Counterfeated green channel fee evasion vehicle identification method and device, medium and equipment

Abstract

The invention discloses a counterfeit green channel fee evasion vehicle identification method and device, a medium and equipment. The method comprises the steps of obtaining first vehicle data of a plurality of to-be-identified vehicles within a preset historical time period from a current moment; constructing a feature factor sequence of each to-be-identified vehicle according to the first vehicle data; inputting the feature factor sequence of each to-be-identified vehicle into a pre-trained green channel vehicle identification model, and outputting a counterfeit green channel probability of each to-be-identified vehicle; obtaining to-be-identified vehicles corresponding to the counterfeit green channel probability in a preset probability interval, and taking the to-be-identified vehicles as a target vehicle set; and performing clustering processing on the target vehicle set, and taking the target vehicles of the same category as a group of counterfeit green channel fee evasion vehicles. Therefore, by adopting the embodiment of the invention, the efficiency of an automatic identification mode is high, real fee evasion vehicles are not easy to miss, and the fee evasion gang can be accurately identified from mass data.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method, device, medium and equipment for identifying vehicles that impersonate green passes and evade tolls. Background Art

[0002] In the field of fresh agricultural product transportation, in order to improve transportation efficiency, reduce losses, and ensure the supply of the people's "vegetable basket", the green channel policy can be implemented to exempt vehicles transporting fresh agricultural products from road tolls. This policy has greatly promoted the circulation of agricultural products, but it has also brought about the problem of some lawless elements taking advantage of policy loopholes to evade fees.

[0003] In the related technologies, the existing technologies mainly rely on manual inspection and random sampling to identify vehicles with fake green pass. Manual inspection is inefficient; random sampling lacks specificity and is prone to miss real evading vehicles; lack of effective data analysis means, it is difficult to accurately identify evasion gangs from massive data. Summary of the invention

[0004] The embodiments of the present application provide a method, device, medium and equipment for identifying vehicles that evade tolls by impersonating Green Pass. In order to have a basic understanding of some aspects of the disclosed embodiments, a simple summary is given below. This summary is not a general review, nor is it intended to identify key / important components or to describe the scope of protection of these embodiments. Its only purpose is to present some concepts in a simple form as a preface to the detailed description that follows.

[0005] In a first aspect, an embodiment of the present application provides a method for identifying a vehicle that impersonates a green pass and evades tolls, the method comprising:

[0006] Acquire first vehicle data of a plurality of vehicles to be identified within a preset historical time period from the current moment;

[0007] Constructing a characteristic factor sequence of each vehicle to be identified according to the first vehicle data;

[0008] Input the characteristic factor sequence of each vehicle to be identified into the pre-trained green pass vehicle identification model, and output the fake green pass probability of each vehicle to be identified;

[0009] Obtain the vehicles to be identified corresponding to the counterfeit green pass probability within the preset probability interval as the target vehicle set;

[0010] The target vehicle set is clustered and the target vehicles of the same category are regarded as a group of toll evasion vehicles pretending to be green pass.

[0011] Optionally, follow the steps below to generate a pre-trained green pass vehicle recognition model, including:

[0012] Obtain the second vehicle data of each toll-evading vehicle in the preset period among the toll-evading vehicles that have historically pretended to be green passes;

[0013] constructing a characteristic factor sequence of each toll-evading vehicle according to the second vehicle data;

[0014] The probability of fake green pass is set for each characteristic factor sequence of toll-evading vehicles as the weight of the second characteristic factor sequence to obtain multiple model training samples;

[0015] A decision tree algorithm is used to create a green pass vehicle recognition model;

[0016] According to each model training sample, machine learning is performed on the green pass vehicle recognition model to optimize the model parameters of the green pass vehicle recognition model and obtain a pre-trained green pass vehicle recognition model.

[0017] Optionally, the vehicle data includes vehicle parking data, vehicle static data and vehicle trajectory data; based on the vehicle data, a characteristic factor sequence of the vehicle is constructed, including:

[0018] According to the vehicle parking data, construct the vehicle's loading point POI label and whether the loading point has the green pass attribute;

[0019] Constructing the attributes of the vehicle's highway mileage and whether it was driven at night based on the vehicle trajectory data;

[0020] According to the static data of the vehicle, the vehicle model, the number of vehicles of the transport company to which it belongs, and the rated load are constructed;

[0021] The vehicle's loading point POI label and whether the loading point has the green pass attribute are classified with the vehicle's highway mileage, vehicle model, number of vehicles of the transport company to which it belongs, rated load, and whether it is driven at night to obtain the vehicle's characteristic factor sequence.

[0022] Optionally, based on the vehicle stop data, construct the vehicle's loading point POI label and whether the loading point has a green pass attribute, including:

[0023] Obtain the stop time of each initial stop point in the vehicle stop data;

[0024] Determine whether the stop time of each initial stop point is greater than a preset time threshold;

[0025] Extracting initial stop points whose stop duration is greater than a preset stop duration threshold as multiple candidate stop points;

[0026] Get the stop type of each alternative stop point;

[0027] Determine whether the stop type of each candidate stop point belongs to the preset POI type;

[0028] Eliminate the candidate stop points whose stop types belong to the preset POI type from the multiple candidate stop points, and obtain multiple valid stop points for the vehicle;

[0029] According to the multiple valid stopping points of the vehicle, the POI label of the loading point of the vehicle and whether the loading point has the green pass attribute are analyzed.

