Inspection method, device and equipment for passing vehicle smuggling and storage medium

Abstract

The invention discloses an inspection method, device and equipment for transfer vehicle smuggling, and a storage medium. Comprising the following steps: acquiring a multi-source data set, and calculating a vehicle transfer behavior factor based on the multi-source data set; constructing a training data set, and training a smuggling prediction model based on the training data set and a LightGBM model; inputting the vehicle turn-off behavior factor into the smuggling prediction model, and outputting a smuggling prediction score; and when the smuggling prediction score is greater than or equal to a preset threshold value, determining that the vehicle is a suspicious vehicle for transferring smuggling. According to the method, based on the vehicle transfer smuggling behavior characteristics, the vehicle transfer behavior factors are mined from the multi-source data set, the smuggling prediction score is accurately output based on the trained large model, accurate clues are provided for customs anti-smuggling departments, and the investigation success rate is improved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method, apparatus, equipment, and storage medium for investigating smuggling by transit vehicles. Background Technology

[0002] Vehicle transit, as an efficient logistics method, is widely used in cross-border freight transport. However, the complexity and difficulty of regulating vehicle transit also provide opportunities for smuggling. Vehicle transit smuggling refers to smuggling activities carried out during transit, employing diverse and covert methods. Common techniques include misdeclaring product names, concealing quantities, smuggling undeclared goods, and exploiting special zones or policy loopholes. These smuggling activities not only lead to the loss of national tax revenue but also severely disrupt normal trade order.

[0003] Traditional regulatory methods rely mainly on manual inspections and paper document reviews, which are inefficient, costly, and unable to cope with increasingly sophisticated smuggling methods. Summary of the Invention

[0004] This application provides a method, apparatus, equipment, and storage medium for inspecting smuggling by transit vehicles, so as to at least solve the technical problem in the related art that it is difficult to efficiently and accurately inspect smuggling by transit vehicles.

[0005] According to one aspect of the embodiments of this application, a method for investigating smuggling by transit vehicles is provided, comprising: Obtain a multi-source dataset and calculate vehicle checkpoint behavior factors based on the multi-source dataset; Construct a training dataset, and train a smuggling prediction model based on the training dataset and the LightGBM model; The vehicle transfer behavior factor is input into the smuggling prediction model, and the smuggling prediction score is output. When the smuggling prediction score is greater than or equal to a preset threshold, the vehicle is identified as a suspected smuggling vehicle.

[0006] According to another aspect of the embodiments of this application, an inspection device for smuggling vehicles in transit is also provided, comprising: The data processing module is used to acquire multi-source datasets and calculate vehicle transfer behavior factors based on the multi-source datasets. The model training module is used to construct a training dataset and train a smuggling prediction model based on the training dataset and the LightGBM model. The model prediction module is used to input the vehicle transfer behavior factors into the smuggling prediction model and output a smuggling prediction score. The result output module is used to determine that a vehicle is a suspected vehicle for smuggling when the smuggling prediction score is greater than or equal to a preset threshold.

[0007] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the above-described method for investigating smuggling of transit vehicles through the computer program.

[0008] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer program, and the computer program is configured to execute the above-described method for investigating smuggling of transit vehicles when it is run.

[0009] The technical solutions provided in this application embodiment may include the following beneficial effects: This application, through in-depth analysis of the characteristics of vehicle-to-border smuggling, mines factors related to vehicle-to-border smuggling from multi-source datasets and utilizes advanced machine learning techniques to construct a large-scale model, accurately outputting smuggling prediction scores. This innovative method provides customs anti-smuggling departments with accurate clues, significantly improving the success rate of detection. It also provides strong support for building an intelligent and efficient customs supervision system. Attached Figure Description

[0010] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of an inspection method for smuggling vehicles in transit according to an embodiment of this application; Figure 2 This is a flowchart of another method for investigating smuggling of transit vehicles according to an embodiment of this application; Figure 3 This is a schematic diagram of an inspection device for smuggling vehicles in transit, according to an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an optional electronic device according to an embodiment of this application. Detailed Implementation

[0011] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0012] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0013] The following is a detailed description of the inspection method for smuggling vehicles in transit according to embodiments of this application, with reference to the accompanying drawings. Figure 1 As shown, the method mainly includes the following steps: S101 acquires multi-source datasets and calculates vehicle transfer behavior factors based on the multi-source datasets.

