An Artificial Intelligence-Based Toll Station Ramp Control Method, Device, and Medium

Abstract

The present application discloses an artificial-intelligence-based toll station ramp control method, device and medium, which collect video streams and sensor data of a section to be controlled, and preprocess the video streams to determine key frame images; analyze the key frame images in combination with sensor data to determine traffic data and traffic impact factor values corresponding to the section to be controlled, and input the traffic data and traffic impact factor values into a trained traffic flow prediction model to output an estimated traffic condition corresponding to a preset time period; generate a recommended planned path for the corresponding driving vehicle based on the vehicle driving requirements and the estimated traffic condition obtained in real time, and optimize the estimated traffic condition based on the confirmation status of the corresponding driving vehicle for the recommended planned path; determine the congestion level corresponding to the optimized estimated traffic condition, and determine a ramp control strategy corresponding to the congestion level based on a preset ramp control library, so as to perform ramp control on the toll station according to the ramp control strategy.

Description

Technical Field

[0001] This application relates to the field of traffic control technology, and particularly to a toll station ramp control method, device, and medium based on artificial intelligence. Background Art

[0002] With the continuous growth of the global economy and the acceleration of the urbanization process, highways, as an important infrastructure connecting major cities and economic regions, have continuously increased in scale and complexity, becoming an indispensable traffic artery in modern society. The rapid development of highways has not only greatly promoted the flow of people and the exchange of goods, but also put forward higher requirements for traffic management and services. Toll stations, as key nodes and traffic hubs in the highway network, undertake multiple functions such as vehicle toll collection, traffic flow regulation, and safety supervision. Their operation efficiency and management level directly affect the smoothness of the entire road network and the user experience.

[0003] Traditional toll station ramp management methods mostly rely on manual toll collection, semi-automatic toll collection systems, and basic video surveillance. In this mode, the real-time monitoring of traffic flow, the rapid identification of vehicle types, and the emergency response capabilities for special traffic conditions (such as accidents and bad weather) are relatively limited. Manual operations are easily affected by factors such as fatigue and emotions, resulting in low efficiency; while simple monitoring systems often lack intelligent analysis functions and are difficult to accurately evaluate traffic conditions in real time and make corresponding adjustments. Under the combined action of these factors, traffic congestion is easily caused, and the vehicle passing time is prolonged, which not only causes waste of time and resources, but also may increase the risk of traffic accidents due to vehicle backlogs, affecting road safety and the public travel experience.

[0004] In recent years, with the rapid development of advanced technologies such as artificial intelligence, big data, and cloud computing, intelligent transportation systems (ITS) have become a new way to solve modern traffic management problems. The application of artificial intelligence technology in the field of traffic management, such as intelligent traffic signal control, vehicle autonomous driving, and road condition prediction, has shown remarkable results, improving the operation efficiency and safety of traffic systems. However, for the specific scenario of highway toll station ramps, although there have been some preliminary explorations, such as automatic license plate recognition and the popularization of the ETC system, there is still a lack of a comprehensive and efficient intelligent control solution as a whole. Summary of the Invention

[0005] Embodiments of this application provide a toll station ramp control method, device, and medium based on artificial intelligence to solve the above technical problems.

[0006] On the one hand, embodiments of this application provide a toll station ramp control method based on artificial intelligence, including:

[0007] Based on the roadside devices preset for the section to be controlled, collect the video stream and sensor data of the section to be controlled, and preprocess the video stream to determine key frame images;

[0008] Analyze the key frame images in combination with the sensor data, determine the traffic data and traffic impact factor values corresponding to the section to be controlled, and input the traffic data and the traffic impact factor values into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to a preset time period;

[0009] Based on the vehicle driving requirements obtained in real time and the estimated traffic conditions, generate a recommended planned path for the corresponding driving vehicle, and optimize the estimated traffic conditions based on the confirmation status of the driving vehicle for the corresponding recommended planned path;

[0010] Determine the congestion level corresponding to the optimized estimated traffic conditions, and based on the preset ramp control library, determine the ramp control strategy corresponding to the congestion level, so as to perform ramp control on the toll station according to the ramp control strategy.

