Unattended toll collection system based on mainline gantry and ramp gate fusion prediction toll

Abstract

The application discloses an unattended toll collection system based on main line gantry and ramp card fusion pre-toll, which comprises a front main line gantry, a ramp online pre-toll card mouth, a ramp special situation induction screen, a quick CPC card self-service toll lane and a toll platform; the front main line gantry is provided with a main line gantry module; the ramp online pre-toll card mouth comprises a ramp card mouth module and a pre-transaction platform, the pre-transaction platform is connected with the toll platform and a local database; the ramp special situation induction screen is provided with a dynamic induction module; the quick CPC card self-service toll lane is provided with an RFID card reader, a payment terminal and an abnormal processing unit. The system captures the vehicles entering the toll station by the ramp card mouth, performs online pre-toll in advance, identifies the transaction state of the vehicles entering the toll station in advance, completes the pre-toll processing, shortens the transaction time of the self-service vehicles through the pre-toll, and improves the passing efficiency of the self-service lane.

Description

Technical Field

[0001] This invention relates to the field of highway toll collection technology, and in particular to an unattended toll collection system based on the fusion of estimated fees from the mainline gantry and ramp cards. Background Technology

[0002] With the continuous growth of highway traffic flow, the traditional manual toll collection mode can no longer meet the demand for rapid passage. In order to improve traffic efficiency, many regions have begun to promote the intelligent transformation of highway toll stations and gradually replace manual toll lanes with self-service payment systems.

[0003] The current self-service payment lane workflow is as follows: After a vehicle enters the toll area, the system automatically recognizes the license plate via camera, calculates the fee based on vehicle type and mileage, and displays the amount and QR code on the screen; the driver completes the payment via mobile payment, and the vehicle is automatically released after system confirmation. Although the self-service payment system improves toll collection efficiency to some extent, it still has limitations. The entire payment process must be performed sequentially, resulting in low traffic efficiency. During peak hours, vehicle queues occur, affecting overall traffic flow. Furthermore, in special circumstances (including card reading failures or billing errors, or the need for manual intervention), these situations occur frequently, severely impacting lane traffic efficiency. Therefore, this solution proposes an unmanned toll collection system based on the integration of mainline gantry and ramp card estimated fees to address the aforementioned problems. Summary of the Invention

[0004] The purpose of this invention is to provide an unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an unattended toll collection system based on the integration of mainline gantry and ramp card estimated fees, including a front mainline gantry, ramp online estimated fee checkpoints, ramp special situation guidance screens, fast CPC card self-service payment lanes, and a toll collection platform; The front mainline gantry is equipped with a mainline gantry module, which is used to collect vehicle passage data and upload it to the tolling platform; The online toll collection point for the ramp includes a ramp checkpoint module and a pre-transaction platform. The ramp checkpoint module is deployed at the entrance of the toll station ramp and is used for license plate information recognition and data fusion processing; the pre-transaction platform connects the billing platform and the local database and is used to perform online billing and media verification. The ramp emergency guidance screen is equipped with a dynamic guidance module, including an emergency classification model and a variable information board; The fast CPC card self-service payment lane is equipped with an RFID card reader, a payment terminal, and an anomaly handling unit for self-service payment.

[0006] Preferably, the vehicle passage data collected by the mainline gantry module includes the vehicle CPC card number, vehicle type, entrance station code, entrance time, and cumulative mileage.

[0007] Preferably, the ramp checkpoint module simultaneously acquires visible light images and infrared feature data of vehicles, and performs correlation matching between the data acquired by the mainline gantry and the real-time capture data. The matching degree of the correlation matching is expressed as: (1); Where SIM(x) represents the string similarity function, This represents the difference between the timestamp of the gantry data and the capture time. This indicates the average travel time. This indicates the maximum permissible time difference threshold. , This represents the dynamic weighting coefficient, and: (2).

[0008] Preferably, the calculation for the data fusion processing is expressed as follows: (3); in, This indicates the final billing amount. This indicates the online billing result. Indicates the amount read from the medium. , Let represent the dynamic weighting coefficients, and satisfy the following: (4).

[0009] Preferably, the dynamic weighting coefficient is obtained through a credibility assessment and is expressed as follows: (5); (6); Among them, C online and C medium These represent the confidence scores for online billing and media reading, respectively.

