Traveling portal recognition model training, path recognition and toll transaction method and device based on base station

Abstract

The invention provides a passing portal recognition model training, path recognition and toll transaction method and device based on a base station, and the method comprises the steps: building the mapping of a communication base station identification code and a portal system identification code on a driving path through a training neural network model, and obtaining a passing portal recognition model; a passing base station identification sequence is constructed by means of base station information obtained by a vehicle-mounted mobile terminal in the vehicle driving process, a passing portal frame identification model is input to restore a passing portal frame system identification prediction sequence in the passing process, the missing and complete passing portal frame sequence is supplemented, accurate charging can be achieved, and the charging efficiency is improved. Transaction errors caused by loss of passing portal information due to information medium damage, signal instability and artificial shielding are prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of expressway transaction and charging, and in particular to a method and device for training a gantry recognition model, identifying a path, and conducting toll transaction based on a base station. Background Art

[0002] During the high-speed passage, the ETC (non-stop toll collection system) gantry system is set up to realize the segmented billing of vehicles. ETC vehicles pass through the highway mainly in the electronic non-stop toll collection mode. The ETC gantry system charges the passing vehicles according to the charging standard of the billing unit and the vehicle model and other information in the passage medium, and writes the billing information into the passage medium. Although the system runs stably, there are still problems to be optimized. It is mainly reflected in: due to the damage of the medium, unstable signals, and some customers manually blocking the billing facilities, the gantry points on the way are missing, resulting in the failure of path restoration and the failure of calling the departmental / provincial online billing. Finally, the minimum fee is used for billing, and it is impossible to charge according to the actual passage path of the vehicle, resulting in toll losses for highway owners. Therefore, a technical solution is urgently needed to solve the problem of path restoration errors caused by missing gantry data. Summary of the invention

[0003] In view of this, an embodiment of the present invention provides a base station-based gantry recognition model training, path recognition, and toll transaction method and device to eliminate or improve one or more defects existing in the prior art and solve the problem that the gantry information passed is missing and difficult to recover during the high-speed toll billing process.

[0004] One aspect of the present invention provides a method for training a traveling gantry recognition model based on base station information, comprising:

[0005] Obtain a relationship comparison table between communication base stations and high-speed gantry systems, wherein the relationship comparison table records one or more communication base stations that can be received by each high-speed gantry system in the target area, and establishes a corresponding relationship; the high-speed gantry system is named according to a first rule to obtain a corresponding gantry system identification code, and the communication base station is named according to a second rule to obtain a communication base station identification code;

[0006] Construct a plurality of high-speed random path samples, and search for a gantry system identification sample sequence that the high-speed random path samples pass through; for each gantry system in the gantry system identification sample sequence, randomly select at least one corresponding communication base station identification code in the relationship comparison table, and establish a communication base station identification sample sequence;

[0007] Constructing a training sample set, the training sample set comprising a plurality of samples, each sample comprising a communication base station identification sample sequence and a gantry system identification sample sequence corresponding to a high-speed passing random path sample;

[0008] Acquire an initial neural network model, the initial neural network model comprising a word embedding coding layer, a feature extraction layer, a sequence modeling layer and a fully connected layer; the initial neural network model takes the communication base station identification sample sequence as input, encodes it through the word embedding coding layer, extracts features through the feature extraction layer, and then mines the long-distance dependency relationship in the sequence based on the self-attention mechanism through the sequence modeling layer, and the fully connected layer predicts and outputs the predicted value of the gantry system identification sample sequence;

[0009] A loss function is constructed based on the deviation between the predicted value and the true value of the gantry system identification sample sequence, and the parameters of the initial neural network model are updated according to the loss function to obtain a gantry recognition model.

[0010] In some embodiments, the gantry system identification code is marked with the provincial administrative division code, ETC gantry type, ETC gantry direction, and ETC gantry serial number as a unique identifier; the communication base station identification code is marked with the mobile network code, tracking area code, and cell unique identifier.

[0011] In some embodiments, the method further comprises: performing data enhancement on each sample in the training sample set in the form of random exchange, noise addition and sequence length transformation.

[0012] In some embodiments, the feature extraction layer adopts a convolutional neural network, and a maximum pooling layer is provided between each convolution kernel in the convolutional neural network; the sequence modeling layer adopts a Transformer structure.