[0011] In an implementation manner of the present application, based on the roadside devices preset for the section to be controlled, collect the video stream and sensor data of the section to be controlled, and preprocess the video stream to determine key frame images, which specifically includes:

[0012] Through the cameras and sensors preset for the section to be controlled, collect the video stream and sensor data corresponding to the section to be controlled in real time, and receive the video stream and sensor data uploaded by the cameras and the sensors;

[0013] Analyze the video stream, identify the latter image in two adjacent images with inconsistent front and back information in the video stream, so as to determine the key frame images corresponding to the video stream.

[0014] In an implementation manner of the present application, analyze the key frame images in combination with the sensor data, determine the traffic data and traffic impact factor values corresponding to the section to be controlled, which specifically includes:

[0015] Obtain the historical ramp control records corresponding to the section to be controlled, and based on the historical ramp control records, determine the traffic impact factors corresponding to the section to be controlled; wherein, the traffic impact factors include time periods, holidays, and weather, and the time periods include peak periods, off-peak periods, and regular periods;

[0016] Segment the key frame images to determine the driving vehicles in the key frame images, and identify the vehicle types, vehicle colors, and license plate numbers corresponding to the driving vehicles;

[0017] Determine the vehicle speed corresponding to the traveling vehicle according to the sensor data corresponding to the adjacent road sections in the road section to be controlled, so as to obtain the traffic data corresponding to the road section to be controlled;

[0018] Obtain the data collection time corresponding to the traffic data, so as to determine the traffic impact factor value corresponding to the road section to be controlled according to the data collection time and in combination with the sensor data.

[0019] In an implementation manner of the present application, before inputting the traffic data and the traffic impact factor value into a trained traffic flow prediction model to output the estimated traffic condition corresponding to a preset time period, the method further includes:

[0020] Construct a traffic flow prediction model corresponding to the road section to be controlled based on a deep learning algorithm;

[0021] Obtain the historical traffic data corresponding to the road section to be controlled and the historical traffic impact factor value corresponding to the historical traffic data, and mark the traffic condition in the next time period corresponding to the historical traffic data and the historical traffic impact factor value;

[0022] Input the historical traffic data and the historical traffic impact factor value into the traffic flow prediction model, and output the corresponding estimated traffic condition until the estimated traffic condition matches the marked traffic condition in the next time period, and complete the training of the traffic flow prediction model.

[0023] In an implementation manner of the present application, inputting the traffic data and the traffic impact factor value into a trained traffic flow prediction model to output the estimated traffic condition corresponding to a preset time period specifically includes:

[0024] Determine whether the data formats corresponding to the traffic data and the traffic impact factor value are consistent with the input format corresponding to the traffic flow prediction model. If not, convert the traffic data and the traffic impact factor value into the input format;

[0025] Input the traffic data and the traffic impact factor value in the input format into a trained traffic flow prediction model, and output the estimated traffic condition corresponding to the road section to be controlled within a preset time period;

[0026] Verify the estimated traffic condition. If the verification fails, correct the estimated traffic condition based on a preset expert library to obtain the target estimated traffic condition.

[0027] In an implementation manner of the present application, generating a recommended planned path for the corresponding traveling vehicle based on the vehicle driving requirements obtained in real time and the estimated traffic condition specifically includes:

[0028] According to the predicted traffic conditions, determine whether there are potential risks in the section to be controlled. If so, obtain the driving demand information corresponding to the vehicles traveling in the section to be controlled; wherein, the driving demand information includes the destination, the expected arrival time, and the preferred route.

[0029] Based on the driving demand information and in combination with the predicted traffic conditions, generate at least one recommended planned route corresponding to the vehicle traveling; wherein, the recommended planned route includes the estimated travel duration, the estimated cost, and the congestion situation.

[0030] Send the recommended planned route to the corresponding vehicle traveling through wireless communication for the vehicle personnel to select and confirm.

[0031] In an implementation manner of the present application, based on the confirmation status of the corresponding recommended planned route by the vehicle traveling, optimize the predicted traffic conditions, specifically including:

[0032] Receive the confirmation situations of all vehicles in the section to be controlled for the corresponding recommended planned route, and dynamically update the traffic impact factor value corresponding to the section to be controlled according to the feedback data.

[0033] According to the feedback data and in combination with the updated traffic impact factor value, perform iterative optimization on the model parameters of the traffic flow prediction model, so as to generate the target predicted traffic conditions through the optimized traffic flow prediction model.