[0010] Preferably, the special incident classification model uses a decision tree algorithm to identify and classify abnormal vehicle billing events in real time; The variable message sign uses a dynamic electronic display screen to provide drivers with guidance information on choosing the payment lane.

[0011] Preferably, the classification judgment of the special situation classification model is: If the entry information of a vehicle waiting to pay at the ramp checkpoint module is missing or matches the blacklist, a Level 1 warning will be issued. If the real-time online billing value minus the baseline billing value exceeds the billing difference threshold, it indicates that there is a deviation between the real-time billing and the expected baseline, and a secondary traffic diversion process will be performed. If the vehicle's remaining mileage is less than the minimum travel distance, a level three interception process will be initiated.

[0012] Preferably, when the special situation classification model performs classification processing; First, obtain the vehicle arrival time through the ramp checkpoint, denoted as t. arrival ; Then, the theoretical travel time from the dynamic electronic display screen to the lane is calculated and expressed as: (8); Where L represents the distance from the dynamic electronic display screen to the lane. Let N represent the average vehicle speed and N represent the number of vehicles currently in the queue. Indicates the processing delay for a single vehicle; Finally, the ratio of manual or autonomous lanes is dynamically adjusted, as shown below: (9); Where P represents the probability of a special situation vehicle, and K represents the maximum traffic capacity coefficient of the lane.

[0013] Preferably, when the exception handling unit detects a CPC card read failure, it initiates a three-level rollback mechanism, wherein the rollback level of the three-level rollback mechanism is represented as follows; The payment is completed using the pre-transaction amount at level 1; billing is based on the vehicle license plate's historical transaction records at level 2; and manual intervention is triggered at level 3, and the process is transferred to manual processing.

[0014] Preferably, the RFID reader adopts a time-division multiplexing protocol, as shown below: (10); (11); Among them, T c T represents the total cycle time. r T represents the data retrieval time. e T represents the data erasure time. s N represents standby time, k represents the safety factor, and N represents the safety factor. c R represents the amount of data in the card. t Indicates the transmission rate.

[0015] The technical effects and advantages of this invention are as follows: This system consists of a front-end mainline gantry, ramp online pre-paid toll booths, ramp emergency guidance screens, fast CPC card self-service toll lanes, and a tolling platform. The mainline gantry, as an existing tolling gantry, provides basic traffic data support for the system. The ramp booths are responsible for capturing images of vehicles entering the toll station and combining them with the data from the mainline gantry to pre-calculate online tolls and identify the transaction status of vehicles entering the toll station in advance to complete the pre-paid toll processing. The ramp emergency guidance screens guide vehicles with abnormal statuses to manual lanes, reducing manual intervention when self-service vehicles encounter special situations, ensuring efficient and smooth lane transactions, and shortening the transaction time of self-service vehicles by pre-calculating tolls, thereby improving the traffic efficiency of self-service lanes. Attached Figure Description

[0016] Figure 1 This is a block diagram of the system composition of the present invention.

[0017] Figure 2 This is a flowchart of the charging operation process of the system of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: The present invention provides as follows Figure 1 The unattended toll collection system shown is based on the integration of mainline gantry and ramp card estimated toll collection, including the front mainline gantry, ramp online estimated toll collection point, ramp special situation guidance screen, fast CPC card self-service payment lane and toll collection platform; A mainline gantry module is installed at the front mainline gantry. The mainline gantry module is used to collect vehicle passage data and upload it to the tolling platform. It should be noted that the mainline gantry module is located at the mainline gantry and is responsible for collecting vehicle passage data and uploading the data to the billing platform in real time for subsequent fee calculation and management.

[0020] Specifically, the vehicle passage data collected by the mainline gantry module includes the vehicle CPC card number, vehicle type, entrance station code, entrance time, and cumulative mileage.

