectory points corresponding to situations such as truck slowing down or making temporary avoidance will not be misidentified as stopping points.

[0035] Next, the stopping points of each truck are sorted from earliest to latest according to the recorded time of the trajectory points corresponding to the stopping points, and the stopping point sequence corresponding to each truck is obtained. The two adjacent stopping points in the stopping point sequence correspond to the two ends of a complete transportation.

[0036] Then, the trajectory segment between two adjacent stops in the stop-point sequence is divided into travel chains, and the two adjacent stops are determined as the two endpoints of the travel chain. It's important to understand that the two endpoints refer to the starting and ending stops of a travel chain in terms of time sequence; the starting stop corresponds to the beginning of the current transport, and the ending stop corresponds to the end. Specifically, two adjacent stops are taken from the stop-point sequence, and the trajectory segment between them is divided into a travel chain. The stop that appears earlier in the time sequence is determined as the first endpoint of the travel chain, and the stop that appears later in the time sequence is determined as the last endpoint. These two endpoints together constitute the two endpoints of the travel chain. In this way, the trajectory segment between each pair of adjacent stops in the stop-point sequence is divided into a travel chain.

[0037] Finally, all the travel chains segmented from each truck are aggregated to obtain a travel chain set. Each travel chain carries the corresponding license plate identifier, the two ends of the stop point and the trajectory segment information, which are then processed one by one for subsequent spatial determination.

[0038] Step S40: Construct a berth interest surface buffer zone based on the berth feature library, and perform spatial determination on the two ends of each travel chain in the travel chain set and the berth interest surface buffer zone to obtain the target berth point that falls into the berth interest surface buffer zone.

[0039] It should be noted that the berth interest surface buffer zone refers to the planar area formed by extending a set distance outward from the berth spatial vector data in the berth feature database. The extended distance is used to accommodate vehicles queuing outside the port gate. In this embodiment, it is set to 200 meters. This distance is determined based on the length of the queuing area outside the common port gate and can be adjusted between 100 meters and 500 meters according to the actual layout of the port area. The target stopping point refers to the endpoint of the travel chain that falls within the berth interest surface buffer zone after spatial determination.

[0040] Specifically, firstly, the spatial vector data of each berth is read from the berth feature database. Based on the boundary of each polygon, a buffer zone of 200 meters is generated outward to form the berth interest surface buffer zone corresponding to each berth. The buffer zone covers the berth operation surface, the yard entrance and exit, and the queuing area outside the gate. Trucks queuing normally to enter the port are therefore included in the judgment range, while road parking far away from the port area will not fall into the buffer zone.

[0041] Then, the set of travel chains is traversed, and the two endpoints of each travel chain are selected. For each endpoint, spatial determination is performed with the berth interest surface buffers of all berths. A ray is drawn horizontally from the endpoint, and a ray intersection algorithm is used to count the number of intersections between the ray and each edge of the berth interest surface buffer. The boundaries of the berth interest surface buffer may have indentations and irregular turns. The ray intersection algorithm does not depend on the shape regularity of the region. If the number of intersections is odd, the endpoint is determined to fall inside the berth interest surface buffer; if the number is even, it is determined to be outside the buffer. It is important to understand that the ray intersection algorithm is a computational geometry algorithm that determines the inclusion relationship between a point and a polygon by drawing a ray in any direction from the point to be determined and counting the number of intersections between the ray and the polygon boundary. If the number of intersections is odd, the point to be determined is inside the polygon; if the number is even, it is outside.

[0042] Next, all stops that are determined to have fallen into the target stop are recorded as the target stop for the corresponding berth. The target stop retains the original travel chain association information, including the travel chain to which it belongs, the license plate identifier, and the corresponding berth number. When a stop falls into the berth interest surface buffer of multiple berths at the same time, it is recorded as multiple target stop corresponding to each berth, and handed over to the subsequent time verification for further identification.

[0043] Step S50: Calculate the actual dwell time based on the travel chain corresponding to the target stop. When the actual dwell time exceeds the preset loading and unloading time threshold, mark the travel chain corresponding to the target stop as a valid collection and distribution trajectory with loading and unloading cargo type label.

[0044] It should be noted that the actual dwell time refers to the duration of the travel chain corresponding to the target berth within the berth's interest surface buffer zone; the preset loading and unloading time threshold refers to the pre-set benchmark time for loading and unloading operations based on cargo type. For container berths, this can be 30 minutes, and for dry bulk cargo and general cargo berths, it can be 60 minutes. The basis for this value is that container loading and unloading has a high degree of mechanization and fast vehicle turnover, while bulk cargo involves weighing, queuing, and leveling processes, which take longer. This threshold can be adjusted according to the port's actual operational efficiency statistics; the effective collection and distribution trajectory refers to the travel chain related to berth loading and unloading behavior, which is confirmed by both spatial determination and time verification.

[0045] Further, step S50 includes: First, extracting the entry and exit times of the travel chain corresponding to the target stop within the berth interest surface buffer zone, and calculating the actual dwell time based on the time difference between the exit and entry times. Specifically, for each target stop, extracting the trajectory segment of its respective travel chain, finding the first trajectory point that falls into the berth interest surface buffer zone within the trajectory segment, and using the corresponding recorded time as the entry time; finding the last trajectory point that falls into the berth interest surface buffer zone, and using the corresponding recorded time as the exit time; the difference between the exit and entry times yields the actual dwell time, which reflects the total dwell time of the truck within the port area from entry to exit, rather than just the stationary time at the stop.