[0030] Optionally, based on multiple valid stops of the vehicle, the POI tag of the vehicle's loading point and whether the loading point has a green pass attribute are analyzed, including:

[0031] From multiple valid stopping points of the vehicle, multiple candidate stopping points with the shortest distances from the highway entrance are obtained, and the absolute value of the distance difference between two adjacent candidate stopping points from the multiple candidate stopping points and the highway entrance is less than a preset distance threshold;

[0032] Determine whether each candidate stop point is within the electronic fence used to characterize the green pass POI type;

[0033] If so, bind each candidate stop point to the green pass POI type corresponding to the electronic fence where it is located, obtain the POI label of each candidate stop point, use it as the loading point POI label of the vehicle, and determine that each candidate stop point has the green pass attribute, and obtain the loading point of the vehicle has the green pass attribute;

[0034] If not, and the stop type of each candidate stop point does not belong to the national POI type in the geographic information system, and there is no national POI type within the preset range around each candidate stop point, it is determined that each candidate stop point has a green pass attribute, and the loading point of the vehicle has a green pass attribute;

[0035] If not and the stop type of each candidate stop point belongs to the national POI type in the geographic information system, the attribute of each candidate stop point is determined to be non-green pass, and the loading point of the vehicle does not have the green pass attribute.

[0036] Optionally, based on the vehicle trajectory data, construct the vehicle's highway mileage and whether it is driven at night, including:

[0037] From the vehicle trajectory data, obtain the driving trajectory between the vehicle's entry point into the highway and the current exit point;

[0038] Calculate the vehicle's high-speed mileage based on the driving trajectory;

[0039] Obtaining the driving time period of the vehicle from the vehicle trajectory data;

[0040] According to the driving time period of the vehicle, it is determined whether the vehicle is driving at night, and an attribute of whether the vehicle is driving at night is obtained.

[0041] Optionally, cluster the target vehicle set and treat the target vehicles of the same category as a group of toll evasion vehicles pretending to be green pass, including:

[0042] Obtain the parking data of each target vehicle in the target vehicle set;

[0043] According to the stop data of each target vehicle, the candidate stop points of each target vehicle are analyzed; the candidate stop points of each target vehicle are obtained by screening based on a preset time threshold and a preset POI type;

[0044] The k-means algorithm and the distance between adjacent candidate stops are used to cluster the candidate stops of all target vehicles to obtain the vehicle clustering area of ​​each target vehicle;

[0045] Use the vehicle gathering area of ​​each target vehicle to associate the national POI type in the geographic information system to obtain the POI attribute of the vehicle gathering area of ​​each target vehicle;

[0046] Based on the existence of the same vehicle gathering areas, POI attributes of the vehicle gathering areas, and the company to which the vehicles belong, the k-nearest neighbor algorithm is used for classification, and the classification results with the same characteristics after classification are divided into a group, obtaining multiple groups of vehicles that pretend to be green passes and evade tolls.

[0047] In a second aspect, an embodiment of the present application provides a device for identifying a vehicle that impersonates a green pass and evades tolls, the device comprising:

[0048] A data acquisition module, used to acquire first vehicle data of a plurality of vehicles to be identified within a preset historical time period from the current moment;

[0049] A characteristic factor construction module, used to construct a characteristic factor sequence of each vehicle to be identified according to the first vehicle data;

[0050] A fake green pass probability output module is used to input the characteristic factor sequence of each vehicle to be identified into a pre-trained green pass vehicle identification model, and output the fake green pass probability of each vehicle to be identified;

[0051] A vehicle acquisition module is used to acquire vehicles to be identified that correspond to the probability of counterfeit green pass within a preset probability interval as a target vehicle set;

[0052] The vehicle clustering module is used to cluster the target vehicle set and treat the target vehicles of the same category as a group of vehicles that pretend to be green pass and evade tolls.

[0053] In a third aspect, an embodiment of the present application provides a computer storage medium, which stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the above-mentioned method steps.

[0054] In a fourth aspect, an embodiment of the present application provides a device, which may include: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the above-mentioned method steps.

[0055] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0056] In the embodiment of the present application, by adopting an automated data acquisition and feature construction method, combined with a pre-trained green pass vehicle identification model, the probability of each vehicle to be identified as a fake green pass can be accurately calculated. Feature construction is based on vehicle parking data, static information, and trajectory data to obtain feature information that characterizes vehicle transportation. Through this feature information, high-risk evasion suspects can be effectively screened out, thereby improving the efficiency of inspection in a targeted manner. Ultimately, through cluster analysis, evasion gangs with similar behavior patterns can be identified, achieving a precise strike against evasion behavior. This application not only improves the accuracy and efficiency of identifying fake green pass vehicles, but also enhances the precise identification of evasion gangs, ensuring the fair implementation of the green channel policy.

[0057] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0059] Figure 1 It is a flowchart of a method for identifying a vehicle that impersonates a green pass and evades tolls provided in an embodiment of the present application;

[0060] Figure 2 It is a flowchart of a model training method for a green pass vehicle recognition model provided by the present application;

[0061] Figure 3 This application provides a data representation of the probability of each vehicle to be identified as a fake green pass;

[0062] Figure 4 This is a schematic diagram of the structure of a device for identifying a vehicle that impersonates a green pass and evades tolls, provided in an embodiment of the present application;

[0063] Figure 5 It is a structural schematic diagram of a device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] The following description and the accompanying drawings sufficiently illustrate specific embodiments of the present application to enable those skilled in the art to practice them.

[0065] It should be clear that the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0066] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Instead, they are only examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims.

[0067] In the description of the present application, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances. In addition, in the description of the present application, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are in an "or" relationship.

[0068] The present application provides a method, device, medium and equipment for identifying toll evasion vehicles that are posing as green pass vehicles, so as to solve the problems existing in the above-mentioned related technical problems. In the embodiment of the present application, by adopting an automated data acquisition and feature construction method, combined with a pre-trained green pass vehicle recognition model, the probability of each vehicle to be identified as a fake green pass can be accurately calculated. The feature construction is based on the vehicle parking data, static information, and trajectory data to obtain characteristic information that characterizes vehicle transportation. Through this characteristic information, high-risk toll evasion suspects can be effectively screened out, thereby improving the efficiency of inspection in a targeted manner. Finally, through cluster analysis, toll evasion gangs with similar behavior patterns can be identified, and accurate strikes on toll evasion behaviors can be achieved. Through this application, not only the accuracy and efficiency of identifying fake green pass vehicles are improved, but also the accurate identification of toll evasion gangs is enhanced, ensuring the fair implementation of the green channel policy. The following uses an exemplary embodiment for detailed description.