[0021] The online toll collection system for toll stations includes a toll station module and a pre-transaction platform. The toll station module is deployed at the entrance of the toll station ramp and is used for license plate information recognition and data fusion processing to ensure the accuracy of toll information. The pre-transaction platform connects the billing platform and the local database to perform online billing and media verification, supporting fast response and efficient toll collection. It should be noted that media verification refers to the process of verifying and confirming media used for payment or billing (including electronic payment cards, mobile payment QR codes, and RFID tags). Specifically, the main contents of media verification include: identity verification, verifying the legality of the payment medium to ensure it is a valid payment tool. This can be achieved by checking the card's expiration date and status (whether it has been reported lost); data consistency check, ensuring that the data read from the payment medium (including card number and account information) matches expectations, avoiding billing issues due to information errors; balance or credit limit confirmation, for some payment methods (prepaid cards or credit cards), the system needs to check whether the user's account balance or credit limit is sufficient to complete the current billing operation; security verification, using encryption technology, dynamic tokens, and other means to ensure the security of the payment process and prevent fraud and unauthorized transactions; real-time feedback, during the media verification process, the system should be able to provide real-time feedback on the verification results for rapid response and processing of subsequent operations.

[0022] Data fusion processing involves capturing vehicle license plates at ramp checkpoints and transmitting the license plate information to the system backend. The backend then queries the transaction records of the upstream gantry based on the license plate information to obtain the vehicle's card and entry information. Simultaneously, it invokes the online tolling program, combining data collected from the mainline gantry and ramp checkpoints to calculate the estimated toll for the vehicle. During the calculation process, if cumulative tolling information from the upstream gantry is obtained, it is used as a backup tolling method to ensure accuracy. The system backend employs a multi-source data fusion algorithm and a distributed real-time computing framework, specifically including the following core technologies: The system employs Apache Kafka or Flink as the real-time data stream processing engine to parse license plate information (license plate number, capture time, and lane position) captured at ramp checkpoints in milliseconds and correlate it with historical transaction records (card number, entry information, and cumulative toll amount) at the mainline gantries. A fast hash table-based retrieval algorithm achieves a one-to-many mapping between license plates and card numbers, ensuring data matching is completed within one second of a vehicle entering the ramp. If mainline gantry data is missing (card not successfully read), a distributed transaction lock mechanism (Redis distributed lock) ensures data read / write consistency for the same vehicle across different gantries, preventing data conflicts caused by concurrent operations.

[0023] Specifically, the ramp checkpoint module simultaneously collects visible light images and infrared feature data of vehicles, and performs correlation matching between the data collected by the mainline gantry and the real-time capture data. The matching degree of the correlation matching is expressed as: (1); Where SIM(x) represents the string similarity function, This represents the difference between the timestamp of the gantry data and the capture time. This indicates the average travel time. This indicates the maximum permissible time difference threshold. , This represents the dynamic weighting coefficient, and: (2).

[0024] It should be noted that when collecting visible light images and infrared feature data of vehicles; The first step is to select cameras, including visible light cameras and infrared cameras. Visible light cameras are used to capture visible light images of the vehicle. Choose cameras with high resolution and wide dynamic range to ensure clear images can be obtained under different lighting conditions. Infrared cameras are used to capture infrared feature data of the vehicle. Especially at night or in low light environments, infrared cameras can provide better image quality, making it easier to identify the vehicle and its features. Then, the data acquisition equipment is configured, including a synchronization controller and a data storage unit. The synchronization controller is used to coordinate the operation of the visible light camera and the infrared camera to ensure that data is acquired simultaneously at the same time. The data storage unit is equipped with sufficient storage space to store the acquired visible light images and infrared data. The data acquisition process is then executed. When a vehicle is detected approaching via inductive loop, laser sensor, or radar equipment, the visible light and infrared cameras are triggered to acquire data. A synchronization controller ensures that the two cameras capture images at the same time to avoid image inconsistencies caused by time deviations. At the moment the vehicle enters the ramp checkpoint, the controller instructs the visible light and infrared cameras to take pictures simultaneously, acquiring the vehicle's visible light image and infrared feature data. A timestamp is added to each image and data record for subsequent processing and comparison.

[0025] Next, data processing is performed, pairing visible light images and infrared images at the same time to ensure that they correspond to different features of the same vehicle; using computer vision technology, vehicle features (license plate, vehicle model) are extracted from the visible light images, and thermal features (engine temperature, vehicle body temperature) are extracted from the infrared images.

[0026] Finally, data analysis is performed, applying image processing and machine learning algorithms to analyze the fused data, identify vehicle type, license plate number and other characteristics, and then perform billing or traffic management. Infrared feature data is analyzed to detect whether there are any abnormalities in the vehicle (overheating, malfunction).