[0046] Secondly, based on the berth identifier corresponding to the berth in the berth interest surface buffer where the target stopping point falls, the loading / unloading category label corresponding to the berth identifier is retrieved from the berth function mapping relationship. It's important to understand that the berth identifier is a number assigned to each berth in the berth feature database, used to associate berth spatial vector data, loading / unloading category labels, and berth function mapping relationships. Specifically, the berth identifier corresponding to the berth in the berth interest surface buffer where the target stopping point falls is read, and the corresponding loading / unloading category label is retrieved from the berth function mapping relationship using the berth identifier as an index. When a truck stops at a berth with a clearly defined cargo type attribute, the category of its transported cargo can be inferred from the berth attribute.

[0047] Next, the preset loading and unloading time thresholds corresponding to the loading and unloading category labels are read from the preset cargo category time lookup table. It's important to understand that the preset cargo category time lookup table records the correspondence between each loading and unloading category label and the preset loading and unloading time threshold. Specifically, the corresponding preset loading and unloading time threshold is read from the preset cargo category time lookup table based on the loading and unloading category label. Since the loading and unloading process times differ significantly for different cargo categories, thresholds are set separately for each cargo category to avoid the problem of a uniform threshold being too strict for fast-loading cargo categories and too lenient for slow-loading cargo categories.

[0048] Then, the types of points of interest (POIs) within the preset range around the target stop are obtained. When the POI type belongs to the preset non-operational type and the actual dwell time is less than the preset rest time threshold, the travel chain corresponding to the target stop is marked as an invalid trajectory. It is important to understand that the preset POI range refers to the distance range of POIs retrieved from the target stop as a reference point. In this example, it is set to 100 meters, based on the fact that maintenance, catering, and other service facilities in the port area are usually located close to the truck parking locations. The POI type refers to the classification of POIs in the map data, which can be categories such as car repair, catering and accommodation, gas stations, and logistics warehouses. The preset non-operational type refers to the pre-defined POI types that are unrelated to port loading and unloading operations. In this example, it includes two categories: car repair and catering and accommodation. The preset rest time threshold is the upper limit of the dwell time used to identify rest-type stays. In this example, it is set to 90 minutes, based on the fact that the stay for meals and short rests usually does not exceed this time, while the dry bulk cargo loading and unloading process generally takes longer than this time. Specifically, based on the target stop point, the system retrieves points of interest from the map data within the preset point of interest range, reads the type of point of interest, and indicates that the truck stop is more in line with the rest characteristics such as dining or maintenance when the point of interest type belongs to the preset non-operation type and the actual stay time is less than the preset rest time threshold. The travel chain corresponding to the target stop point is marked as an invalid trajectory. This marking is only triggered when there happens to be such service facilities near the berth gate and the stay characteristics match. The long stay of normal loading and unloading will not be affected.

[0049] Finally, when the actual dwell time exceeds the preset loading / unloading time threshold and the travel chain corresponding to the target docking point is not marked as an invalid trajectory, the travel chain corresponding to the target docking point is marked as a valid collection and distribution trajectory with a loading / unloading cargo category label. It is important to understand that an invalid trajectory refers to a travel chain that has been identified as unrelated to port loading / unloading activities. Specifically, for each target docking point, a final confirmation is performed. When the actual dwell time exceeds the preset loading / unloading time threshold and the corresponding travel chain is not marked as an invalid trajectory, the travel chain is marked as a valid collection and distribution trajectory with the corresponding loading / unloading cargo category label. When a travel chain falls into the berth interest surface buffer zone of multiple berths simultaneously, only the locations that meet the corresponding berth duration verification are marked; locations that do not meet the verification are not marked.

[0050] Step S60: Extract the landside stops located on the landside from each effective collection and distribution trajectory, and perform density clustering on the landside stops with the same loading and unloading cargo type labels. Generate the core cargo source agglomeration area based on the clusters obtained from the clustering.

[0051] It should be noted that landside docking points refer to the docking points at one end of the effective collection and distribution trajectory that do not fall into the berth interest surface buffer zone, corresponding to the actual place of origin or consumption of goods; core cargo source agglomeration area refers to the area enclosed by the outer boundary of the clusters formed by density clustering of landside docking points, corresponding to the industrial agglomeration area with stable cargo sources in reality.

[0052] Further, step S60 includes: First, extracting the docking points at one end of each effective collection and distribution trajectory that do not fall into the berth interest surface buffer zone as landside docking points, and grouping the landside docking points according to the loading and unloading cargo category label to obtain the landside docking point set corresponding to each loading and unloading cargo category label. It should be understood that the landside docking point set refers to the set of all landside docking points with the same loading and unloading cargo category label. Specifically, traversing each effective collection and distribution trajectory, comparing the positional relationship between the docking points at both ends and the berth interest surface buffer zone, extracting the docking points at the one end that do not fall into the berth interest surface buffer zone as landside docking points. The landside docking points correspond to the actual place of origin or consumption of goods inland. Then, grouping the landside docking points according to the loading and unloading cargo category label carried by each effective collection and distribution trajectory, and grouping the landside docking points with the same label into the same landside docking point set. The source locations of different cargo categories differ significantly in industrial distribution, and the grouping process makes the clustering results correspond to the industrial spatial pattern of each cargo category.