[0069] The following will be combined with the attached Figure 1 -Attached Figure 3 , a method for identifying a vehicle that evades tolls by pretending to be a green pass provided by an embodiment of the present application is introduced in detail. The method can be implemented by a computer program and can be run on a device for identifying a vehicle that evades tolls by pretending to be a green pass based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application.

[0070] Please refer to Figure 1 , which is a schematic flowchart of a method for identifying toll-evading vehicles posing as green channel vehicles provided by an embodiment of the present application. As Figure 1 shown, the method of the embodiment of the present application may include the following steps:

[0071] S101, Obtain first vehicle data of multiple vehicles to be identified within a preset historical time period from the current moment;

[0072] Among them, the multiple vehicles to be identified are historical vehicles recorded by an identification device in a database, such as ten thousand historical vehicles. The preset historical time period refers to a time range set for analyzing vehicle behavior, and the data during this period will be used for identification and analysis. The first vehicle data includes vehicle parking data, vehicle static data, and vehicle trajectory data. Vehicle parking data refers to the relevant data of the parking points of the vehicle within the preset historical time period, including information such as the parking time, location, and duration. Vehicle static data refers to the basic information of the vehicle, such as the vehicle type, license plate number, affiliated company, load capacity, etc. Vehicle trajectory data is data used to represent the trajectory position points during the vehicle's driving process.

[0073] In some embodiments, for example, when the regulatory department wants to identify a possible toll-evading gang posing as green channel vehicles on a certain highway, a preset historical time period can be set at this time, such as the most recent 72 hours. The system automatically collects the parking data and vehicle static data of all vehicles within the most recent 30 days. For example, a vehicle parked at Service Area No. H123 for 45 minutes at 6:00 am on August 28, 2024, can be used as the first parking data of this vehicle; the system also obtains the static information of this vehicle, including the license plate number "Beijing A12345", the vehicle type is a closed van, the affiliated company is "Lucky Transport Co., Ltd.", and the rated load capacity is 10 tons, which can be used as the vehicle static data of this vehicle.

[0074] S102, Construct a characteristic factor sequence for each vehicle to be identified according to the first vehicle data;

[0075] Among them, the first vehicle data includes vehicle parking data, vehicle static data, and vehicle trajectory data. For the sake of simplified description, the vehicle to be identified can be directly referred to as a vehicle.

[0076] In some embodiments of the present application, the specific process of constructing a characteristic factor sequence for each vehicle to be identified based on vehicle data includes: constructing the vehicle's loading point POI label and whether the loading point has a green pass attribute based on the vehicle stop data; constructing the vehicle's highway mileage and whether it is driven at night based on the vehicle trajectory data; constructing the vehicle's model, the number of vehicles of the transportation company to which it belongs, and the rated load based on the vehicle's static data; classifying the vehicle's loading point POI label and whether the loading point has a green pass attribute with the vehicle's highway mileage, model, number of vehicles of the transportation company to which it belongs, rated load, and whether it is driven at night to obtain the vehicle's characteristic factor sequence.

[0077] Among them, the loading point POI label refers to the category or attribute of the loading point obtained by matching the geographical location information of the vehicle loading point with the national POI data. Whether the loading point has the green channel attribute refers to whether the loading point is located in the fresh agricultural product transportation area or farmland area stipulated by the green channel policy. The mileage of the highway section refers to the mileage traveled by the vehicle on the highway. The vehicle model refers to the specification type of the vehicle, such as closed vans, container trucks, refrigerated trucks, etc. The number of vehicles of the transport company refers to the total number of vehicles registered under the name of the transport company to which the vehicle belongs. The rated load refers to the standard load of the vehicle as specified by the design and regulations. Whether to drive at night refers to whether the vehicle chooses to drive at night or in the early morning.

[0078] In the embodiment of the present application, by carefully analyzing the vehicle parking data, vehicle static data, and vehicle trajectory data, a detailed characteristic factor sequence of each vehicle can be constructed. These factor sequences not only include whether the vehicle's loading point meets the green pass attribute, but also cover the vehicle's driving habits, the size of the company to which it belongs, the vehicle's load capacity, and its driving time preference. By classifying these characteristic factor sequences, vehicles with a tendency to evade fees can be identified more accurately, thereby improving the pertinence and efficiency of the audit.

[0079] In some embodiments of the present application, the specific process of constructing a loading point POI label of a vehicle and whether the loading point has a green pass attribute based on vehicle stop data includes: obtaining the stop time of each initial stop point in the vehicle stop data; judging whether the stop time of each initial stop point is greater than a preset time threshold; extracting initial stop points whose stop time is greater than the preset time threshold as multiple candidate stop points; obtaining the stop type of each candidate stop point; judging whether the stop type of each candidate stop point belongs to a preset POI type; from multiple candidate stop points, eliminating the alternative stop points whose stop type belongs to the preset POI type to obtain multiple valid stop points for the vehicle; and analyzing the loading point POI label of the vehicle and whether the loading point has a green pass attribute based on the multiple valid stop points of the vehicle.

[0080] For example, obtain the vehicle parking data of each vehicle to be recognized within the most recent N hours, and take the parking points with a parking duration > M minutes and exclude the parking points of the specified POI types (such as gas stations, service areas, toll stations, highway rest areas, ferry crossings) as valid parking points.