[0034] In an implementation manner of the present application, determine the congestion level corresponding to the optimized predicted traffic conditions, and based on the preset ramp control library, determine the ramp control strategy corresponding to the congestion level, specifically including:

[0035] Based on the preset congestion level division standard and according to the optimized predicted traffic conditions, determine the congestion level corresponding to the section to be controlled.

[0036] According to the congestion level, determine the corresponding ramp control strategy in the preset ramp control library, and determine the actual situation corresponding to the road to be controlled according to the traffic data collected in real time for the section to be controlled.

[0037] According to the actual situation, adjust the ramp control strategy and generate the target ramp control strategy corresponding to the section to be controlled.

[0038] On the other hand, the embodiment of the present application further provides a toll ramp control device based on artificial intelligence, and the device includes:

[0039] At least one processor;

[0040] And a memory communicatively connected to the at least one processor;

[0041] Among them, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a toll station ramp control method based on artificial intelligence as described above.

[0042] On the other hand, an embodiment of the present application also provides a non-volatile computer storage medium storing computer-executable instructions, which when executed, implement a toll station ramp control method based on artificial intelligence as described above.

[0043] The embodiment of the present application provides a toll station ramp control method, device and medium based on artificial intelligence, which at least include the following beneficial effects:

[0044] Through efficient roadside device collection and preprocessing technologies, traffic videos and sensor data of the section to be controlled can be obtained in real time and accurately, providing a solid foundation for subsequent traffic analysis and prediction; by fusing video and sensor data, a more comprehensive understanding of the traffic conditions of the section to be controlled can be achieved, including key indicators such as vehicle flow, speed, and density, as well as influencing factors such as weather and road conditions; by inputting this data into a trained traffic flow prediction model, the traffic conditions in the next period of time can be predicted more accurately, providing a basis for traffic management and decision-making; by obtaining the driving needs of vehicles in real time and combining the estimated traffic conditions, a personalized recommended planned route can be provided for the driving vehicles, which helps to reduce the vehicle driving time, improve the road traffic efficiency, and at the same time reduce the possibility of traffic congestion and accidents; by collecting the confirmation status of the recommended planned route by the driving vehicles, the estimated traffic conditions can be timely feedback and adjusted, which helps to improve the accuracy and practicality of traffic prediction and better meet the actual traffic needs; by determining the congestion level of the optimized estimated traffic conditions, the current and future traffic conditions can be more accurately evaluated, and combined with the preset ramp control library, corresponding ramp control strategies can be quickly formulated and executed to effectively relieve traffic congestion and improve the road traffic capacity; by implementing the ramp control strategy, the vehicle flow entering the toll station can be reasonably regulated to avoid excessive congestion at the toll station entrance, which helps to improve the toll station traffic efficiency, reduce the vehicle waiting time, and at the same time enhance the overall traffic fluency and safety. Description of the Drawings

[0045] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation to the present application. In the drawings:

[0046] Figure 1Schematic flowchart of a toll station ramp control method based on artificial intelligence provided by an embodiment of the present application;

[0047] Figure 2 Internal structure schematic diagram of a toll station ramp control device based on artificial intelligence provided by an embodiment of the present application. Specific embodiments

[0048] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the scope of protection of the present application.

[0049] The following will detail the technical solutions provided by each embodiment of the present application with reference to the drawings.

[0050] Figure 1 Schematic flowchart of a toll station ramp control method based on artificial intelligence provided by an embodiment of the present application.

[0051] The implementation of the analysis method involved in the embodiments of the present application can be a terminal device or a server, and the present application does not make special restrictions on this. For the convenience of understanding and description, the following embodiments will be described in detail taking the server as an example.

[0052] It should be noted that the server can be a single device or a system composed of multiple devices, that is, a distributed server, and the present application does not make specific limitations on this.

[0053] As Figure 1 shown, a toll station ramp control method based on artificial intelligence provided by an embodiment of the present application includes:

[0054] 101. Based on the roadside devices preset for the section to be controlled, collect the video stream and sensor data of the section to be controlled, and preprocess the video stream to determine the key frame images.

[0055] Specifically, in an embodiment of the present application, based on the roadside devices preset for the section to be controlled, collect the video stream and sensor data of the section to be controlled, and preprocess the video stream to determine the key frame images, which specifically includes:

[0056] Through the cameras and sensors preset for the section to be controlled, collect the video stream and sensor data corresponding to the section to be controlled in real time, and receive the video stream and sensor data uploaded by the cameras and sensors;

[0057] Analyze the video stream to identify the latter image among two adjacent images with inconsistent information before and after in the video stream, so as to determine the key frame image corresponding to the video stream.