[0053] Next, the number of neighboring docking points within a preset neighborhood radius of each landside docking point in the landside docking point set is calculated. Landside docking points with a neighboring docking point count not less than a preset minimum threshold are defined as core docking points. It's important to understand that the preset neighborhood radius refers to the distance range used in density clustering to count neighboring docking points; in this example, it's 2km. This value is based on the fact that the spatial scale of typical provincial industrial parks or logistics bases in China is approximately 2km, and this radius can cover scattered loading / unloading platforms and parking locations within the same cargo origin. The number of neighboring docking points refers to the number of other landside docking points within the preset neighborhood radius of a given landside docking point. The preset minimum threshold is the minimum number of docking points required to determine that an area has statistical significance; in this example, it's 50, based on the fact that a cargo origination area needs to reach a docking scale of 50 vehicles within the observation period to indicate the existence of stable industrial cargo sources; clusters below this scale are mostly sporadic transportation. Core docking points are landside docking points with a neighboring docking point count not less than the preset minimum threshold. Specifically, for each set of landside docking points, the spherical distance formula is used to calculate the distance between each landside docking point in the set. The number of neighboring docking points within a preset neighborhood radius of each landside docking point is counted. Landside docking points with a number of neighboring docking points not less than a preset minimum number of points are determined as core docking points. A large number of trucks with the same type of cargo gather around the location of the core docking point, corresponding to a stable cargo source generation area.

[0054] Then, starting from each core stop, landside stops with density reachable within a preset neighborhood radius are grouped into the same cluster, and landside stops not grouped into any cluster are removed from the landside stop set, resulting in various clusters. It's important to understand that density reachability refers to a chain-like connection between two landside stops through several intermediate stops within a preset neighborhood radius. Specifically, starting from each core stop, cluster expansion is performed, grouping all landside stops within the preset neighborhood radius of the core stop into the same cluster. For newly grouped core stops, the stops within their neighborhood are further grouped into the same cluster, and this expansion continues along the density reachable relationship until the cluster boundary is sparsely populated and no stops can be grouped into it. Landside stops not grouped into any cluster are considered isolated points formed by accidental transport of goods along the driver's route or location drift, and are removed from the landside stop set, resulting in various clusters.

[0055] Finally, the coordinates of the outermost landside docking points in each cluster are extracted, and polygonal regions are generated based on these coordinates. These polygonal regions are then identified as core cargo aggregation areas. Specifically, for each cluster, the coordinates of the outermost landside docking points are extracted, and a convex hull algorithm is used to generate polygonal regions with the outermost landside docking points as vertices. These polygonal regions are identified as core cargo aggregation areas, each labeled with a corresponding loading / unloading cargo type, and their boundaries correspond to the distribution of industrial parks and logistics bases in reality.

[0056] Step S70: Calculate the freight connection frequency contribution of each core cargo agglomeration area to the corresponding berth based on each effective collection and distribution trajectory, and divide each core cargo agglomeration area according to the cumulative proportion of freight connection frequency contribution to generate berth-level hinterland division results.

[0057] It should be noted that the contribution of freight connection frequency refers to the proportion of the freight connection frequency from a certain core cargo agglomeration area to the corresponding berth to the sum of the freight connection frequencies of all core cargo agglomeration areas of that berth; the berth-level hinterland division result refers to the hinterland spatial structure formed after classifying each core cargo agglomeration area according to level, including three levels: direct core hinterland, indirect competitive hinterland, and peripheral radiating hinterland. The direct core hinterland refers to the area with the strongest connection to the berth's cargo sources, the indirect competitive hinterland refers to the area with a certain degree of cargo mobility and is contested by multiple ports, and the peripheral radiating hinterland refers to the area with only scattered cargo source connections.

[0058] Specifically, the effective operating frequency of each truck between each core cargo aggregation area and its corresponding berth is statistically analyzed within a preset observation period. The effective operating frequency of each truck is summarized and divided by the number of days corresponding to a preset time unit to obtain the freight connection frequency between each core cargo aggregation area and its corresponding berth. The proportion of each freight connection frequency to the sum of the freight connection frequencies of all core cargo aggregation areas at the corresponding berth is calculated to obtain the freight connection frequency contribution. The core cargo aggregation areas are then ranked from largest to smallest according to the freight connection frequency contribution and accumulated sequentially to obtain the cumulative proportion of each core cargo aggregation area. Following the Pareto principle (80 / 20 rule), a hierarchical division is performed, with the top 80% of the cumulative proportion designated as the direct core hinterland, and the area between 80% and 95% designated as... The remaining 5% of the area is designated as the indirect competitive hinterland, and the peripheral radiation hinterland. For core cargo gathering areas that simultaneously transport the same type of goods to berths A and B, the smaller of the freight connection frequency to berth A and the freight connection frequency to berth B is taken as twice, and divided by the sum of the two to obtain the hinterland overlap coefficient. The hinterland overlap coefficient is calculated for all berth pairs with common cargo source areas. Finally, the direct core hinterland, indirect competitive hinterland, peripheral radiation hinterland, and hinterland overlap coefficients of each berth pair are integrated into the berth-level hinterland division results. The hinterland at each level can be visualized in the form of different color blocks and the hinterland overlap coefficient in the form of a heat map through a geographic information system, which intuitively presents the monopoly area, competitive area, and blank area in the water transport market.