[0081] In some embodiments of the present application, the specific process of analyzing the loading point POI label of the vehicle and whether the loading point has the green channel attribute based on multiple valid parking points of the vehicle includes: obtaining, from the multiple valid parking points of the vehicle, multiple candidate parking points with the shortest distance from the highway entrance, and the absolute value of the distance difference between adjacent two candidate parking points among the multiple candidate parking points from the highway entrance is less than a preset distance threshold; determining whether each candidate parking point is within the electronic fence for characterizing the green channel POI type; if so, binding each candidate parking point to the corresponding green channel POI type of the electronic fence where it is located to obtain the POI label of each candidate parking point as the loading point POI label of the vehicle, and determining that each candidate parking point has the green channel attribute to obtain that the loading point of the vehicle has the green channel attribute; if not and the parking type of each candidate parking point does not belong to the national POI type in the geographic information system and there is no national POI type coverage within the preset range around each candidate parking point, determining that each candidate parking point has the green channel attribute to obtain that the loading point of the vehicle has the green channel attribute; if not and the parking type of each candidate parking point belongs to the national POI type in the geographic information system, determining that each candidate parking point has a non-green channel attribute to obtain that the loading point of the vehicle does not have the green channel attribute.

[0082] For example, after obtaining the valid parking points of the vehicle, take the parking point 1, parking point 2 (stop 1 - stop 2 < r km), and parking point 3 (stop 1 - stop 2 < r km, stop 2 - stop 3 < r km) with the shortest distance from the highway entrance as candidate parking points, determine whether each candidate parking point is within the electronic fence for characterizing the green channel POI type, if not within the electronic fence, further determine whether each candidate parking point belongs to the national POI type in the geographic information system, if not and there is no POI type coverage within the preset range r2 meters around the candidate parking point, it is considered to be a farmland area, and at this time, it can be considered that the vehicle has the green channel attribute, otherwise it has a non-green channel attribute.

[0083] In some embodiments of the present application, the specific process of constructing the driving mileage of the vehicle on the highway section and the attribute of whether it is driving at night based on the vehicle trajectory data includes: obtaining, from the vehicle trajectory data, the driving trajectory between the entry point where the vehicle enters the highway and the current exit point; calculating the driving mileage of the vehicle on the highway section according to the driving trajectory; obtaining, from the vehicle trajectory data, the driving time period of the vehicle; and determining whether the vehicle is driving at night according to the driving time period of the vehicle to obtain the attribute of whether the vehicle is driving at night.

[0084] S103, inputting the characteristic factor sequence of each vehicle to be identified into a pre-trained green pass vehicle identification model, and outputting the counterfeit green pass probability of each vehicle to be identified;

[0085] Among them, the pre-trained Green Pass vehicle identification model refers to a machine learning model that has been trained using a large amount of known data (including data of real Green Pass vehicles and fake Green Pass vehicles). The model can predict whether a vehicle is likely to be a fake Green Pass based on the input feature factor sequence. The fake Green Pass probability refers to the probability that the vehicle to be identified is a fake Green Pass calculated by the model based on the input feature factor sequence. The probability is a value between 0 and 1. The higher the value, the greater the possibility that the vehicle is a fake Green Pass.

[0086] In some embodiments of the present application, a pre-trained green pass vehicle identification model is generated according to the following steps, specifically including: obtaining second vehicle data of each toll-evading vehicle that has historically impersonated a green pass within a preset period; constructing a characteristic factor sequence for each toll-evading vehicle based on the second vehicle data; setting a probability of counterfeiting a green pass for each toll-evading vehicle's characteristic factor sequence as a weight of the second characteristic factor sequence, and obtaining multiple model training samples; using a decision tree algorithm to create a green pass vehicle identification model; and performing machine learning on the green pass vehicle identification model based on each model training sample to optimize the model parameters of the green pass vehicle identification model, and obtaining a pre-trained green pass vehicle identification model.

[0087] The second vehicle data also includes vehicle parking data, vehicle static data and vehicle trajectory data.

[0088] It should be noted that the specific process of constructing the characteristic factor sequence of each toll-evading vehicle based on the second vehicle data is the same as the specific process of constructing the characteristic factor sequence of the vehicle based on the vehicle data in step S102. For details, please refer to step S102 and will not be repeated here.

[0089] In an embodiment of the present application, an accurate green pass vehicle recognition model is created. The model is trained using data on historical toll evasion vehicles through a machine learning algorithm to learn the characteristics of toll evasion behavior, thereby optimizing model parameters and improving recognition accuracy.

[0090] The characteristic factor sequence of each vehicle to be identified is input into the pre-trained green pass vehicle identification model, and the counterfeit green pass probability of each vehicle to be identified is output, for example Figure 2 As shown, it includes characteristic factors, vehicle ID and probability of counterfeiting green pass. The characteristic factors include vehicle type, rated load, number of vehicles of the transport company, whether the loading point is green pass, mileage of the highway section, and whether it is driven at night. The vehicle ID includes vehicle 001-vehicle 005. Different values ​​can represent the specific meaning of the characteristic factors.

[0091] S104, obtaining vehicles to be identified corresponding to the counterfeit green pass probability within a preset probability interval as a target vehicle set;

[0092] The preset probability interval refers to a probability range set in the Green Pass vehicle identification model to screen out vehicles with a high suspicion of evading charges by impersonating the Green Pass. The fake Green Pass probability refers to the probability value calculated by the Green Pass vehicle identification model based on the characteristic factor sequence of the vehicle to be identified, indicating the possibility of the vehicle evading charges by impersonating the Green Pass. The target vehicle set refers to the set of vehicles with a high suspicion of evading charges that are screened out based on the fake Green Pass probability.