[0058] In one embodiment, deploy hardware facilities such as cameras, radars, infrared sensors, and pressure sensing devices at the entrances and exits of toll station ramps to achieve real-time collection of information such as vehicle types, vehicle speeds, and traffic volumes. Deploy a license plate recognition system inside the toll station area to monitor vehicle information (license plate numbers, vehicle types, driving directions, etc.) entering and leaving the toll station. Use a 5G high-speed communication network to transmit the collected data to the back-end server or cloud platform to ensure the real-time nature and integrity of the data.

[0059] The server cleans the video stream and sensor data, including: removing abnormal data, filling in missing values, normalizing vehicle speed and traffic volume data, etc., to ensure the accuracy and consistency of the data, and uses the random forest algorithm to predict missing values. Use other non-missing features as inputs to predict the missing values. At the same time, classify the historical traffic data and mark it according to time periods such as peak hours, non-peak hours, and holidays for subsequent training and prediction of the AI model.

[0060] In one embodiment, in order to more effectively manage the sections of the expressway to be controlled in a certain city, the traffic management department decides to adopt a set of advanced intelligent monitoring systems. The system mainly includes a plurality of high-definition cameras and various sensors preset along the sections to be controlled, such as vehicle detectors, speed sensors, and meteorological sensors. These cameras and sensors are carefully arranged at key positions to collect the corresponding video stream and various sensor data of the sections to be controlled in real time. The high-definition cameras are responsible for capturing the real-time traffic pictures of the sections, while the sensors are responsible for collecting key information such as vehicle flow, speed, vehicle type distribution, weather, and road surface conditions. The collected video stream and sensor data are uploaded to the central monitoring center in real time through a wireless network to ensure the timeliness and accuracy of the data.

[0061] In the central monitoring center, a dedicated video analysis software deeply analyzes the received video stream. The software uses advanced image recognition algorithms and can automatically detect two adjacent images with inconsistent information before and after in the video stream. For example, when a vehicle suddenly changes lanes or stops, this change will produce images with inconsistent information before and after in the video stream. The software can accurately identify this change and mark the latter image as the key frame image. In this way, the system can automatically screen out the most important image frames in the video stream, providing a key basis for subsequent traffic analysis and prediction.

[0062] The determined key-frame images are used for further analysis of traffic conditions. Combining sensor data, the system can more accurately understand the real-time traffic conditions of the section to be controlled, including vehicle behavior, traffic flow changes, and potential safety hazards. This information is used to generate real-time traffic reports and warnings, helping traffic management departments take necessary measures in a timely manner, such as adjusting traffic signals, dispatching rescue vehicles, or implementing ramp control, to ensure the safety and smoothness of the road.

[0063] 102. Analyze the key-frame images in combination with sensor data to determine the traffic data and traffic impact factor values corresponding to the section to be controlled, and input the traffic data and traffic impact factor values into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to the preset time period.

[0064] Specifically, in an embodiment of the present application, analyzing the key-frame images in combination with sensor data to determine the traffic data and traffic impact factor values corresponding to the section to be controlled specifically includes:

[0065] Obtain the historical ramp control records corresponding to the section to be controlled, and based on the historical ramp control records, determine the traffic impact factors corresponding to the section to be controlled; among them, the traffic impact factors include time period, holiday, and weather, and the time period includes peak period, off-peak period, and regular period;

[0066] Segment the key-frame images to determine the driving vehicles in the key-frame images, and identify the vehicle types, vehicle colors, and license plate numbers corresponding to the driving vehicles;

[0067] According to the sensor data corresponding to the driving vehicles on the adjacent sections in the section to be controlled, determine the vehicle speeds corresponding to the driving vehicles to obtain the traffic data corresponding to the section to be controlled;

[0068] Obtain the data collection time corresponding to the traffic data, and based on the data collection time and in combination with sensor data, determine the traffic impact factor values corresponding to the section to be controlled.