[0027] Specifically, the computational representation of data fusion processing is as follows: (3); in, This indicates the final billing amount. This indicates the online billing result. Indicates the amount read from the medium. and Let represent the dynamic weighting coefficients, and satisfy the following: (4).

[0028] Furthermore, the dynamic weighting coefficients are obtained through credibility assessment and are expressed as follows: (5); (6); Among them, C online and C medium These represent the confidence scores for online billing and media reading, respectively.

[0029] It should be noted that credibility assessment is the process of evaluating the reliability and effectiveness of the system's dynamic weighting coefficients. Its evaluation indicators include: accuracy, measuring how closely the system's output matches reality, including the accuracy rate of license plate recognition; reliability, the system's stability and consistency under different conditions (different weather, lighting, traffic flow); a highly reliable system should maintain the same performance in various environments; timeliness, the speed of the system's response and the efficiency of data processing; timeliness assessment ensures that the system can respond and process data quickly during peak traffic periods; security, assessing the system's security during data transmission and storage, including preventing data leakage and tampering risks; and user satisfaction, evaluating user acceptance and satisfaction with the system through user feedback and surveys, including ease of use and intuitive perception of effectiveness.

[0030] The evaluation methods include: data analysis, which assesses the system's performance in actual operation by collecting and analyzing historical data, including calculating error rate and success rate; experimental verification, which verifies the system's performance under specific conditions through controlled experiments or simulations, including testing license plate recognition rate under different vehicle speeds and lighting conditions; model evaluation, which evaluates the performance of models using machine learning or deep learning methods such as cross-validation and confusion matrices; security audit, which conducts a security assessment of the system, including audits of code, network security, and data storage to identify potential security vulnerabilities; and user research, which collects user experience and feedback through questionnaires, interviews, etc., to assess user trust and satisfaction with the system. The implementation steps include: Define the evaluation objectives, clarifying the aspects and specific goals (accuracy, reliability) that need to be evaluated; select evaluation indicators, choosing appropriate indicators based on the evaluation objectives to quantify the evaluation results; Data collection: Collect relevant data, including system operation data, user feedback, and environmental influencing factors; Analysis and calculation: Analyze the collected data and calculate the values ​​of each indicator; Results interpretation: Based on the calculation results, assess the reliability of the system and identify possible directions for improvement.

[0031] The ramp emergency guidance screen is equipped with a dynamic guidance module, including an emergency classification model and a variable message sign; Specifically, the special incident classification model uses the decision tree algorithm. The decision tree can generate easy-to-understand classification rules based on input features (vehicle type, vehicle speed, and payment status) for real-time identification and classification of abnormal vehicle billing events. The variable message signs use dynamic electronic displays to provide drivers with guidance on choosing the payment lane.

[0032] It should be noted that the data input for the special incident classification model comes from vehicle information collected by the ramp checkpoint module, including vehicle passage data, license plate images, and infrared feature data. Historical data is used to train the decision tree model to improve its recognition rate of billing anomalies. The model is optimized through cross-validation and model pruning techniques to avoid overfitting. When a vehicle passes through the ramp, the dynamic guidance module analyzes the vehicle data in real time to identify billing anomalies (unpaid, failed payment, and card reading errors). Once a billing anomaly is identified, the system will generate corresponding warning messages or suggestions, which will be prepared to be issued to the driver through variable message signs.

[0033] Furthermore, the classification criteria of the special situation classification model are as follows: If the entry information of a vehicle waiting to pay at the ramp checkpoint module is missing or matches the blacklist, a Level 1 warning will be issued. If the real-time online billing value minus the baseline billing value exceeds the billing difference threshold, it indicates that there is a deviation between the real-time billing and the expected baseline, and a secondary traffic diversion process will be performed. If the vehicle's remaining mileage is less than the minimum travel distance, a level three interception process will be initiated.