[0059] This embodiment constructs a functional mapping relationship between berth spatial vector data and loading / unloading cargo category labels to clean and segment the trajectories of heavy-duty trucks. It employs a dual mechanism of berth interest surface spatial determination and loading / unloading dwell time verification based on cargo category attributes to accurately extract effective collection and distribution trajectories. Furthermore, it uses landside berth density clustering to obtain core cargo aggregation areas and divides the hinterland into multi-level zones based on the contribution of freight connection frequency. This application effectively eliminates noisy trajectories such as transit, maintenance, and parking routes, improving the accuracy of hinterland division and providing reliable support for port resource allocation and competitive strategy formulation.

[0060] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 The port hinterland delineation method based on heavy-duty truck trajectories, step S70, further includes steps S201 to S205: Step S201: Calculate the effective operating frequency of each truck between each core cargo gathering area and its corresponding berth. Summarize the effective operating frequency of each truck and divide it by the number of days corresponding to the preset time unit to obtain the freight connection frequency between each core cargo gathering area and its corresponding berth.

[0061] It should be noted that, for each berth, effective collection and distribution trajectories located at both ends within the core cargo aggregation area and the corresponding berth's interest surface buffer zone are selected. Based on license plate identification and the round-trip core cargo aggregation area-berth pair, the effective operating frequency of each truck between each aggregation area-berth pair is calculated. The effective operating frequencies of all trucks traveling between the same core cargo aggregation area and the corresponding berth are summarized and divided by the total number of days in the preset observation period to obtain the average daily freight connection frequency between the core cargo aggregation area and the corresponding berth. The specific formula is as follows: in, Core cargo gathering area to berth Freight contact frequency, The total number of days in the observation period. Traveling to and from the core cargo gathering area during the observation period berths The collection of all the trucks between them. For trucks In core cargo gathering areas berths Effective operational frequency between the two. Freight link frequency refers to the number of truck trips between the core cargo gathering area and the corresponding berth per unit time. It is used to characterize the frequency of freight activities between the two. The higher the freight link frequency, the closer the freight link between the two.

[0062] Step S202: Calculate the proportion of the freight connection frequency of each core cargo source cluster area to the sum of the freight connection frequencies of all core cargo source cluster areas of the corresponding berth, and obtain the freight connection frequency contribution of each core cargo source cluster area.

[0063] It should be noted that, for each berth, the freight connection frequency between each core cargo source cluster and that berth is divided by the sum of the freight connection frequencies of all core cargo source clusters corresponding to that berth. This yields the contribution of each core cargo source cluster to the freight connection frequency of that berth, calculated using the following formula: in, Core cargo gathering area Contribution to the frequency of cargo transportation connections at berth j The frequency of freight connections between core cargo aggregation area i and berth j. This refers to the collection of all core cargo gathering areas that have effective collection and distribution connections with berth j. This represents the sum of freight connection frequencies for all core cargo aggregation areas corresponding to berth j. The freight connection frequency contribution rate is used to characterize the relative proportion of each core cargo aggregation area in all freight connections for the corresponding berth. The larger the freight connection frequency contribution rate, the closer the connection between the corresponding core cargo aggregation area and the berth.

[0064] Step S203: Sort each core cargo gathering area in descending order of its contribution to freight connection frequency, and then accumulate the contribution to freight connection frequency in turn to obtain the cumulative proportion of each core cargo gathering area.

[0065] It should be noted that, for each berth, its corresponding core cargo aggregation areas are ranked from highest to lowest according to their contribution to freight connection frequency. Starting from the top-ranked core cargo aggregation area, the freight connection frequency contribution of each core cargo aggregation area is accumulated sequentially to obtain the cumulative percentage up to the current core cargo aggregation area. The specific formula is as follows: in, The cumulative percentage of the core cargo cluster area corresponding to berth j when sorted up to the m-th position. This represents the contribution of the qth core cargo cluster to berth j in terms of freight connection frequency, sorted from largest to smallest.

[0066] The cumulative percentage of the top-ranked core cargo cluster is equal to its own contribution to freight connection frequency; the cumulative percentage of the second-ranked core cargo cluster is equal to the sum of the freight connection frequency contributions of the top two core cargo clusters, and so on. This cumulative percentage is used to subsequently divide each core cargo cluster into direct core hinterland, indirect competitive hinterland, and peripheral radiation hinterland.

[0067] Step S204: The core cargo agglomeration areas with a cumulative proportion within the first preset proportion range are classified as direct core hinterlands, the core cargo agglomeration areas with a cumulative proportion within the second preset proportion range are classified as indirect competitive hinterlands, and the core cargo agglomeration areas with a cumulative proportion within the third preset proportion range are classified as peripheral radiation hinterlands.

[0068] It should be noted that, based on the Pareto principle, the core cargo gathering areas with a cumulative share within the first preset share range are classified as direct core hinterlands. These areas have formed a stable cargo dependence relationship with the berths and are the main source of berth throughput. The core cargo gathering areas with a cumulative share within the second preset share range are classified as indirect competitive hinterlands. The cargo in these areas has a certain degree of mobility and is often located in the middle of multiple ports or multiple modes of transportation. The core cargo gathering areas with a cumulative share within the third preset share range are classified as peripheral radiation hinterlands. These areas are mostly characterized by occasional and scattered long-tail freight demand.