[0013] On the other hand, the present invention provides a method for restoring a gantry sequence based on base station information, the method being executed by a high-speed toll transaction platform, and comprising the following steps:

[0014] Receive the target vehicle entry prompt information sent by the entrance toll station, and send a base station information collection instruction to the preset mobile terminal deployed on the vehicle according to the prompt information, so as to collect and record the identification code of the communication base station connected during the journey and its corresponding timestamp;

[0015] During high-speed passage, receiving vehicle passage data sent by multiple gantry systems when the target vehicle passes by, the vehicle passage data including a timestamp, a license plate number of the detected target vehicle, and a gantry identification code;

[0016] Receiving a settlement instruction for the target vehicle sent by the exit toll station, finding the missing path according to the gate identification code in the acquired vehicle traffic data, and finding the missing time period according to the timestamps at the beginning and end of the missing path;

[0017] Sending a base station information request instruction to the service library to request the preset mobile terminal to generate the communication base station identification code and its corresponding timestamp during the missing period, and constructing a base station identification sequence in chronological order;

[0018] Input the passing base station identification sequence corresponding to the missing period into the passing gantry identification model in the training method of the passing gantry identification model based on base station information to output the passing gantry identification prediction sequence in the missing period;

[0019] The predicted sequence of gantry identifications during the missing period is incorporated into the vehicle traffic data to obtain a complete gantry sequence.

[0020] In some embodiments, incorporating the predicted sequence of gantry identifications during the missing period into the vehicle traffic data comprises:

[0021] When there are multiple missing time periods in a single trip, the gantry identification prediction sequence is inserted into the gantry identification coding sequence detected and provided by the gantry system in chronological order, and the repeated gantry identification codes are deleted to obtain a complete gantry sequence.

[0022] In some embodiments, the method further comprises: performing a path accessibility check on the complete gantry sequence, and generating an error message when an inaccessible path exists.

[0023] In another aspect, the present invention provides a highway toll transaction method, which is executed by a highway toll transaction platform, and comprises the following steps:

[0024] Obtaining the complete gantry sequence obtained in the above method of restoring the gantry sequence based on base station information;

[0025] Call the departmental or provincial billing port to calculate the highway toll based on the complete gantry sequence and send it to the exit toll station for transaction.

[0026] On the other hand, the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, which implements the steps of the above method when executed by a processor.

[0027] On the other hand, the present invention provides a computer program product, comprising a computer program / instruction, characterized in that the computer program / instruction implements the steps of the above method when executed by a processor.

[0028] The beneficial effects of the present invention are at least:

[0029] The base station-based gantry recognition model training, path recognition, and toll transaction method and device described in the present invention establish a mapping between the communication base station identification code and the gantry system identification code on the driving path by training a neural network model to obtain a gantry recognition model, construct a base station identification sequence with the help of the base station information obtained by the on-board mobile terminal during the vehicle's driving, input the gantry recognition model to restore the predicted sequence of the gantry system identification during the passage process, fill in the missing and restore the complete gantry sequence, which can achieve accurate billing and prevent transaction errors caused by loss of gantry information due to damage to the signal medium, unstable signals, and artificial shielding.

[0030] Additional advantages, purposes, and features of the present invention will be described in part in the following description, and will become apparent to those skilled in the art after studying the following, or may be learned from the practice of the present invention. The purposes and other advantages of the present invention may be achieved and obtained by the structures specifically indicated in the specification and the accompanying drawings.

[0031] Those skilled in the art will appreciate that the objectives and advantages that can be achieved with the present invention are not limited to the above specific description, and the above and other objectives that can be achieved by the present invention will be more clearly understood from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of the present application, and do not constitute a limitation of the present invention. In the drawings:

[0033] Figure 1 The present invention is a flowchart of a method for training a gantry recognition model based on base station information according to an embodiment of the present invention.

[0034] Figure 2 This is a relationship diagram between gantry and base station signals in a method for restoring a gantry sequence based on base station information described in an embodiment of the present invention.

[0035] Figure 3 This is a schematic diagram of the model structure in the training method of the traveling gantry recognition model based on base station information described in one embodiment of the present invention.

[0036] Figure 4 The present invention is a flowchart of a method for restoring a gantry sequence based on base station information according to an embodiment of the present invention.