[0014] In one implementation, a multi-source dataset is collected. This includes nationwide POI fence data, vehicle trajectory data, vehicle static data, vehicle route details, nationwide road data, vehicle transit declaration details, and vehicle trajectory patterns obtained based on the vehicle trajectory data. The vehicle route details include the vehicle's unique identifier (vid), license plate number, origin, destination, start time, and end time.

[0015] Furthermore, vehicle transfer behavior factors are calculated based on multi-source datasets. In one implementation, vehicle transfer behavior factors include: vehicle stop time, vehicle stop duration, vehicle stop latitude and longitude, vehicle stop point POI type, number of vehicles clustered together, vehicle origin POI type, whether the trajectory pattern of the clustered vehicles is abnormal, and the percentage of routes traveled.

[0016] Specifically, based on multi-source datasets, the transit vehicle's stopping time, transit vehicle's stopping duration, transit vehicle's stopping point latitude and longitude, and transit vehicle's stopping POI type are calculated.

[0017] In one implementation, detailed parking data of the vehicle is calculated based on the vehicle's trajectory data.

[0018] First, the trajectory data is cleaned to extract basic trajectory data such as the truck's license plate number, longitude, latitude, current timestamp, speed, direction, and road segment markers (links). Then, by analyzing points where the speed is 0, the vehicle's stopping location and stopping time are determined, thereby generating detailed stopping data for the vehicle.

[0019] Furthermore, based on the license plate number of the transit vehicle's customs declaration details, the parking data of the transit vehicle is obtained by associating the license plate number of the vehicle's parking details. Based on the parking data of the transit vehicle, the parking time, parking duration, and latitude and longitude of the transit vehicle's parking point during the transit period are determined.

[0020] First, key information such as license plate numbers is extracted from the detailed customs declaration data of transit vehicles. Then, based on the extracted license plate numbers, they are correlated with the vehicle parking details to match the corresponding parking records of the transit vehicles. In this way, the specific parking time, duration, and latitude and longitude information of each transit vehicle during its transit period can be determined. This process utilizes data association technology, combining timestamps and geographic coordinate data, to accurately track and analyze vehicle parking behavior.

[0021] Furthermore, the stopping points of transit and customs declaration vehicles with a stopping time greater than a preset first threshold and less than a preset second threshold are matched with national POI data to obtain the POI type of transit and customs declaration vehicles.

[0022] First, based on the stopping time of transit vehicles, stopping points with stopping times exceeding a first preset threshold and less than a second preset threshold are selected, thus identifying stopping points with reasonable stopping times. Then, the latitude and longitude information of these stopping points is matched with national POI (Point of Interest) data to determine the POI type corresponding to each stopping point. This allows for accurate identification of the specific location type where vehicles stop.

[0023] Furthermore, the percentage of trips on each route is calculated based on multi-source datasets.

[0024] First, based on the origin and destination of the transit declaration details of the transit vehicles, the origin and destination of the vehicle route details are associated, and the route details of all vehicles with the same origin and destination within a preset time period are obtained. Based on all the route details of vehicles with the same origin and destination, the total number of times the vehicles travel on the routes with the same origin and destination within the preset time period is counted.

[0025] In this embodiment, origin and destination information is extracted from the customs declaration details of transit vehicles. Then, this origin and destination information is correlated with the origin and destination in the vehicle route details, filtering out vehicle route details that have the same origin and destination as those in the customs declaration details within a preset time period. This process, through data matching and time filtering technology, can accurately identify all vehicle routes with the same origin and destination within a specific time period, such as within the past year.

[0026] After obtaining the vehicle route details that match the origin and destination in the customs declaration details, the total number of trips between the origin and destination for all vehicles within a preset time period is calculated for these detailed records. By analyzing the vehicle route detail data, trip records that meet the criteria are counted; that is, every trip between the origin and destination, regardless of whether it is a repeated trip by the same vehicle, is counted. In this way, the total number of trips between the origin and destination is obtained.

[0027] Furthermore, the target road segment to which the transfer vehicle stop belongs is determined, and based on the detailed information of the vehicle road segment associated with the target road segment, the vehicles traveling on the target road segment and their travel time can be found.

[0028] By associating the links (road segment identifiers) in the calculated details of vehicle transfer stops with the links in the vehicle road details, vehicles traveling on the links of the transfer stops and their travel times can be filtered out.

[0029] Based on vehicle identification and travel time, the route details of vehicles traveling on the target road segment and vehicles with the same origin and destination are associated, and vehicles traveling on the target road segment with the same origin and destination are filtered out; based on the travel route details of vehicles with the same origin and destination on the target road segment, the number of times vehicles with the same origin and destination travel on the target road segment within a preset time period is counted.