[0069] Step S205: For the same type of loading and unloading cargo, identify the common core cargo agglomeration area that has effective collection and distribution trajectories with both the first and second berths. Obtain the first freight connection frequency between the common core cargo agglomeration area and the first berth, and the second freight connection frequency between the common core cargo agglomeration area and the second berth. Calculate the hinterland overlap coefficient based on the first and second freight connection frequencies, and generate berth-level hinterland division results based on the hinterland overlap coefficient, direct core hinterland, indirect competing hinterland, and peripheral radiation hinterland.

[0070] It should be noted that the hinterlands of different berths often overlap spatially. Cargo from the same inland area may simultaneously flow to two berths. To quantify this competition for cargo resources, it is necessary to calculate the hinterland overlap coefficient between berths. For a core cargo cluster area that simultaneously supplies the same type of cargo to both berths, the first cargo transport frequency between this core cargo cluster area and the first berth, and the second cargo transport frequency between this core cargo cluster area and the second berth are obtained. The smaller of the first and second cargo transport frequencies is taken as twice, and divided by the sum of the first and second cargo transport frequencies to obtain the hinterland overlap coefficient. The specific formula is as follows: in, Let be the hinterland overlap coefficient of the core cargo gathering area i relative to the first berth A and the second berth B. The first freight transport frequency between the core cargo aggregation area i and the first berth A. Let be the second freight connection frequency between the core cargo aggregation area i and the second berth B, and min represents taking the smaller value between the first and second freight connection frequencies.

[0071] When the frequency of freight connections between the core cargo agglomeration area and the two berths is equal, the hinterland overlap coefficient is 1, indicating that the degree of freight connection between the core cargo agglomeration area and the two berths is the same; the greater the difference between the first freight connection frequency and the second freight connection frequency, the closer the hinterland overlap coefficient is to 0, indicating that the freight connection of the core cargo agglomeration area is more inclined to one of the berths.

[0072] For each berth pair sharing a common core cargo cluster, the hinterland overlap coefficient is calculated. This coefficient is then integrated with the corresponding direct core hinterland, indirect competing hinterland, and peripheral radiating hinterland to generate berth-level hinterland delineation results. These results can be visualized using a geographic information system (GIS), where different color blocks represent different levels of hinterland, and heatmaps display the hinterland overlap coefficients to illustrate the connection tendencies and competition levels between core cargo clusters and different berths.

[0073] This embodiment determines the frequency of freight connections between core cargo gathering areas and corresponding berths by summarizing the effective operating frequencies of each truck within a preset observation period. The trajectory records are then transformed into a quantitative indicator reflecting the closeness of freight connections between cargo areas and berths. By ranking the contribution of freight connection frequencies and calculating the cumulative proportion, the berth hinterland is divided into three levels: direct core hinterland, indirect competitive hinterland, and peripheral radiation hinterland, reflecting the connection strength and relative importance between different cargo areas and corresponding berths. Through the calculation of the hinterland overlap coefficient, a unified indicator is used to characterize the degree and tendency of overlap in freight connections between common core cargo gathering areas and different berths. This allows the berth-level hinterland division results to provide a reference for port cargo organization, inland terminal layout, and the formulation of differentiated competitive strategies.

[0074] Based on the first embodiment of this application, this application also provides a port hinterland delineation device based on heavy-duty truck trajectories. Please refer to... Figure 3 The device includes: The berth feature library construction module 10 is used to obtain the berth spatial vector data and loading / unloading category labels of each berth in the target area, and to construct the berth function mapping relationship between the berth spatial vector data and the loading / unloading category labels to obtain the berth feature library.

[0075] The trajectory cleaning module 20 is used to acquire the trajectory dataset of trucks whose load reaches the preset load threshold within the target area, and to remove anomalies and repair missing data in the trajectory dataset to obtain a cleaned trajectory dataset.

[0076] The trip chain segmentation module 30 is used to identify the stop points of the continuous trajectories of each truck in the cleaning trajectory dataset, and to segment the continuous trajectories into trip chains based on the identified stop points, thereby obtaining a trip chain set.

[0077] The spatial determination module 40 is used to construct a berth interest surface buffer based on the berth feature library, and to perform spatial determination on the two ends of each travel chain in the travel chain set and the berth interest surface buffer to obtain the target berth that falls into the berth interest surface buffer.

[0078] The duration verification module 50 is used to calculate the actual dwell time based on the travel chain corresponding to the target stop. When the actual dwell time exceeds the preset loading and unloading time threshold, the travel chain corresponding to the target stop is marked as a valid collection and distribution trajectory with loading and unloading cargo type label.

[0079] Clustering module 60 is used to extract landside stops located on the landside in each effective collection and distribution trajectory, and to perform density clustering on landside stops with the same loading and unloading cargo category labels, and to generate core cargo source agglomeration areas based on the clusters obtained from the clustering.

[0080] The hinterland division module 70 is used to calculate the contribution of freight connection frequency from each core cargo agglomeration area to the corresponding berth based on each effective collection and distribution trajectory, and to divide each core cargo agglomeration area according to the cumulative proportion of freight connection frequency contribution, generating berth-level hinterland division results.