[0069] In an embodiment, in order to more finely manage the section to be controlled of the expressway in a certain city, the traffic management department introduced an intelligent traffic analysis system. The system aims to provide a scientific basis for ramp control by analyzing historical data and real-time data. The system first obtains the historical ramp control records corresponding to the section to be controlled. These records include detailed information such as the implementation time, reasons, and effects of ramp control in the past period. Based on these historical data, the system uses data mining technology to determine the main factors affecting the traffic of the section to be controlled, including time period (such as peak period, off-peak period, and regular period), holiday, and weather conditions. These factors are used as important references for subsequent traffic analysis and prediction.

[0070] The system receives key-frame images from preset cameras on the section to be controlled. These images are automatically screened by the system according to video stream analysis and contain important traffic information. By segmenting the key-frame images, the system can accurately identify the moving vehicles in the images. Through further image recognition technology, the system can also determine the vehicle type, vehicle color, and license plate number of each vehicle, providing detailed information for subsequent traffic data analysis.

[0071] The system not only analyzes key-frame images but also combines sensor data corresponding to the moving vehicles on adjacent sections in the section to be controlled. These data include key information such as the passing time and speed of the vehicles. By comparing the sensor data of adjacent sections, the system can accurately calculate the vehicle speed of the moving vehicles before entering the section to be controlled, thereby obtaining the traffic data corresponding to the section to be controlled. The system obtains the data acquisition moment corresponding to the traffic data. This moment is the time point when the system records the data and is used to analyze subsequent changes in traffic conditions. Combining the data acquisition moment and sensor data, the system can determine the traffic impact factor values at a specific time point in the section to be controlled. These values reflect the degree to which the section traffic conditions are affected by factors such as time periods, holidays, and weather, providing a scientific basis for ramp control decisions.

[0072] In an embodiment of the present application, before inputting the traffic data and traffic impact factor values into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to a preset time period, it further includes:

[0073] Based on deep learning algorithms, construct a traffic flow prediction model corresponding to the section to be controlled;

[0074] Obtain the historical traffic data corresponding to the section to be controlled and the historical traffic impact factor values corresponding to the historical traffic data, and label the traffic conditions in the next time period corresponding to the historical traffic data and historical traffic impact factor values;

[0075] Input the historical traffic data and historical traffic impact factor values into the traffic flow prediction model and output the corresponding estimated traffic conditions until the estimated traffic conditions match the labeled traffic conditions in the next time period, completing the training of the traffic flow prediction model.

[0076] In an embodiment, in order to improve the traffic management level of the section to be controlled on a certain urban arterial road, the traffic management department decides to use deep learning technology to construct a traffic flow prediction model. The model aims to learn the changing rules of traffic flow through historical data and thus accurately predict the traffic conditions in future time periods. The traffic management department first constructs a traffic flow prediction model corresponding to the section to be controlled based on deep learning algorithms such as long short-term memory network (LSTM) or convolutional neural network (CNN). These algorithms have excellent performance in processing time series data and image data and can capture complex patterns and trends in traffic flow.

[0077] Next, the system obtains the historical traffic data corresponding to the section to be controlled, including key indicators such as vehicle flow, speed, and density. At the same time, it also obtains the historical traffic impact factor values corresponding to these historical traffic data, such as time periods, holidays, weather conditions, etc. In order to train the model, these historical data need to be labeled. The specific approach is to correspond the historical traffic data and impact factor values at each time point with the traffic conditions in the next time period to form a training data set. In this way, the model can learn the mapping relationship from historical data to future traffic conditions.

[0078] The labeled historical traffic data and impact factor values are input into the traffic flow prediction model. The model gradually approaches the real traffic conditions by continuously learning and adjusting parameters. During the training process, the system uses appropriate loss functions and optimization algorithms, such as mean square error (MSE) and stochastic gradient descent (SGD), etc., to evaluate the prediction performance of the model and optimize the model parameters. The training process continues until the estimated traffic conditions of the model match the labeled traffic conditions in the next time period, that is, a high prediction accuracy is achieved. At this time, it can be considered that the model has successfully learned the change rules of traffic flow.

[0079] The trained traffic flow prediction model can be used to predict the future traffic conditions of the section to be controlled in real time. By inputting the current traffic data and impact factor values, the model can output the estimated traffic conditions in the next time period, that is, traffic conditions such as the flow peak time and congestion duration within a certain future time period, providing a scientific decision-making basis for traffic management departments. For example, before the peak period arrives, the model can predict the congestion situation of the section, thereby helping the traffic management department take ramp control measures in advance, such as adjusting signal timing, restricting vehicle entry, etc., to relieve traffic pressure. According to the real-time collected data and historical data, continuously optimize this AI model so that it can more accurately predict future traffic flows and dynamically adjust the ramp control strategy.