[0030] Understandably, after obtaining the vehicles and their travel times along the links connecting to border crossing stops, this data is further correlated with the route details for vehicles with the same origin and destination calculated in the previous step. The correlation criteria are that the vehicle ID (vid) is equal and the travel time falls within the start and end time range of the route details for vehicles with the same origin and destination. Through this correlation, it is possible to filter out the route details for all vehicles with the same origin and destination along the links connecting to border crossing stops within the past year.

[0031] After obtaining the detailed travel routes of vehicles with the same origin and destination on the target road segment, these detailed records are analyzed to count the number of times each vehicle travels on the target road segment. By counting travel records that meet the condition that each trip of a vehicle with the same origin and destination on the target road segment is counted, regardless of whether it is a repeated trip by the same vehicle, the total number of trips of vehicles with the same origin and destination on the target road segment is obtained.

[0032] Finally, the percentage of trips on the route is obtained based on the ratio of the number of trips on the target road segment to the total number of trips.

[0033] The higher the percentage of vehicles traveling on a given route, the more vehicles use the road section near the checkpoint, suggesting a lower probability of irregularities. Conversely, the lower the percentage of vehicles traveling on a given route, the fewer vehicles use the road section near the checkpoint, suggesting a more remote location and a higher probability of smuggling.

[0034] In one implementation, the calculation of the number of transit vehicles and the POI type of the origin of the transit vehicles is based on a multi-source dataset, including: acquiring transit vehicle parking data and detailed parking data; when the parking distance between other vehicles and transit vehicles is less than a preset distance threshold and the transit vehicle's parking time is greater than a preset time threshold, the vehicle is identified as a transit vehicle; the number of transit vehicles is counted to obtain the number of transit vehicles.

[0035] First, the stopping location (latitude and longitude) and stopping time information are extracted from the stopping point data of transit vehicles. Then, these stopping points are correlated with the stopping point details of all vehicles, filtering out vehicles whose distance from the transit vehicle's stopping point is less than a preset threshold (e.g., 500 meters) and whose cluster stopping time exceeds a preset time threshold (e.g., 45 minutes). In this way, information on vehicles clustered with transit vehicles is obtained, and the total number of clustered stopping vehicles is calculated based on these results, thus obtaining the number of transit and customs declaration clustered stopping vehicles. Information such as the license plate number, stopping point latitude and longitude, and stopping duration of clustered stopping vehicles is also obtained. Further, the stopping points of the clustered stopping vehicles before the preset cluster stopping time are obtained; these stopping points before the preset time are matched with national POI data to obtain the POI type of the clustered vehicle's origin.

[0036] First, the license plate numbers of vehicles clustered with transit vehicles are calculated, and these license plate numbers are associated with vehicle parking details to filter out the parking point information of these vehicles before they clustered together (e.g., 2 hours ago). Then, the latitude and longitude information of these parking points is matched with national POI (Point of Interest) data, and spatial analysis techniques are used to determine the POI type corresponding to each parking point. Finally, the POI type of the departure point of the vehicles clustered with transit vehicles is obtained.

[0037] In one implementation, calculating whether the trajectory pattern of clustered vehicles is abnormal based on a multi-source dataset includes: determining the trajectory pattern of clustered parked vehicles based on the trajectory data of clustered parked vehicles, wherein the trajectory pattern is a trajectory segment composed of a parking segment and a driving segment; and determining whether the trajectory pattern of clustered vehicles is abnormal based on the type of trajectory pattern.

[0038] Specifically, trajectory data related to clustered and parked vehicles is extracted. Based on this trajectory data, the trajectory pattern of each clustered and parked vehicle is determined. The trajectory pattern is represented by the shape of the trajectory segment composed of the parking segment and the driving segment, such as at least one of collinearity, small rectangle, large rectangle, regular rectangle, and elongated rectangle. Based on the type of trajectory pattern, it is determined whether the trajectory pattern of the clustered vehicles is abnormal. The loading and unloading points of the vehicles are determined based on the trajectory pattern; if it is a loading and unloading point, it is considered abnormal. This application's solution integrates multi-source datasets to deeply mine the behavioral feature data of transit vehicles, providing comprehensive and accurate data support for subsequent model predictions.

[0039] S102 constructs a training dataset, and based on the training dataset and the LightGBM model, trains a smuggling prediction model.