[0093] In some embodiments, it is assumed that the regulatory authorities set a preset probability interval of 0.7 to 1.0, which means that only vehicles with a probability of counterfeiting Green Pass greater than or equal to 70% will be considered high-risk vehicles. The model output results show that the probability of counterfeiting Green Pass for 10 vehicles falls within the range of 0.7 to 1.0. For example, a truck with a license plate number of "Yu A88888" has a probability of counterfeiting Green Pass of 0.78; another truck with a license plate number of "Su B12345" has a probability of counterfeiting Green Pass of 0.85. These two vehicles and the other 8 vehicles together constitute the target vehicle set.

[0094] S105, clustering the target vehicle set, and treating the target vehicles of the same category as a group of toll evasion vehicles that pretend to be green pass.

[0095] In some embodiments of the present application, a target vehicle set is clustered, and the specific process of treating the target vehicles of the same category as a group of counterfeit green pass toll evading vehicles includes: obtaining the stop data of each target vehicle in the target vehicle set; analyzing the candidate stop points of each target vehicle based on the stop data of each target vehicle; the candidate stop points of each target vehicle are obtained by screening based on a preset time threshold and a preset POI type; clustering the candidate stop points of all target vehicles using the k-means algorithm and the distance between adjacent candidate stop points to obtain the vehicle gathering area of ​​each target vehicle; using the vehicle gathering area of ​​each target vehicle to associate the national POI type in the geographic information system to obtain the POI attributes of the vehicle gathering area of ​​each target vehicle; based on the existence of the same vehicle gathering area, the POI attributes of the vehicle gathering area and the company to which the vehicle belongs, classification is performed using the k nearest neighbor algorithm, and the classification results with the same characteristics after classification are divided into a group to obtain multiple groups of counterfeit green pass toll evading vehicles.

[0096] Among them, the k-means algorithm is a commonly used clustering algorithm, which is used to divide data points into K clusters so that the points within the clusters are as similar as possible and the points between clusters are as different as possible. The vehicle clustering area refers to the spatial clustering area of ​​the candidate stop points of the target vehicle obtained through cluster analysis.

[0097] In the embodiment of the present application, by obtaining the parking data of each vehicle in the target vehicle set and screening out candidate parking points based on the preset duration threshold and POI type, the present application can effectively identify the common parking areas of vehicles. The k-means algorithm is used to perform cluster analysis on these candidate parking points, and combined with the POI type in the geographic information system, the vehicle gathering area and its attributes can be accurately determined. This method not only improves the accuracy of identifying fake green pass evasion gangs, but also reveals the spatial distribution pattern and behavioral characteristics of evasion vehicles.

[0098] In the embodiment of the present application, by adopting an automated data acquisition and feature construction method, combined with a pre-trained green pass vehicle identification model, the probability of each vehicle to be identified as a fake green pass can be accurately calculated. Feature construction is based on vehicle parking data, static information, and trajectory data to obtain feature information that characterizes vehicle transportation. Through this feature information, high-risk evasion suspects can be effectively screened out, thereby improving the efficiency of inspection in a targeted manner. Ultimately, through cluster analysis, evasion gangs with similar behavior patterns can be identified, achieving a precise strike against evasion behavior. This application not only improves the accuracy and efficiency of identifying fake green pass vehicles, but also enhances the precise identification of evasion gangs, ensuring the fair implementation of the green channel policy.

[0099] See also Figure 3 , which is a flow chart of a model training method for a green pass vehicle recognition model provided in an embodiment of the present application. Figure 3 As shown, the method of the embodiment of the present application may include the following steps:

[0100] S201, obtaining second vehicle data of each toll-evading vehicle in a preset period among the toll-evading vehicles that have historically faked Green Pass;

[0101] S202, constructing a characteristic factor sequence of each toll-evading vehicle based on the second vehicle data;

[0102] S203, setting a counterfeit green pass probability for each characteristic factor sequence of the toll evading vehicle as a weight of the second characteristic factor sequence, and obtaining a plurality of model training samples;

[0103] S204, using a decision tree algorithm to create a green pass vehicle recognition model;

[0104] S205, performing machine learning on the green pass vehicle recognition model according to each model training sample, so as to optimize the model parameters of the green pass vehicle recognition model and obtain a pre-trained green pass vehicle recognition model.

[0105] In some embodiments of the present application, the specific process of performing machine learning on the green pass vehicle recognition model according to each model training sample includes: inputting each model training sample into the green pass vehicle recognition model, outputting the model loss value, and generating a pre-trained green pass vehicle recognition model when the model loss value reaches the minimum; or when the model loss value has not reached the minimum, backpropagating the model loss value to update the parameters of the green pass vehicle recognition model, and continuing to execute the step of inputting each model training sample into the green pass vehicle recognition model until the model loss value reaches the minimum.

[0106] In an embodiment of the present application, an accurate green pass vehicle recognition model is created. The model is trained using data on historical toll evasion vehicles through a machine learning algorithm to learn the characteristics of toll evasion behavior, thereby optimizing model parameters and improving recognition accuracy.

[0107] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0108] See also Figure 4 , which shows a schematic diagram of the structure of a device for identifying a toll-evading vehicle that impersonates a green pass provided by an exemplary embodiment of the present application. The device for identifying a toll-evading vehicle that impersonates a green pass can be implemented as all or part of a device through software, hardware, or a combination of both. The device 1 includes a data acquisition module 10, a characteristic factor construction module 20, a fake green pass probability output module 30, a vehicle acquisition module 40, and a vehicle clustering module 50.

[0109] The data acquisition module 10 is used to acquire first vehicle data of a plurality of vehicles to be identified within a preset historical time period from the current moment;

[0110] A characteristic factor construction module 20, used to construct a characteristic factor sequence of each vehicle to be identified according to the first vehicle data;

[0111] A fake green pass probability output module 30 is used to input the characteristic factor sequence of each vehicle to be identified into a pre-trained green pass vehicle identification model, and output the fake green pass probability of each vehicle to be identified;

[0112] The vehicle acquisition module 40 is used to acquire the vehicles to be identified corresponding to the counterfeit green pass probability within a preset probability interval as a target vehicle set;

[0113] The vehicle clustering module 50 is used to perform clustering processing on the target vehicle set, and treat the target vehicles of the same category as a group of toll evasion vehicles pretending to be green pass.