[0081] The port hinterland division device based on heavy-duty truck trajectories provided in this application adopts the port hinterland division method based on heavy-duty truck trajectories in the above embodiments, which can solve the technical problem of how to achieve accurate quantitative division of port hinterland in two dimensions: berths and subdivided cargo types, based on heavy-duty truck operation trajectory data. Compared with the prior art, the beneficial effects of the port hinterland division device based on heavy-duty truck trajectories provided in this application are the same as those of the port hinterland division method based on heavy-duty truck trajectories provided in the above embodiments, and other technical features in the port hinterland division device based on heavy-duty truck trajectories are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0082] This application provides a port hinterland delineation device based on heavy-duty truck trajectories. The port hinterland delineation device based on heavy-duty truck trajectories includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the port hinterland delineation method based on heavy-duty truck trajectories in the above embodiment 1.

[0083] The following is for reference. Figure 4 The diagram illustrates a structural schematic of a port hinterland delineation device based on heavy-duty truck trajectories suitable for implementing embodiments of this application. The port hinterland delineation device based on heavy-duty truck trajectories in this application embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle-mounted terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The port hinterland delineation device based on heavy-duty truck trajectories shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0084] like Figure 4As shown, the port hinterland delineation device based on heavy-duty truck trajectories may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1002 or a program loaded from storage device 1003 into random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the port hinterland delineation device based on heavy-duty truck trajectories. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the port hinterland delineation equipment based on heavy-duty truck tracks to communicate wirelessly or wiredly with other devices to exchange data. Although various port hinterland delineation devices based on heavy-duty truck tracks are shown in the figures, it should be understood that implementation or possession of all of them is not required. More or fewer devices may be implemented alternatively.

[0085] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0086] The port hinterland delineation device based on heavy-duty truck trajectories provided in this application adopts the port hinterland delineation method based on heavy-duty truck trajectories in the above embodiments, which can solve the technical problem of how to achieve accurate quantitative delineation of port hinterland based on both berths and subdivided cargo types by relying on heavy-duty truck operation trajectory data. Compared with the prior art, the beneficial effects of the port hinterland delineation device based on heavy-duty truck trajectories provided in this application are the same as those of the port hinterland delineation method based on heavy-duty truck trajectories provided in the above embodiments, and other technical features in the port hinterland delineation device based on heavy-duty truck trajectories are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0087] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0088] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0089] This application provides a computer-readable medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the port hinterland delineation method based on heavy-duty truck trajectories in the above embodiments.

[0090] The computer-readable medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor devices, or any combination thereof. More specific examples of computer-readable media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable medium may be any tangible medium containing or storing a program that can be executed by instructions, used by a device, or used in conjunction with it. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0091] The aforementioned computer-readable medium may be included in a port hinterland delineation device based on heavy-duty truck trajectories; or it may exist independently and not be assembled into a port hinterland delineation device based on heavy-duty truck trajectories.

[0092] The aforementioned computer-readable medium carries one or more programs that, when executed by a port hinterland delineation device based on heavy-duty truck trajectories, enable the port hinterland delineation device to write computer program code for performing the operations of this application in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0093] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of this application. In this regard, all blocks in the flowcharts or block diagrams may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that all blocks in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using dedicated hardware-based implementations that perform the specified functions or operations, or using a combination of dedicated hardware and computer instructions.

[0094] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0095] The readable medium provided in this application is a computer-readable medium, which stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned port hinterland division method based on heavy-duty truck trajectories. This solves the technical problem of how to achieve accurate quantitative division of port hinterland based on both berth and cargo category dimensions using heavy-duty truck trajectory data. Compared with the prior art, the beneficial effects of the computer-readable medium provided in this application are the same as those of the port hinterland division method based on heavy-duty truck trajectories provided in the above embodiments, and will not be elaborated upon here.

[0096] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the port hinterland delineation method based on heavy-duty truck trajectories described above.

[0097] The computer program product provided in this application can solve the technical problem of how to accurately quantify and divide the port hinterland in two dimensions—berths and specific cargo types—based on the trajectory data of heavy-duty trucks. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the port hinterland division method based on heavy-duty truck trajectories provided in the above embodiments, and will not be repeated here.

[0098] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for port hinterland delineation based on heavy-duty truck trajectories, characterized in that, The method includes: Acquire berth spatial vector data and cargo loading / unloading category labels for each berth within the target area, and construct the berth function mapping relationship between the berth spatial vector data and the cargo loading / unloading category labels to obtain a berth feature library; Obtain the trajectory dataset of trucks whose load reaches a preset load threshold within the target area, and perform anomaly removal and missing data repair on the trajectory dataset to obtain a cleaned trajectory dataset. Stop points are identified for the continuous trajectories of each truck in the cleaning trajectory dataset, and the continuous trajectories are divided into travel chains based on the identified stop points to obtain a travel chain set; A berth interest surface buffer is constructed based on the berth feature library, and spatial determination is performed on the two ends of each travel chain in the travel chain set and the berth interest surface buffer to obtain the target berth point that falls into the berth interest surface buffer. The actual dwell time is calculated based on the travel chain corresponding to the target stop. When the actual dwell time exceeds the preset loading and unloading time threshold, the travel chain corresponding to the target stop is marked as a valid collection and distribution trajectory with the loading and unloading cargo type label. Extract the landside stops located on the landside from each of the effective collection and distribution trajectories, and perform density clustering on the landside stops with the same loading and unloading cargo type labels. Generate core cargo source agglomeration areas based on the clusters obtained from the clustering. The contribution of freight transport frequency from each core cargo gathering area to the corresponding berth is calculated based on the effective collection and distribution trajectories. The core cargo gathering areas are then divided according to the cumulative proportion of the contribution of freight transport frequency, generating berth-level hinterland division results.