[0037] Figure 5 The figure is a logical diagram of a highway toll transaction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0039] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, only structures and / or processing steps closely related to the solutions according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0040] It should be emphasized that the term “include / comprises” when used herein refers to the presence of features, elements, steps or components, but does not exclude the presence or addition of one or more other features, elements, steps or components.

[0041] Due to media damage, unstable signals, and some customers' manual blocking of billing facilities, the gantry points along the way are missing, which may lead to failure of path restoration and failure of calling departmental / provincial online billing. Ultimately, the minimum fee is used for billing, and it is impossible to charge according to the vehicle's actual route, resulting in toll losses for highway owners.

[0042] Therefore, the present invention first provides a training method for a gantry recognition model based on base station information, such as Figure 1 As shown, it includes steps S101 to S105:

[0043] Step S101: Obtain a relationship comparison table between communication base stations and high-speed gantry systems, in which one or more communication base stations that can be received by each high-speed gantry system in the target area are recorded, and a corresponding relationship is established; the high-speed gantry system is named according to a first rule to obtain a corresponding gantry system identification code, and the communication base station is named according to a second rule to obtain a communication base station identification code.

[0044] Step S102: construct multiple high-speed random path samples, and find the gantry system identification sample sequence that the high-speed random path samples pass through; for each gantry system in the gantry system identification sample sequence, randomly select at least one corresponding communication base station identification code from the relationship comparison table to establish a communication base station identification sample sequence.

[0045] Step S103: construct a training sample set, the training sample set includes multiple samples, each sample includes a communication base station identification sample sequence and a gantry system identification sample sequence corresponding to a high-speed random path sample.

[0046] Step S104: Obtain an initial neural network model, which includes a word embedding coding layer, a feature extraction layer, a sequence modeling layer and a fully connected layer; the initial neural network model takes the communication base station identification sample sequence as input, encodes it through the word embedding coding layer, extracts features through the feature extraction layer, and then mines the long-distance dependency relationship in the sequence based on the self-attention mechanism through the sequence modeling layer, and then the fully connected layer predicts and outputs the predicted value of the gantry system identification sample sequence.

[0047] Step S105: construct a loss function based on the deviation between the predicted value and the true value of the gantry system identification sample sequence, and update the parameters of the initial neural network model according to the loss function to obtain a gantry recognition model.

[0048] In step S101, on the road section where the gantry system is located, the mobile terminal devices on the passing vehicles can connect to the nearby base stations. Since the base station has limited ability to establish a connection, the base stations that the vehicle can connect to when passing through the gantry system are relatively fixed, and there is a certain corresponding relationship. This also provides the possibility of searching for the passing gantry based on the base station information.

[0049] The relationship comparison table records the identification information of all base stations that the vehicle may connect to when passing through each gantry system, and forms a comparison. The high-speed gantry system is named according to the first rule to obtain the corresponding gantry system identification code, which can be specifically marked with the provincial administrative division code, ETC gantry type, ETC gantry direction, and ETC gantry serial number as the unique identifier. The communication base station identification code is marked with the mobile network code, tracking area code, and cell unique identifier.

[0050] In step S102, for a gantry system with multiple corresponding base stations, the vehicle will only connect to one of them during the driving process. When there are adjacent or intersecting sections or the signal conditions are good, the signal of the same base station may be connected to multiple gantry systems, that is, a single base station may have multiple corresponding gantry systems in the relationship comparison table. Figure 2 As shown, there are multiple arrival paths between gantries 1 to 5, gantry 2 and gantry 5 are adjacent, and the vehicle can receive base station signals 1 to 6 when passing through gantry 2, and can receive base station signals 4, 7 and 8 when passing through gantry 5. The vehicle travels from gantry 1 to gantry 3 from time t1 to time t2, and it may connect to at least one of the base station signals 1 to 8 during this period of time.

[0051] Therefore, during the driving process of the vehicle, based on the relatively fixed relationship between the gantry and the base station, there is still uncertainty in the base station signal received by the vehicle when passing through the gantry. In view of this situation, in order to establish an effective mapping, it is necessary to provide conditions that meet the actual scene conditions. Therefore, by constructing a random path sample, combining the path sample to determine the gantry system that must be passed, and then randomly selecting a corresponding base station of the gantry system through the relationship comparison table as the base station that the mobile terminal may connect to. Construct a communication base station identification sample sequence and a gantry system identification sample sequence, which is closer to the actual driving process.