[0034] Furthermore, when the special situation classification model performs classification processing; First, obtain the vehicle arrival time through the ramp checkpoint, denoted as t. arrival ; Then, the theoretical travel time from the dynamic electronic display screen to the lane is calculated and expressed as: (8); Where L represents the distance from the dynamic electronic display screen to the lane. Let N represent the average vehicle speed and N represent the number of vehicles currently in the queue. Indicates processing delay for a single vehicle; Finally, the ratio of manual or autonomous lanes is dynamically adjusted, as shown below: (9); Where P represents the probability of a special situation vehicle, and K represents the maximum traffic capacity coefficient of the lane.

[0035] The CPC card self-service payment lane is equipped with RFID card readers, payment terminals and anomaly handling units for self-service payment.

[0036] Specifically, the RFID reader uses a time-division multiplexing protocol, which is represented as follows: (10); (11); Among them, T c T represents the total cycle time. r T represents the data retrieval time. e T represents the data erasure time. s N represents standby time, k represents the safety factor, and N represents the safety factor. c R represents the amount of data in the card. t Indicates the transmission rate.

[0037] Specifically, when the exception handling unit detects a CPC card read failure, it initiates a three-level rollback mechanism, with the rollback level of the three-level rollback mechanism represented as follows: The payment is completed using the pre-transaction amount at level 1; billing is based on the vehicle license plate's historical transaction records at level 2; and manual intervention is triggered at level 3, and the process is transferred to manual processing.

[0038] Example 2, reference Figure 2 As shown, the business processing flow of the unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees in Embodiment 1 is as follows: Checkpoints are set up on the ramps to capture the license plates of all vehicles entering the toll station, thereby identifying all vehicles entering the toll plaza.

[0039] Based on the license plate information captured by the checkpoint, the transaction records of the upstream gantry are queried to obtain the vehicle's card information (including card number, vehicle type, accumulated toll amount, and accumulated trip) and entry information (entry station code, entry lane number, and entry time). Simultaneously, the online tolling interface provided by the provincial center is called (with input parameters: card number, license plate, and entry information) to calculate the estimated toll for all vehicles. Furthermore, by integrating the ramp toll amount and card-related information read from the upstream gantry, the cumulative toll amount for each vehicle's license plate is calculated as a backup tolling method.

[0040] Simultaneously, pre-transaction status is determined: based on the legality of vehicle information, vehicles are classified into three categories: normal ETC vehicles, normal CPC card vehicles, and special vehicles (including blacklisted vehicles, vehicles with abnormal billing, vehicles without entry information, and vehicles with insufficient balance). Abnormal vehicles are screened and guided to manual lanes through the control guidance screen, thereby reducing the number of abnormal vehicles entering self-service lanes and improving the efficiency of self-service lanes; normal vehicles are pushed with estimated fee information to the self-service payment machine.

[0041] When a vehicle with a valid CPC card enters the fast self-service transaction lane, the license plate recognition device immediately displays the payment amount after recognizing the license plate, prompts the user to insert or drop the card, and simultaneously enables QR code payment, thereby shortening the payment time.

[0042] After the self-service payment device detects the card, it will handle the situation in three ways: if the card information is read and the payment is successful, the card information will be cleared, the gate will be raised after the card is collected; if the card information is not read and the payment is successful, the gate will be raised after the card is collected; if the card information is not read and the estimated fee is not billed, the card collection device will flash a red light to prompt manual handling, and the operator will be guided out of the lane to handle the situation.

[0043] It should be noted that the transaction status determination and processing are as follows: The system classifies vehicles based on the legality of their information. For abnormal ETC vehicles (determined to be blacklisted, have insufficient balance, have removed tags, or have no entry information) and abnormal CPC card users (determined to have abnormal routes, abnormal accumulated amounts, or have no entry information), the system guides them directly to the manual lane via a guidance screen. For normal CPC card vehicles, the estimated fee information is pushed to the self-service payment machine. After the vehicle enters the fast self-service transaction lane, the license plate recognition device identifies the license plate and displays the payment amount, prompting the user to insert or drop the card, and simultaneously enabling QR code payment. After the self-service payment device detects the card, it processes it according to different situations. If the card information is read and the payment is successful, the card information is cleared, the gate is raised after the card is collected, and the vehicle is allowed to pass. If the card information cannot be read but the payment is successful, the card is collected directly and the gate is raised, and the vehicle is allowed to pass. If the card information cannot be read and no billing information for the estimated fee is obtained, the card collection device flashes a red light to prompt manual processing, and the vehicle is manually guided out of the lane for processing. For vehicles with special circumstances, the system controls the special circumstances guidance screen on the ramp to guide them to the manual lane for special circumstances processing.