2. The method as described in claim 1, characterized in that, The steps of acquiring berth spatial vector data and cargo loading / unloading category tags for each berth within the target area, and constructing a berth function mapping relationship between the berth spatial vector data and the cargo loading / unloading category tags to obtain a berth feature library include: Acquire the polygonal boundary data of the working face and the polygonal boundary data of the rear storage yard entrance and exit of each berth within the target area, and merge the polygonal boundary data of the working face and the polygonal boundary data of the rear storage yard entrance and exit into berth spatial vector data; Obtain the ship reporting data and port production operation ledger of the target area, and extract the loading and unloading tags of each berth from the ship reporting data and port production operation ledger; Establish a one-to-one mapping relationship between the spatial vector data of each berth and the cargo loading / unloading category label to obtain the berth function mapping relationship; The berth spatial vector data, cargo type labels, and berth function mapping relationships of each berth are stored in a preset database to obtain a berth feature library.

3. The method as described in claim 1, characterized in that, The step of obtaining a trajectory dataset of trucks with a load capacity reaching a preset load threshold within the target area, and performing anomaly removal and missing data repair on the trajectory dataset to obtain a cleaned trajectory dataset, includes: The location trajectory records and vehicle rated loads of each truck in the target area are obtained from the road freight vehicle supervision platform. The location trajectory records corresponding to trucks whose rated loads are less than a preset load threshold are filtered out to obtain a trajectory dataset. The location trajectory records include the recording time, longitude, latitude, instantaneous speed and azimuth. The calculated velocity and calculated acceleration are obtained by calculating the longitude, latitude and recording time of two adjacent trajectory points in the trajectory dataset, and trajectory points whose calculated velocity is greater than a preset velocity upper limit threshold or whose calculated acceleration exceeds a preset acceleration range are removed from the trajectory dataset to obtain the first trajectory dataset. Calculate the absolute value of the azimuth change of two adjacent trajectory points in the first trajectory dataset whose sampling interval is less than a preset sampling interval threshold, and remove the next trajectory point whose absolute value of the azimuth change is greater than a preset azimuth jump threshold from the first trajectory dataset to obtain the second trajectory dataset; The second trajectory dataset is spatially constrained according to the road network buffer zone corresponding to the preset road level, and trajectory points that deviate from the road network buffer zone are removed from the second trajectory dataset to obtain the third trajectory dataset; The trajectory segments in the third trajectory dataset with sampling intervals greater than a preset missing interval threshold are interpolated and filled to obtain a cleaned trajectory dataset.

4. The method as described in claim 1, characterized in that, The step of identifying stop points on the continuous trajectories of each truck in the cleaning trajectory dataset, and dividing the continuous trajectories into travel chains based on the identified stop points to obtain a set of travel chains includes: The cleaning trajectory dataset is traversed to obtain the continuous trajectories of each truck; Based on the continuous trajectory, the set of trajectory points that continuously fall into a circular area with a preset dwell radius threshold and whose cumulative dwell time is not less than a preset dwell time threshold are converged into a stopping point; The stopping points of each truck are arranged in chronological order according to the recording time of the trajectory points corresponding to the stopping points, so as to obtain the stopping point sequence corresponding to each truck. The trajectory segment between two adjacent stops in the stop sequence is divided into a travel chain, and the two adjacent stops are determined as the two ends of the travel chain. By summing up all the travel chains segmented by each of the aforementioned trucks, a travel chain set is obtained.

5. The method as described in claim 1, characterized in that, The step of calculating the actual dwell time based on the travel chain corresponding to the target stop, and marking the travel chain corresponding to the target stop as a valid collection and distribution trajectory with the loading and unloading cargo type label when the actual dwell time exceeds a preset loading and unloading time threshold, includes: Extract the entry and exit times of the travel chain corresponding to the target stop within the interest surface buffer of the berth, and calculate the actual dwell time based on the time difference between the exit time and the entry time; Based on the berth identifier of the berth corresponding to the berth interest surface buffer where the target berth falls, retrieve the loading / unloading category label corresponding to the berth identifier from the berth function mapping relationship; Read the preset loading and unloading time threshold corresponding to the loading and unloading cargo category label from the preset cargo category time lookup table, wherein the preset cargo category time lookup table records the correspondence between each loading and unloading cargo category label and the preset loading and unloading time threshold; Obtain the type of point of interest within a preset range of points of interest surrounding the target stop. When the type of point of interest belongs to a preset non-work type and the actual stay time is less than a preset rest time threshold, mark the travel chain corresponding to the target stop as an invalid trajectory. When the actual dwell time exceeds the preset loading and unloading time threshold and the travel chain corresponding to the target stop is not marked as the invalid trajectory, the travel chain corresponding to the target stop is marked as a valid collection and distribution trajectory with the loading and unloading cargo type label.