[0052] In step S103, the training sample set is established based on the base station identification sample sequence and the gantry system identification sample sequence. In practice, in order to obtain more accurate recognition results, more parameters related to the base station signal may be introduced, such as the base station signal strength.

[0053] In step S104, in order to train the model, a mapping between the communication base station identification code and the communication base station identification sample sequence is established, such as Figure 3 As shown, the initial neural network model used in the present invention includes a word embedding coding layer, a feature extraction layer, a sequence modeling layer and a fully connected layer;

[0054] Among them, the word embedding coding layer (embedding coding layer) encodes the base station sequence and the gantry number. First, a unique index is assigned to each base station number, and an Embedding layer is created to convert the discrete base station number into a continuous vector representation.

[0055] The feature extraction layer uses a convolutional neural network (CNN) to extract local features. A maximum pooling layer is provided between each convolution kernel in the convolutional neural network. Multiple convolution kernels are used to extract local features of the base station sequence. The feature vector size is reduced through maximum pooling to retain significant information.

[0056] In the sequence modeling layer, Transformer can capture long-distance dependencies in the sequence. Through the self-attention mechanism, the model can understand the global information of the entire sequence.

[0057] Fully connected layer, used to predict the output gantry.

[0058] In some embodiments, the method further sets a data enhancement layer to perform data enhancement on each sample in the training sample set in the form of random exchange, noise addition and sequence length transformation.

[0059] Specifically, the base station number sequence is processed to increase the generalization and anti-interference capabilities of the model during the training process. The base station positions in the sequence are randomly swapped to increase the model's adaptability to different orders; the base station sequence length is changed to increase the model's adaptability to sequence length; some random noise (digital replacement) is added to the base station number to simulate the data inaccuracy that may be encountered in the real environment.

[0060] In step S105, the initial neural network is trained based on the data-enhanced training sample set, so that a gantry recognition model that can establish a mapping relationship between the passed communication base station identification sample sequence and the gantry system identification sample sequence can be obtained.

[0061] On the other hand, the present invention provides a method for restoring the gantry sequence based on base station information, the method being executed by a high-speed toll transaction platform, such as Figure 4 As shown, the method includes the following steps S201 to S206:

[0062] Step S201: Receive the target vehicle entry prompt information sent by the entrance toll station, and send a base station information collection instruction to the preset mobile terminal deployed on the vehicle according to the prompt information, so as to collect and record the communication base station identification code and its corresponding timestamp generated during the journey.

[0063] Step S202: During high-speed passage, vehicle passage data sent by multiple gantry systems when a target vehicle is detected passing through is received. The vehicle passage data includes a timestamp, a license plate number of the detected target vehicle, and an identification code of the gantry.

[0064] Step S203: receiving a settlement instruction for the target vehicle sent by the exit toll station, finding the missing path according to the gantry identification code in the acquired vehicle traffic data, and finding the missing period according to the timestamps at the beginning and end of the missing path.

[0065] Step S204: Send a base station information request instruction to the service library to request the preset mobile terminal to generate a connected communication base station identification code and its corresponding timestamp during the missing period, and construct a base station identification sequence in chronological order.

[0066] Step S205: input the base station identification sequence corresponding to the missing period into the gantry identification model in the training method of the gantry identification model based on base station information described in steps S101 to S104 to output the predicted sequence of gantry identifications in the missing period.

[0067] Step S206: Incorporate the predicted sequence of gantry identifications during the missing period into the vehicle traffic data to obtain a complete gantry sequence.

[0068] The gantry recognition model trained in the above steps S101 to S104 is applied to the completion of path gantries in actual scenarios in steps S201 to S206 of this embodiment. The base station information identified by the owner and the test vehicle on the highway is collected regularly based on the APP (or mini program), and the gantry information passed by the vehicle is collected. The gantry and its surrounding identified base stations are associated according to the timestamp, and the corresponding relationship between the gantry and the base station is stored in the database. Through algorithm model training, when the missing gantry is identified, the gantry information is restored by inputting the identified base station information.

[0069] In step S201, when the target vehicle passes through the entrance toll station, the entrance toll station can sense and collect the license plate information of the target vehicle through cameras, ground sensors and other equipment. At the same time, a prompt message is generated and sent to the high-speed toll transaction platform, and the high-speed toll transaction platform sends a base station information collection instruction to the mobile terminal loaded on the target vehicle. The mobile terminal can be a dedicated device or a mobile phone loaded with a setting program or software. In practice, the driver can be required to collect base station information through the mobile phone by scanning the code installed in the APP or applet. The form of the mobile terminal is not fixed, and it should be understood that any form that can load a specific program and perform base station information collection tasks belongs to the scope of protection of the present invention.