[0080] In an embodiment of the present application, the traffic data and traffic impact factor values are input into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to the preset time period, specifically including:

[0081] Determine whether the data format corresponding to the traffic data and traffic impact factor values is consistent with the input format corresponding to the traffic flow prediction model. If not, convert the traffic data and traffic impact factor values into the input format;

[0082] Input the traffic data and traffic impact factor values in the input format into the trained traffic flow prediction model, and output the estimated traffic conditions corresponding to the section to be controlled within the preset time period;

[0083] Verify the estimated traffic conditions. If the verification fails, correct the estimated traffic conditions based on a preset expert database to obtain the target estimated traffic conditions.

[0084] In one embodiment, in order to more accurately predict the future traffic conditions in the area to be controlled on a busy road section in a certain city, the traffic management department decides to adopt an advanced traffic flow prediction model and combine data format conversion and expert verification mechanisms to ensure the accuracy of the prediction. First, the system collects traffic data and traffic impact factor values from various sensors and cameras. These data may exist in different formats, such as CSV, JSON, or database records, etc. The system checks whether the format of these data is consistent with the input format required by the traffic flow prediction model. If inconsistencies are found, the system will automatically perform data format conversion to ensure that the traffic data and impact factor values can be correctly read and processed by the model. This step is crucial for avoiding data errors and improving the prediction accuracy of the model.

[0085] Once the data format conversion is completed, the system inputs the traffic data and traffic impact factor values that conform to the input format into the trained traffic flow prediction model. The model uses these data and combines the traffic flow laws learned internally to output the estimated traffic conditions corresponding to the road section to be controlled within a preset time period. These estimated data may include key indicators such as vehicle flow, average speed, and congestion index.

[0086] To ensure the accuracy of the prediction, the system verifies the output estimated traffic conditions. The verification process may include comparison with historical data, trend analysis, and reasonableness checks, etc. If the verification fails, that is, there is a large deviation between the estimated traffic conditions and the actual situation or it does not conform to logic, the system will trigger a correction mechanism. This mechanism is based on a preset expert database, which contains the knowledge and experience of traffic experts. The system corrects the estimated traffic conditions according to the rules and suggestions in the expert database to obtain a more accurate and reliable target estimated traffic conditions. The correction process may involve adjusting prediction parameters, introducing new impact factors, or adopting a more complex prediction model, etc. Finally, the target estimated traffic conditions after verification and correction are used to guide the decision-making of the traffic management department. For example, according to the prediction results, the management department can adjust traffic signals in advance, deploy police forces, or implement ramp control measures to relieve traffic congestion and improve road traffic efficiency.

[0087] 103. Generate a recommended planned path for the corresponding driving vehicle based on the real-time obtained driving requirements of the vehicle and the estimated traffic conditions, and optimize the estimated traffic conditions based on the confirmation status of the driving vehicle for the corresponding recommended planned path.

[0088] Specifically, in one embodiment of the present application, generating a recommended planned path for the corresponding driving vehicle based on the real-time obtained driving requirements of the vehicle and the estimated traffic conditions specifically includes:

[0089] Based on the estimated traffic conditions, determine whether there are potential risks in the section to be controlled. If so, obtain the driving demand information corresponding to the vehicles traveling in the section to be controlled; among them, the driving demand information includes the destination, the expected arrival time, and the preferred route.

[0090] Based on the driving demand information and combined with the estimated traffic conditions, generate at least one recommended planned route corresponding to the vehicle traveling; among them, the recommended planned route includes the estimated travel duration, the estimated cost, and the congestion situation.

[0091] Send the recommended planned route to the corresponding vehicle traveling through wireless communication for the vehicle personnel to select and confirm.

[0092] In one embodiment, in order to optimize the traffic management of the section to be controlled in the core area of a certain city and ensure that the vehicles traveling can pass through efficiently and safely, the traffic management department has developed an intelligent route planning system. This system combines the estimated traffic conditions and the driving demand information to provide personalized driving route suggestions for vehicles. First, based on the estimated traffic conditions, conduct a potential risk assessment on the section to be controlled. The factors considered in the assessment include vehicle flow, average speed, congestion index, and historical accident data, etc. If the system determines that there are potential risks, such as severe congestion or frequent accidents, subsequent steps will be triggered to obtain the driving demand information of the vehicles traveling.