[0040] In an exemplary embodiment, 800 vehicles are selected as the training dataset and 200 vehicles are selected as the test dataset. The vehicle turning behavior factors in step S101 are extracted and weighted according to the preset weights. These weighted factors are then input into the LightGBM algorithm model for model parameter training.

[0041] During the model tuning phase, a grid search method is used to optimize the hyperparameters. By traversing all possible combinations of values ​​within a specified range, the optimal hyperparameter configuration is found to achieve the best model performance. Finally, the optimized model is deployed and used for inference to achieve accurate prediction of the behavior of vehicles turning around.

[0042] The LightGBM model is an efficient machine learning algorithm based on the gradient boosting framework, particularly suitable for handling large-scale datasets. Its model structure is based on decision trees, building a powerful predictive model by combining multiple weak learners (typically decision trees). Each decision tree consists of multiple nodes, including a root node, internal nodes, and leaf nodes.

[0043] Compared to traditional gradient boosting algorithms like XGBoost, the LightGBM model offers significant advantages. It employs a histogram-based algorithm and a leaf-wise growth strategy, greatly improving training speed and efficiency. Simultaneously, it achieves excellent prediction accuracy while maintaining high efficiency. Furthermore, LightGBM provides a clear ranking of feature importance, enhancing model interpretability to aid decision-making, and supports parallel and distributed learning, demonstrating strong scalability and the ability to efficiently handle massive amounts of data.

[0044] S103 inputs vehicle transfer behavior factors into the smuggling prediction model and outputs a smuggling prediction score.

[0045] In one implementation, the vehicle transfer behavior factor is input into the smuggling prediction model, and the smuggling prediction score is output. This includes: weighting the vehicle transfer behavior factor based on preset weights to obtain weighted vehicle transfer behavior data; inputting the weighted vehicle transfer behavior data into the smuggling prediction model to obtain the smuggling prediction score; and dynamically adjusting the weights of the vehicle transfer behavior factor based on the inspection results.

[0046] First, vehicle border crossing behavior factors are extracted from multi-source datasets. Then, these behavior factors are weighted according to pre-defined weights to obtain weighted vehicle border crossing behavior data. This weighted data is input into a smuggling prediction model, which analyzes this data to generate a smuggling prediction score to assess whether a vehicle poses a smuggling risk.

[0047] Finally, the weights of the vehicle transit behavior factors are dynamically adjusted based on the inspection results. If the inspection results show a deviation between the model's predictions and the actual situation, the weights of the corresponding factors will be adjusted according to the deviation to optimize the model's predictive performance. This process, through a data-driven weight adjustment mechanism, ensures that the model can continuously adapt to new smuggling methods and business needs, improving the accuracy and reliability of predictions.

[0048] S104 determines a vehicle as a suspected smuggling vehicle when the smuggling prediction score is greater than or equal to a preset threshold.

[0049] In the analysis of vehicle transit behavior, by inputting vehicle transit behavior factors into the smuggling prediction model, the model outputs a smuggling prediction score based on these factors. When the score is greater than or equal to a preset threshold, the system marks the vehicle as a suspected smuggling vehicle.

[0050] Furthermore, if the score is greater than or equal to a preset threshold, the prediction result is pushed to the customs anti-smuggling department. Real-time alerts are issued at the customs anti-smuggling department's terminal, and the pushed prediction results include basic vehicle information such as license plate number, owner information, driving trajectory, and key data such as the prediction score. This information provides regulatory authorities with clear clues for investigation, helping them quickly locate and handle suspicious vehicles, improving the efficiency and accuracy of investigations, and effectively combating smuggling activities.

[0051] To facilitate understanding of the methods in the embodiments of this application, the following description is provided in conjunction with the appendix. Figure 2 Further description.

[0052] like Figure 2As shown, the process begins with data collection, including information such as stop time, stop duration, stop latitude and longitude, POI type, number of vehicles clustered, origin POI type, trajectory pattern, and percentage of trips. Next, a training and testing set is constructed, and this data is used to build the LightGBM model. The model outputs a smuggling prediction score; if the score is greater than or equal to a set threshold, the vehicle is identified as suspicious; otherwise, it is identified as a legitimate vehicle. This forms a complete vehicle smuggling prediction system.