[0070] The communication base station identification code is marked with the mobile network code MNC, the tracking area code TAC and the cell unique identifier ECI. By setting the timestamp and license plate number fields, and searching for the base station identified by the gantry within the preset time range, the gantry and surrounding base station identification data are obtained. The data sample is shown below. The unique base station information can be obtained by splicing, so it is spliced ​​as the unique base station number. The specific form is shown in Table 1 below.

[0071] Table 1 Base station raw data collection field table

[0072]

[0073]

[0074] By collecting the timestamp and license plate number fields and synchronously recording the number of the gantry system through which the target vehicle passes, the data recorded by the corresponding gantry system is obtained, as shown in Table 2 below.

[0075] Table 2 Gantry original data collection field table

[0076] Field Name type Example Timestamp TIMESTAMP 1642992000 License plate number VARCHAR(10) Beijing A·123BC Mast No. VARCHAR(20) S010163001000320010

[0077] In step S202, when the target vehicle passes through, if the gantry system detects the target vehicle, the vehicle passing data is recorded and sent to the highway toll transaction platform. This information is authentic and can be used for settlement.

[0078] In step S203, the vehicle traffic data may be missing due to medium damage, signal instability, human shielding and other issues. The missing period can be found by searching the timestamp.

[0079] In steps S204 to S206, when it is identified that the gantry data is missing, the base station data passed by the owner's mobile phone during the missing time period is queried, and the corresponding gantry information is restored based on the pre-trained model to achieve fitting of the vehicle trajectory. Specifically, the preset mobile terminal is requested to collect the base station identification codes collected during the missing period, and the base station identification sequence is constructed in chronological order. The base station identification sequence of each missing period is input into steps S101 to S104 for training to obtain the predicted sequence of gantry identifications during the missing period. This sequence is a gantry identification coding sequence output in chronological order, which can fill in the missing period to obtain a complete gantry sequence.

[0080] In some embodiments, in step S206, the predicted sequence of gantry identification during the missing time period is incorporated into the vehicle traffic data, including: when there are multiple missing time periods in a single trip, the predicted sequence of gantry identification is inserted into the gantry identification coding sequence detected and provided by the gantry system in chronological order, and the duplicate gantry identification codes are deleted to obtain a complete gantry sequence.

[0081] In some embodiments, the method further comprises: performing a path accessibility check on the complete gantry sequence, and generating an error message when an inaccessible path exists.

[0082] In another aspect, the present invention provides a highway toll transaction method, the method being executed by a highway toll transaction platform, referring to Figure 5 The method comprises the following steps S301-S302:

[0083] Step S301: Acquire the complete gantry sequence obtained in the method of restoring the gantry sequence based on base station information described in the above steps S201 to S206.

[0084] Step S302: Call the ministerial or provincial billing port, calculate the highway toll according to the complete gantry sequence, and send it to the exit toll station for transaction.

[0085] Corresponding to the above method, the present invention also provides an apparatus / system, which includes a computer device, the computer device includes a processor and a memory, the memory stores computer instructions, the processor is used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the apparatus / system implements the steps of the method described above.

[0086] The embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the aforementioned edge computing server deployment method are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0087] In summary, the base station-based gantry recognition model training, path recognition, and toll transaction method and device described in the present invention establish a mapping between the communication base station identification code and the gantry system identification code on the driving path by training a neural network model to obtain a gantry recognition model, and use the base station information obtained by the on-board mobile terminal during vehicle driving to construct a base station identification sequence, input the gantry recognition model to restore the gantry system identification prediction sequence during the passage process, fill in the missing and restore the complete gantry sequence, which can achieve accurate billing and prevent transaction errors caused by loss of gantry information due to damage to the signal medium, unstable signals, and artificial shielding.