[0044] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An unattended toll collection system based on the fusion of estimated fees for mainline gantry and ramp cards, characterized in that: This includes the front mainline gantry, ramp online estimated toll gates, ramp emergency guidance screens, fast CPC card self-service toll lanes, and tolling platform; A mainline gantry module is installed at the front mainline gantry, and the mainline gantry module is used to collect vehicle passage data and upload it to the tolling platform. The online toll collection point for ramps includes a ramp collection point module and a pre-transaction platform; The ramp checkpoint module is deployed at the entrance of the toll station ramp for license plate information recognition and data fusion processing; The pre-transaction platform connects the billing platform and the local database to perform online billing and media verification; The ramp emergency guidance screen is equipped with a dynamic guidance module, including an emergency classification model and a variable information board; The fast CPC card self-service payment lane is equipped with an RFID card reader, a payment terminal, and an anomaly handling unit for self-service payment.

2. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 1, characterized in that, The vehicle passage data collected by the mainline gantry module includes the vehicle CPC card number, vehicle type, entrance station code, entrance time, and cumulative mileage. The ramp checkpoint module simultaneously collects visible light image data of the vehicles and performs correlation matching between the mainline gantry data and the real-time captured data. The matching degree of this correlation matching is expressed as: （1）； in, SIM ( x () represents the string similarity function. This represents the difference between the timestamp of the gantry data and the capture time. This indicates the average travel time. This indicates the maximum permissible time difference threshold. , This represents the dynamic weighting coefficient, and: （2）。 3. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 1, characterized in that, The computational representation of the data fusion processing is as follows: （3）； in, This indicates the final billing amount. Indicates the online billing result. Indicates the amount read from the medium. , Let represent the dynamic weighting coefficients, and satisfy: （4）。 4. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 3, characterized in that, The dynamic weighting coefficients are obtained through credibility assessment and are expressed as follows: （5）； （6）； Among them, C online and C medium These represent the confidence scores for online billing and media reading, respectively.

5. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 1, characterized in that, The special incident classification model uses a decision tree algorithm to identify and classify abnormal vehicle billing events in real time; the variable message sign uses a dynamic electronic display screen to provide drivers with guidance information on choosing payment channels.

6. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 5, characterized in that, The classification criteria of the special situation classification model are as follows: If the entry information of a vehicle waiting to pay at the ramp checkpoint module is missing or matches the blacklist, a Level 1 warning will be issued. If the real-time online billing value minus the baseline billing value exceeds the billing difference threshold, it indicates that there is a deviation between the real-time billing and the expected baseline, and a secondary traffic splitting process will be performed.

7. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 6, characterized in that, When the special situation classification model performs classification processing; First, obtain the vehicle arrival time through the ramp checkpoint, denoted as t. arrival ; Then, the theoretical travel time from the dynamic electronic display screen to the lane is calculated and expressed as: （8）； Where L represents the distance from the dynamic electronic display screen to the lane. Let N represent the average vehicle speed and N represent the number of vehicles currently in the queue. Indicates the processing delay for a single vehicle; Finally, the ratio of manual or autonomous lanes is dynamically adjusted, as shown below: （9）； Where P represents the probability of a special situation vehicle, and K represents the maximum traffic capacity coefficient of the lane.

8. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 1, characterized in that, When the exception handling unit detects a CPC card reading failure, it initiates a three-level rollback mechanism. The rollback levels of the three-level rollback mechanism are as follows: Level 1: Complete payment using the pre-transaction amount; Level 2: Billing based on historical transaction records associated with the license plate; Level 3: Triggering a manual intervention protocol to switch to manual processing.

9. The unattended toll collection system based on the fusion of mainline gantry and ramp card estimated fees as described in claim 1, characterized in that, The RFID reader uses a time-division multiplexing protocol, as shown below: （10）； （11）； Among them, T c T represents the total cycle time. r T represents the data retrieval time. e T represents the data erasure time. s N represents standby time, k represents the safety factor, and N represents the safety factor. c R represents the amount of data in the card. t Indicates the transmission rate.

Images