[0053] This application integrates multi-dimensional behavioral characteristics such as stopping time, location, POI type, and trajectory pattern. The system can accurately depict vehicle transit behavior, significantly improving the ability to identify covert and organized smuggling activities and providing more precise clues for customs anti-smuggling departments. Secondly, based on the interpretability of the LightGBM model, the system can analyze the impact of various factors (such as the number of clustered vehicles and the proportion of vehicles traveling along the route) on the risk score, assisting customs officers in understanding the model's decision-making basis. Finally, this solution transforms the traditional analysis mode relying on manual experience into an automated and intelligent early warning process, significantly shortening the time for clue discovery, reducing the risk of missed detections, and reducing interference with legally passing vehicles by accurately identifying suspicious targets.

[0054] According to another aspect of the embodiments of this application, a device for inspecting smuggled goods in transit vehicles is also provided for implementing the above-described method for inspecting smuggled goods in transit vehicles. For example... Figure 3 As shown, the device includes: Data processing module 301 is used to acquire multi-source datasets and calculate vehicle transfer behavior factors based on the multi-source datasets; Model training module 302 is used to build a training dataset and train a smuggling prediction model based on the training dataset and the LightGBM model. Model prediction module 303 is used to input vehicle transfer behavior factors into the smuggling prediction model and output a smuggling prediction score; The result output module 304 is used to determine that a vehicle is a suspected vehicle for smuggling when the smuggling prediction score is greater than or equal to a preset threshold.

[0055] It should be noted that the above-described embodiments of the inspection device for smuggling transit vehicles, when executing the inspection method for smuggling transit vehicles, are only illustrative examples of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the inspection device for smuggling transit vehicles and the inspection method embodiments provided in the above embodiments belong to the same concept, and their implementation process is detailed in the method embodiments, which will not be repeated here.

[0056] According to another aspect of the embodiments of this application, an electronic device corresponding to the inspection method for smuggling transit vehicles provided in the foregoing embodiments is also provided, so as to execute the above-mentioned inspection method for smuggling transit vehicles.

[0057] Please refer to Figure 4 This illustrates a schematic diagram of an electronic device provided by some embodiments of this application. For example... Figure 4 As shown, the electronic device includes: a processor 400, a memory 401, a bus 402, and a communication interface 403. The processor 400, the communication interface 403, and the memory 401 are connected via the bus 402. The memory 401 stores a computer program that can run on the processor 400. When the processor 400 runs the computer program, it executes the inspection method for smuggling of transit vehicles provided in any of the foregoing embodiments of this application.

[0058] The memory 401 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 403 (which can be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.

[0059] Bus 402 can be an ISA bus, PCI bus, or EISA bus, etc. Buses can be divided into address buses, data buses, control buses, etc. Memory 401 is used to store programs. After receiving execution instructions, processor 400 executes the program. The inspection method for smuggling of transit vehicles disclosed in any of the aforementioned embodiments of this application can be applied to processor 400, or implemented by processor 400.

[0060] The processor 400 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 400 or by instructions in software form. The processor 400 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 401. The processor 400 reads the information in memory 401 and, in conjunction with its hardware, completes the steps of the above method.

[0061] The electronic device provided in this application embodiment and the inspection method for smuggling of transit vehicles provided in this application embodiment are based on the same inventive concept and have the same beneficial effects as the methods they adopt, operate or implement.

[0062] According to another aspect of the present application, a computer-readable storage medium corresponding to the inspection method for smuggling transit vehicles provided in the foregoing embodiments is also provided, wherein a computer program (i.e., a program product) is stored thereon, and when the computer program is run by a processor, it executes the inspection method for smuggling transit vehicles provided in any of the foregoing embodiments.

[0063] It should be noted that examples of computer-readable storage media may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0064] The computer-readable storage medium provided in the above embodiments of this application and the inspection method for smuggling of transit vehicles provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0065] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0066] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A method for inspecting smuggling by transit vehicles, characterized in that, include: Obtain a multi-source dataset and calculate vehicle checkpoint behavior factors based on the multi-source dataset; Construct a training dataset, and train a smuggling prediction model based on the training dataset and the LightGBM model; The vehicle transfer behavior factor is input into the smuggling prediction model, and the smuggling prediction score is output. When the smuggling prediction score is greater than or equal to a preset threshold, the vehicle is identified as a suspected smuggling vehicle.

2. The method according to claim 1, characterized in that, The vehicle checkpoint behavior factor is calculated based on the multi-source dataset, including: Based on the multi-source dataset, the following parameters are calculated: transit vehicle stopping time, transit vehicle stopping duration, transit vehicle stopping point latitude and longitude, transit vehicle stopping POI type, number of transit vehicles clustered together, origin POI type of clustered vehicles, whether the trajectory pattern of clustered vehicles is abnormal, and the percentage of routes traveled.