[0114] It should be noted that the device for identifying vehicles that impersonate Green Pass and evade tolls provided in the above-mentioned embodiment only uses the division of the above-mentioned functional modules as an example when executing the method for identifying vehicles that impersonate Green Pass and evade tolls. In actual applications, the above-mentioned 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. In addition, the device for identifying vehicles that impersonate Green Pass and evade tolls provided in the above-mentioned embodiment and the method for identifying vehicles that impersonate Green Pass and evade tolls belong to the same concept. The implementation process is detailed in the method embodiment and will not be repeated here.

[0115] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0116] In the embodiment of the present application, by adopting an automated data acquisition and feature construction method, combined with a pre-trained green pass vehicle identification model, the probability of each vehicle to be identified as a fake green pass can be accurately calculated. Feature construction is based on vehicle parking data, static information, and trajectory data to obtain feature information that characterizes vehicle transportation. Through this feature information, high-risk evasion suspects can be effectively screened out, thereby improving the efficiency of inspection in a targeted manner. Ultimately, through cluster analysis, evasion gangs with similar behavior patterns can be identified, achieving a precise strike against evasion behavior. This application not only improves the accuracy and efficiency of identifying fake green pass vehicles, but also enhances the precise identification of evasion gangs, ensuring the fair implementation of the green channel policy.

[0117] The present application also provides a computer-readable medium having program instructions stored thereon, which, when executed by a processor, implement the method for identifying vehicles that impersonate green passes and evade tolls provided in the above-mentioned method embodiments.

[0118] The present application also provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the method for identifying vehicles that impersonate Green Pass and evade tolls in the above-mentioned various method embodiments.

[0119] See also Figure 5 , which is a schematic diagram of the structure of a device provided in the embodiment of the present application. Figure 5 As shown, the device 1000 may include: at least one processor 1001 , at least one network interface 1004 , a user interface 1003 , a memory 1005 , and at least one communication bus 1002 .

[0120] The communication bus 1002 is used to realize the connection and communication between these components.

[0121] The user interface 1003 may include a display screen (Display) and a camera (Camera), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.

[0122] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0123] Among them, the processor 1001 may include one or more processing cores. The processor 1001 uses various interfaces and lines to connect the various parts of the entire device 1000, and executes various functions and processes data of the device 1000 by running or executing instructions, programs, code sets or instruction sets stored in the memory 1005, and calling data stored in the memory 1005. Optionally, the processor 1001 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 1001 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes operating devices, user interfaces and applications, etc.; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 1001, and it can be implemented separately through a chip.

[0124] Among them, the memory 1005 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing the operating device, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 1005 may also be optionally at least one storage device located away from the aforementioned processor 1001. As Figure 5As shown, the memory 1005 as a computer storage medium may include an operating device, a network communication module, a user interface module, and an application for identifying vehicles that impersonate Green Pass and evade tolls.

[0125] exist Figure 5 In the device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 1001 can be used to call the fake green pass toll evasion vehicle identification application stored in the memory 1005, and specifically perform the following operations:

[0126] Acquire first vehicle data of a plurality of vehicles to be identified within a preset historical time period from the current moment;

[0127] Constructing a characteristic factor sequence of each vehicle to be identified according to the first vehicle data;

[0128] Input the characteristic factor sequence of each vehicle to be identified into the pre-trained green pass vehicle identification model, and output the fake green pass probability of each vehicle to be identified;

[0129] Obtain the vehicles to be identified corresponding to the counterfeit green pass probability within the preset probability interval as the target vehicle set;

[0130] The target vehicle set is clustered and the target vehicles of the same category are regarded as a group of toll evasion vehicles pretending to be green pass.

[0131] In one embodiment, when the processor 1001 generates a pre-trained green pass vehicle recognition model, the processor 1001 specifically performs the following operations:

[0132] Obtain the second vehicle data of each toll-evading vehicle in the preset period among the toll-evading vehicles that have historically pretended to be green passes;

[0133] constructing a characteristic factor sequence of each toll-evading vehicle according to the second vehicle data;

[0134] The probability of fake green pass is set for each characteristic factor sequence of toll-evading vehicles as the weight of the second characteristic factor sequence to obtain multiple model training samples;

[0135] A decision tree algorithm is used to create a green pass vehicle recognition model;

[0136] According to each model training sample, machine learning is performed on the green pass vehicle recognition model to optimize the model parameters of the green pass vehicle recognition model and obtain a pre-trained green pass vehicle recognition model.

[0137] In one embodiment, the processor 1001 executes the following operations when the vehicle data includes vehicle parking data, vehicle static data, and vehicle trajectory data; and constructs a characteristic factor sequence of the vehicle according to the vehicle data:

[0138] Based on the vehicle parking data, construct the vehicle's loading point POI label and whether the loading point has the green pass attribute;

[0139] Constructing the attributes of the vehicle's highway mileage and whether it was driven at night based on the vehicle trajectory data;

[0140] According to the static data of the vehicle, the vehicle model, the number of vehicles of the transport company to which it belongs, and the rated load are constructed;

[0141] The vehicle's loading point POI label and whether the loading point has the green pass attribute are classified with the vehicle's highway mileage, vehicle model, number of vehicles of the transport company to which it belongs, rated load, and whether it is driven at night to obtain the vehicle's characteristic factor sequence.