6. The method as described in claim 1, characterized in that, The step of extracting landside stops located on the landside from each of the effective collection and distribution trajectories, performing density clustering on the landside stops with the same loading and unloading cargo type labels, and generating core cargo source agglomeration areas based on the clustered clusters includes: Extract the docking point at one end of each effective collection and distribution trajectory that does not fall into the berth interest surface buffer zone as the landside docking point, and group the landside docking points according to the loading and unloading cargo category label to obtain the landside docking point set corresponding to each loading and unloading cargo category label; Calculate the number of neighboring stops included in the preset neighborhood radius of each landside stop in the set of landside stops, and determine the landside stops whose number of neighboring stops is not less than the preset minimum number of stops threshold as core stops; Starting from each of the core docking points, landside docking points with achievable density within the preset neighborhood radius are grouped into the same cluster, and landside docking points not grouped into any cluster are removed from the set of landside docking points to obtain various clusters; Extract the coordinates of the outermost landside docking points in each cluster, and generate a polygonal region based on the coordinates of the outermost landside docking points. Then, determine the polygonal region as the core cargo aggregation area.

7. The method as described in claim 1, characterized in that, The steps of calculating the freight connection frequency contribution from each core cargo agglomeration area to the corresponding berth based on each effective collection and distribution trajectory, and dividing each core cargo agglomeration area according to the cumulative proportion of the freight connection frequency contribution to generate berth-level hinterland division results include: The effective operating frequency of each truck between each core cargo gathering area and its corresponding berth is statistically analyzed. The effective operating frequency of each truck is summarized and divided by the number of days corresponding to a preset time unit to obtain the freight connection frequency between each core cargo gathering area and its corresponding berth. Calculate the proportion of the freight connection frequency corresponding to each core cargo agglomeration area to the sum of the freight connection frequencies of all core cargo agglomeration areas of the corresponding berth, and obtain the freight connection frequency contribution of each core cargo agglomeration area. The core cargo gathering areas are sorted from largest to smallest according to their contribution to freight connection frequency, and the contribution to freight connection frequency is accumulated in turn to obtain the cumulative proportion of each core cargo gathering area. The core cargo agglomeration areas with the cumulative proportion within the first preset proportion range are divided into direct core hinterlands, the core cargo agglomeration areas with the cumulative proportion within the second preset proportion range are divided into indirect competitive hinterlands, and the core cargo agglomeration areas with the cumulative proportion within the third preset proportion range are divided into peripheral radiation hinterlands. For the same type of cargo, a common core cargo agglomeration area is identified that has an effective collection and distribution trajectory with both the first berth and the second berth. The first freight connection frequency between the common core cargo agglomeration area and the first berth and the second freight connection frequency between the common core cargo agglomeration area and the second berth are obtained. The hinterland overlap coefficient is calculated based on the first freight connection frequency and the second freight connection frequency. The berth-level hinterland division result is generated based on the hinterland overlap coefficient, the direct core hinterland, the indirect competing hinterland, and the peripheral radial hinterland.

8. A port hinterland delineation device based on heavy-duty truck trajectories, characterized in that, The device is applied to the port hinterland delineation method based on heavy-duty truck trajectories as described in any one of claims 1 to 7, and the device comprises: The berth feature library construction module is used to obtain the berth spatial vector data and loading / unloading category labels of each berth in the target area, and to construct the berth function mapping relationship between the berth spatial vector data and the loading / unloading category labels to obtain the berth feature library; The trajectory cleaning module is used to acquire the trajectory dataset of trucks whose load reaches a preset load threshold within the target area, and to remove anomalies and repair missing data in the trajectory dataset to obtain a cleaned trajectory dataset. The trip chain segmentation module is used to identify the stop points of the continuous trajectories of each truck in the cleaning trajectory dataset, and to segment the continuous trajectories into trip chains based on the identified stop points to obtain a trip chain set. The spatial determination module is used to construct a berth interest surface buffer based on the berth feature library, and to perform spatial determination on the two ends of each travel chain in the travel chain set and the berth interest surface buffer to obtain the target berth that falls into the berth interest surface buffer. The duration verification module is used to calculate the actual dwell time based on the travel chain corresponding to the target stop. When the actual dwell time exceeds the preset loading and unloading time threshold, the travel chain corresponding to the target stop is marked as a valid collection and distribution trajectory with the loading and unloading cargo type label. The clustering module is used to extract the landside stops located on the landside in each of the effective collection and distribution trajectories, and to perform density clustering on the landside stops with the same loading and unloading cargo category labels, and to generate core cargo source agglomeration areas based on the clusters obtained from the clustering. The hinterland division module is used to calculate the freight connection frequency contribution of each core cargo agglomeration area to the corresponding berth based on each effective collection and distribution trajectory, and to divide each core cargo agglomeration area according to the cumulative proportion of the freight connection frequency contribution, thereby generating berth-level hinterland division results.

9. A port hinterland delineation device based on heavy-duty truck trajectories, characterized in that, The device includes: a memory, a processor, and a port hinterland delineation program based on heavy-duty truck trajectories stored in the memory and running on the processor, the port hinterland delineation program based on heavy-duty truck trajectories being configured to implement the steps of the port hinterland delineation method based on heavy-duty truck trajectories as described in any one of claims 1-7.

10. A storage medium, characterized in that, The storage medium stores a port hinterland division program based on heavy-duty truck trajectories. When the port hinterland division program based on heavy-duty truck trajectories is executed by the processor, it implements the steps of the port hinterland division method based on heavy-duty truck trajectories as described in any one of claims 1-7.

Images