3. The method according to claim 2, characterized in that, Based on the multi-source dataset, the following parameters are calculated: transit vehicle stopping time, transit vehicle stopping duration, transit vehicle stopping point latitude and longitude, and transit vehicle stopping POI type: Calculate detailed vehicle parking data based on vehicle trajectory data; Based on the license plate number of the transit vehicle's customs declaration details, the parking data of the transit vehicle is obtained by associating the license plate number of the vehicle parking details. Based on the parking data of the transit vehicle, the parking time, parking duration, and latitude and longitude of the transit vehicle's parking point during the transit period are determined. The stopping points of the transit and customs declaration vehicles whose stopping time is greater than a preset first threshold and less than a preset second threshold are matched with national POI data to obtain the stopping POI type of the transit and customs declaration vehicles.

4. The method according to claim 2, characterized in that, The percentage of trips on the route is calculated based on the multi-source dataset, including: Based on the origin and destination of the transit vehicle's customs declaration details, associate the origin and destination of the vehicle's route details, and obtain the route details of all vehicles that have the same origin and destination within a preset time period. Based on the route details of all vehicles with the same origin and destination, count the total number of times vehicles travel on routes with the same origin and destination within a preset time period. Determine the target road segment to which the transfer vehicle stop belongs, and based on the road segment details associated with the target road segment, query the vehicles traveling on the target road segment and their travel time; Based on vehicle identification and travel time, the route details of vehicles traveling on the target road segment and vehicles with the same origin and destination are associated, and vehicles traveling on the target road segment with the same origin and destination are filtered out; based on the travel route details of vehicles with the same origin and destination on the target road segment, the number of times vehicles with the same origin and destination travel on the target road segment within a preset time period is counted. The percentage of vehicle trips on the target road segment is obtained by comparing the number of trips on that segment with the total number of trips on that route.

5. The method according to claim 2, characterized in that, The number of transit vehicles clustered and stopped at customs based on multi-source datasets, and the POI type of the origin of the clustered vehicles, including: Obtain parking data and detailed parking information for transit vehicles; When the stopping distance between other vehicles and transfer vehicles is less than a preset distance threshold, and the stopping time with transfer vehicles is greater than a preset time threshold, the vehicle is determined to be a transfer vehicle. The number of vehicles that gather and stop is counted to obtain the number of vehicles that gather and stop for customs declaration; the stopping point of the vehicles that gather and stop before the preset gathering and stopping time is obtained; The stops before the preset time period are matched with national POI data to obtain the POI type of the origin of the aggregated vehicles.

6. The method according to claim 5, characterized in that, Calculate whether the aggregated vehicle trajectory patterns are abnormal based on multi-source datasets, including: Based on the trajectory data of the clustered parked vehicles, the trajectory pattern of the clustered parked vehicles is determined, and the trajectory pattern is a trajectory segment composed of a parking segment and a driving segment. Based on the type of trajectory pattern, determine whether the trajectory pattern of the aggregated vehicles is abnormal.

7. The method according to claim 1, characterized in that, The vehicle transit behavior factors are input into the smuggling prediction model, and the smuggling prediction score is output, including: The vehicle transfer behavior factors are weighted based on preset weights to obtain weighted vehicle transfer behavior data. The weighted vehicle transit behavior data is input into the smuggling prediction model to obtain a smuggling prediction score; The weights of the vehicle transfer behavior factors are dynamically adjusted based on the inspection results.

8. A device for detecting smuggling in transit vehicles, characterized in that, include: The data processing module is used to acquire multi-source datasets and calculate vehicle transfer behavior factors based on the multi-source datasets. The model training module is used to construct a training dataset and train a smuggling prediction model based on the training dataset and the LightGBM model. The model prediction module is used to input the vehicle transfer behavior factors into the smuggling prediction model and output a smuggling prediction score. The result output module is used to determine that a vehicle is a suspected vehicle for smuggling when the smuggling prediction score is greater than or equal to a preset threshold.

9. An electronic device, characterized in that, It includes a processor and a memory storing program instructions, the processor being configured to, when executing the program instructions, perform the inspection method for smuggling of transit vehicles as described in any one of claims 1 to 7.

10. A computer-readable medium, characterized in that, It stores computer-readable instructions that are executed by a processor to implement a method for investigating smuggling of transit vehicles as described in any one of claims 1 to 7.

Images