[0142] In one embodiment, when the processor 1001 constructs the POI tag of the vehicle's loading point and whether the loading point has a green pass attribute according to the vehicle parking data, the processor 1001 specifically performs the following operations:

[0143] Obtain the stop time of each initial stop point in the vehicle stop data;

[0144] Determine whether the stop time of each initial stop point is greater than a preset time threshold;

[0145] Extracting initial stop points whose stop duration is greater than a preset stop duration threshold as multiple candidate stop points;

[0146] Get the stop type of each alternative stop point;

[0147] Determine whether the stop type of each candidate stop point belongs to the preset POI type;

[0148] Eliminate the candidate stop points whose stop types belong to the preset POI type from the multiple candidate stop points, and obtain multiple valid stop points for the vehicle;

[0149] According to the multiple valid stopping points of the vehicle, the POI label of the loading point of the vehicle and whether the loading point has the green pass attribute are analyzed.

[0150] In one embodiment, when the processor 1001 analyzes the POI tag of the loading point of the vehicle and whether the loading point has a green pass attribute according to multiple valid stopping points of the vehicle, the processor 1001 specifically performs the following operations:

[0151] From multiple valid stopping points of the vehicle, multiple candidate stopping points with the shortest distances from the highway entrance are obtained, and the absolute value of the distance difference between two adjacent candidate stopping points from the multiple candidate stopping points and the highway entrance is less than a preset distance threshold;

[0152] Determine whether each candidate stop point is within the electronic fence used to characterize the green pass POI type;

[0153] If so, bind each candidate stop point to the green pass POI type corresponding to the electronic fence where it is located, obtain the POI label of each candidate stop point, use it as the loading point POI label of the vehicle, and determine that each candidate stop point has the green pass attribute, and obtain the loading point of the vehicle has the green pass attribute;

[0154] If not, and the stop type of each candidate stop point does not belong to the national POI type in the geographic information system, and there is no national POI type within the preset range around each candidate stop point, it is determined that each candidate stop point has a green pass attribute, and the loading point of the vehicle has a green pass attribute;

[0155] If not and the stop type of each candidate stop point belongs to the national POI type in the geographic information system, the attribute of each candidate stop point is determined to be non-green pass, and the loading point of the vehicle does not have the green pass attribute.

[0156] In one embodiment, when the processor 1001 constructs the attributes of the vehicle's highway mileage and whether it is driven at night based on the vehicle trajectory data, the processor 1001 specifically performs the following operations:

[0157] From the vehicle trajectory data, obtain the driving trajectory between the vehicle's entry point into the highway and the current exit point;

[0158] Calculate the vehicle's high-speed mileage based on the driving trajectory;

[0159] Obtaining the driving time period of the vehicle from the vehicle trajectory data;

[0160] According to the driving time period of the vehicle, it is determined whether the vehicle is driving at night, and an attribute of whether the vehicle is driving at night is obtained.

[0161] In one embodiment, the processor 1001 performs clustering processing on the target vehicle set, and when the target vehicles of the same category are regarded as a group of toll evasion vehicles pretending to be green pass, the following specific operations are performed:

[0162] Obtain the parking data of each target vehicle in the target vehicle set;

[0163] According to the stop data of each target vehicle, the candidate stop points of each target vehicle are analyzed; the candidate stop points of each target vehicle are obtained by screening based on a preset time threshold and a preset POI type;

[0164] The k-means algorithm and the distance between adjacent candidate stops are used to cluster the candidate stops of all target vehicles to obtain the vehicle clustering area of ​​each target vehicle;

[0165] Use the vehicle gathering area of ​​each target vehicle to associate the national POI type in the geographic information system to obtain the POI attribute of the vehicle gathering area of ​​each target vehicle;

[0166] Based on the existence of the same vehicle gathering areas, POI attributes of the vehicle gathering areas, and the company to which the vehicles belong, the k-nearest neighbor algorithm is used for classification, and the classification results with the same characteristics after classification are divided into a group, obtaining multiple groups of vehicles that pretend to be green passes and evade tolls.

[0167] In the embodiment of the present application, by adopting an automated data acquisition and feature construction method, combined with a pre-trained green pass vehicle identification model, the probability of each vehicle to be identified as a fake green pass can be accurately calculated. Feature construction is based on vehicle parking data, static information, and trajectory data to obtain feature information that characterizes vehicle transportation. Through this feature information, high-risk evasion suspects can be effectively screened out, thereby improving the efficiency of inspection in a targeted manner. Ultimately, through cluster analysis, evasion gangs with similar behavior patterns can be identified, achieving a precise strike against evasion behavior. This application not only improves the accuracy and efficiency of identifying fake green pass vehicles, but also enhances the precise identification of evasion gangs, ensuring the fair implementation of the green channel policy.

[0168] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program for identifying vehicles that evade tolls by impersonating Green Pass can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium of the program for identifying vehicles that evade tolls by impersonating Green Pass can be a disk, an optical disk, a read-only storage memory, or a random access storage memory, etc.

[0169] The above disclosure is only the preferred embodiment of the present application, which certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A method for identifying vehicles that impersonate green pass and evade tolls, characterized in that: The method comprises: Acquire first vehicle data of a plurality of vehicles to be identified within a preset historical time period from the current moment; constructing a characteristic factor sequence of each vehicle to be identified according to the first vehicle data; Inputting the characteristic factor sequence of each vehicle to be identified into a pre-trained green pass vehicle identification model, and outputting the counterfeit green pass probability of each vehicle to be identified; Obtain the vehicles to be identified corresponding to the counterfeit green pass probability within the preset probability interval as the target vehicle set; The target vehicle set is clustered, and target vehicles of the same category are regarded as a group of toll evasion vehicles pretending to be green pass.