[0088] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0089] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0090] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A training method for a traveling gantry recognition model based on base station information, characterized in that: include: Obtain a relationship comparison table between communication base stations and high-speed gantry systems, wherein the relationship comparison table records one or more communication base stations that can be received by each high-speed gantry system in the target area, and establishes a corresponding relationship; The high-speed gantry system is named according to the first rule to obtain a corresponding gantry system identification code, and the communication base station is named according to the second rule to obtain a communication base station identification code; Construct a plurality of high-speed random path samples, and search for a gantry system identification sample sequence that the high-speed random path samples pass through; for each gantry system in the gantry system identification sample sequence, randomly select at least one corresponding communication base station identification code in the relationship comparison table, and establish a communication base station identification sample sequence; Constructing a training sample set, the training sample set comprising a plurality of samples, each sample comprising a communication base station identification sample sequence and a gantry system identification sample sequence corresponding to a high-speed passing random path sample; Acquire an initial neural network model, the initial neural network model comprising a word embedding coding layer, a feature extraction layer, a sequence modeling layer and a fully connected layer; the initial neural network model takes the communication base station identification sample sequence as input, encodes it through the word embedding coding layer, extracts features through the feature extraction layer, and then mines the long-distance dependency relationship in the sequence based on the self-attention mechanism through the sequence modeling layer, and the fully connected layer predicts and outputs the predicted value of the gantry system identification sample sequence; A loss function is constructed based on the deviation between the predicted value and the true value of the gantry system identification sample sequence, and the parameters of the initial neural network model are updated according to the loss function to obtain a gantry recognition model.

2. The training method of the traveling gantry recognition model based on base station information according to claim 1 is characterized in that: The gantry system identification code is marked with the provincial administrative division code, ETC gantry type, ETC gantry direction, and ETC gantry serial number as a unique identifier; the communication base station identification code is marked with the mobile network code, tracking area code, and cell unique identifier.

3. The training method of the traveling gantry recognition model based on base station information according to claim 1 is characterized in that: The method further comprises: Data enhancement is performed on each sample in the training sample set in the form of random exchange, noise addition and sequence length transformation.

4. The training method of the traveling gantry recognition model based on base station information according to claim 1 is characterized in that: The feature extraction layer adopts a convolutional neural network, and a maximum pooling layer is provided between each convolution kernel in the convolutional neural network; the sequence modeling layer adopts a Transformer structure.

5. A method for restoring a gantry sequence based on base station information, characterized in that: The method is performed by a high-speed toll transaction platform, and the method comprises the following steps: Receive the target vehicle entry prompt information sent by the entrance toll station, and send a base station information collection instruction to the preset mobile terminal deployed on the vehicle according to the prompt information, so as to collect and record the identification code of the communication base station connected during the journey and its corresponding timestamp; During high-speed passage, receiving vehicle passage data sent by multiple gantry systems when the target vehicle passes by, the vehicle passage data including a timestamp, a license plate number of the detected target vehicle, and a gantry identification code; Receiving a settlement instruction for the target vehicle sent by the exit toll station, finding the missing path according to the gantry identification code obtained in the vehicle traffic data, and finding the missing period according to the timestamps at the beginning and end of the missing path; Sending a base station information request instruction to the service library to request the preset mobile terminal to generate the communication base station identification code and its corresponding timestamp during the missing period, and constructing a base station identification sequence in chronological order; Input the passing base station identification sequence corresponding to the missing period into the passing gantry identification model in the training method of the passing gantry identification model based on base station information as claimed in any one of claims 1 to 4 to output the passing gantry identification prediction sequence in the missing period; The predicted sequence of gantry identifications during the missing period is incorporated into the vehicle traffic data to obtain a complete gantry sequence.

6. The method for restoring the gantry sequence based on base station information according to claim 5 is characterized in that: Incorporating the predicted sequence of gantry identifications during the missing period into the vehicle traffic data comprises: When there are multiple missing time periods in a single trip, the gantry identification prediction sequence is inserted into the gantry identification coding sequence detected and provided by the gantry system in chronological order, and the repeated gantry identification codes are deleted to obtain a complete gantry sequence.

7. The method for restoring the gantry sequence based on base station information according to claim 5 is characterized in that: The method further comprises: The path accessibility is checked for the complete gantry sequence and an error message is generated when an unreachable path exists.

8. A highway toll transaction method, characterized in that: The method is performed by a high-speed toll transaction platform, and the method comprises the following steps: Obtaining a complete passing gantry sequence obtained in the method for restoring a passing gantry sequence based on base station information as described in any one of claims 5 to 7; Call the departmental or provincial billing port to calculate the highway toll based on the complete gantry sequence and send it to the exit toll station for transaction.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images