2. The method according to claim 1, characterized in that Follow these steps to generate a pre-trained green pass vehicle recognition model, including: Obtain the second vehicle data of each toll-evading vehicle in the preset period among the toll-evading vehicles that have historically pretended to be green passes; constructing a characteristic factor sequence of each toll-evading vehicle according to the second vehicle data; For each of the characteristic factor sequences of the toll-evading vehicles, a probability of a fake green pass is set as a weight of the second characteristic factor sequence to obtain a plurality of model training samples; A decision tree algorithm is used to create a green pass vehicle recognition model; According to each model training sample, machine learning is performed on the green pass vehicle recognition model to optimize the model parameters of the green pass vehicle recognition model, thereby obtaining a pre-trained green pass vehicle recognition model.

3. The method according to claim 1 or 2, characterized in that: Vehicle data includes vehicle parking data, vehicle static data and vehicle trajectory data; According to the vehicle data, construct the vehicle's characteristic factor sequence, including: Based on the vehicle parking data, construct the vehicle's loading point POI label and whether the loading point has the green pass attribute; Constructing the attributes of the vehicle's highway mileage and whether it was driven at night based on the vehicle trajectory data; According to the static data of the vehicle, the vehicle model, the number of vehicles of the transport company to which it belongs, and the rated load are constructed; The loading point POI label of the vehicle and whether the loading point has the green pass attribute are classified with the attributes of the vehicle's highway mileage, vehicle type, number of vehicles of the transportation company to which it belongs, rated load, and whether it is driven at night, to obtain the characteristic factor sequence of the vehicle.

4. The method according to claim 3, characterized in that The step of constructing the POI tag of the vehicle's loading point and whether the loading point has a green pass attribute according to the vehicle parking data includes: Obtaining the stop time of each initial stop point in the vehicle stop data; Determine whether the stop time of each initial stop point is greater than a preset time threshold; Extracting initial stop points whose stop duration is greater than a preset stop duration threshold as multiple candidate stop points; Obtaining the stop type of each candidate stop point; Determine whether the stop type of each candidate stop point belongs to a preset POI type; Eliminate the candidate stop points whose stop type belongs to the preset POI type from the multiple candidate stop points to obtain multiple valid stop points for the vehicle; According to the multiple valid stop points of the vehicle, the loading point POI tag of the vehicle and whether the loading point has a green pass attribute are analyzed.

5. The method according to claim 4, characterized in that The step of analyzing the POI tag of the loading point of the vehicle and whether the loading point has a green pass attribute according to the multiple valid stopping points of the vehicle includes: Acquire multiple candidate stopping points with the shortest distances from a highway entrance from the multiple valid stopping points of the vehicle, wherein the absolute value of the distance difference between two adjacent candidate stopping points from the multiple candidate stopping points and the highway entrance is less than a preset distance threshold; Determine whether each candidate stop point is within an electronic fence used to characterize a green pass POI type; If, the candidate stops are bound to the green pass POI type corresponding to the electronic fence where they are located, the POI tags of the candidate stops are obtained as the loading point POI tags of the vehicle, and it is determined that the candidate stops have the green pass attribute, and the loading point of the vehicle has the green pass attribute; If not, and the stop type of each candidate stop point does not belong to the national POI type in the geographic information system, and the national POI type does not exist within a preset range around each candidate stop point, it is determined that each candidate stop point has a green pass attribute, and the loading point of the vehicle has a green pass attribute; If not and the stop type of each candidate stop point belongs to the national POI type in the geographic information system, it is determined that each candidate stop point has a non-green pass attribute, and the loading point of the vehicle does not have a green pass attribute.

6. The method according to claim 3, characterized in that The step of constructing the attributes of the vehicle's highway mileage and whether the vehicle is driven at night based on the vehicle trajectory data includes: Obtaining, from the vehicle trajectory data, a driving trajectory of the vehicle between an entry point of the expressway and a current exit point; Calculating the high-speed road mileage of the vehicle according to the driving trajectory; Obtaining a driving time period of the vehicle from the vehicle trajectory data; According to the driving time period of the vehicle, it is determined whether the vehicle is driving at night, and an attribute of whether the vehicle is driving at night is obtained.

7. The method according to claim 1, characterized in that The clustering process of the target vehicle set is performed to treat the target vehicles of the same category as a group of vehicles that are faking green pass and evading tolls, including: Obtaining the parking data of each target vehicle in the target vehicle set; Analyze the candidate stop points of each target vehicle according to the stop data of each target vehicle; the candidate stop points of each target vehicle are obtained by screening based on a preset time threshold and a preset POI type; The k-means algorithm and the distance between adjacent candidate stops are used to cluster the candidate stops of all target vehicles to obtain the vehicle clustering area of ​​each target vehicle; Use the vehicle gathering area of ​​each target vehicle to associate the national POI type in the geographic information system to obtain the POI attribute of the vehicle gathering area of ​​each target vehicle; Based on the existence of the same vehicle gathering area, the POI attributes of the vehicle gathering area and the company to which the vehicles belong, classification is performed using the k-nearest neighbor algorithm, and the classification results with the same characteristics after classification are divided into a group, thereby obtaining multiple groups of vehicles that pretend to be green passes and evade tolls.

8. A device for identifying vehicles that impersonate Green Pass and evade tolls, characterized in that: The device comprises: A data acquisition module, used to acquire first vehicle data of a plurality of vehicles to be identified within a preset historical time period from the current moment; A characteristic factor construction module, used to construct a characteristic factor sequence of each vehicle to be identified according to the first vehicle data; A fake green pass probability output module, used to input the characteristic factor sequence of each vehicle to be identified into a pre-trained green pass vehicle identification model, and output the fake green pass probability of each vehicle to be identified; A vehicle acquisition module is used to acquire vehicles to be identified that correspond to the probability of counterfeit green pass within a preset probability interval as a target vehicle set; The vehicle clustering module is used to cluster the target vehicle set and treat the target vehicles of the same category as a group of toll-evading vehicles that pretend to be green pass vehicles.

9. A computer storage medium, characterized in that The computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the method according to any one of claims 1 to 7.

10. A device, characterized in that: include: A processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the method according to any one of claims 1 to 7.

Images