[0093] The system communicates with the vehicles traveling in the section to be controlled through wireless communication methods, such as vehicle networking or mobile communication networks. Obtain the driving demand information from the vehicles, including the destination, the expected arrival time, and the preferred route, etc. These information are crucial for generating a planned route that meets the driver's expectations. Based on the obtained driving demand information, the system combines the estimated traffic conditions to generate at least one recommended planned route for each vehicle traveling. The planned route contains detailed driving guidance, including the estimated travel duration, the estimated cost (such as fuel cost or electricity cost), and the congestion situation along the way. These information help the driver make a wise decision.

[0094] The system sends the generated recommended planned route to the corresponding vehicle traveling through wireless communication. The vehicle personnel can view these routes on the in-vehicle display screen or mobile device and select and confirm according to their own needs and preferences. The confirmed route will be used as the driving guidance for the vehicle. During the vehicle's travel, the system continuously monitors the changes in traffic conditions and updates and optimizes the planned route in real time according to the actual situation. If new congestion or accident situations are encountered, the system will immediately notify the vehicle personnel and provide alternative route suggestions to ensure the safety and efficiency of the travel.

[0095] In one embodiment of the present application, based on the confirmation status of a traveling vehicle for a corresponding recommended planned path, the predicted traffic conditions are optimized, specifically including:

[0096] Receiving the confirmation situations of all vehicles in the to-be-controlled section for the corresponding recommended planned paths, and dynamically updating the traffic impact factor values corresponding to the to-be-controlled section according to the feedback data;

[0097] According to the feedback data, and in combination with the updated traffic impact factor values, iteratively optimizing the model parameters of the traffic flow prediction model, so as to generate the target predicted traffic conditions through the optimized traffic flow prediction model.

[0098] In one embodiment, in order to further improve the traffic management efficiency of an important to-be-controlled section in a certain city and ensure the accuracy and practicability of the traffic flow prediction model, the traffic management department implemented a dynamic feedback and optimization system. This system collects the confirmation situations of vehicles for the recommended planned paths, dynamically updates the traffic impact factor values, and iteratively optimizes the parameters of the traffic flow prediction model. First, it receives the confirmation situations of all vehicles in the to-be-controlled section for the corresponding recommended planned paths. These feedback data include whether the vehicle travels according to the recommended path, the actual traffic conditions during the travel, and the driver's satisfaction, etc. By collecting these feedback data, the system can understand the actual effect of the recommended path in real time, as well as the suggestions and opinions of the driver on the path planning.

[0099] Based on the collected feedback data, the system dynamically updates the traffic impact factor values corresponding to the to-be-controlled section. These impact factor values may include the actual traffic capacity of the section, the driving habits of the drivers, the signal timing of the traffic lights, etc. By updating the traffic impact factor values, the system can more accurately reflect the real-time traffic conditions of the section and provide a more reliable data basis for subsequent traffic flow prediction. The system combines the updated traffic impact factor values and the feedback data to iteratively optimize the model parameters of the traffic flow prediction model. The optimization process may include adjusting the input features of the model, modifying the prediction algorithm of the model, or introducing new impact factors, etc. Through iterative optimization, the system can continuously improve the accuracy and generalization ability of the traffic flow prediction model, so that it can better adapt to the changes in the actual traffic conditions.

[0100] The traffic flow prediction model after iterative optimization is used to generate the target predicted traffic conditions. These predicted data include key indicators such as vehicle flow, average speed, congestion index, etc. in the future time period. The target predicted traffic conditions provide a scientific decision-making basis for the traffic management department, which helps it formulate more effective ramp control measures and traffic guidance plans. The system continuously monitors the traffic conditions of the to-be-controlled section and regularly collects feedback data for model update and optimization. Through continuous improvement and perfection, the system can gradually improve the traffic management efficiency and the driver's satisfaction.

[0101] 104. Determine the congestion level corresponding to the predicted traffic condition after optimization, and based on a preset ramp control library, determine the ramp control strategy corresponding to the congestion level, so as to perform ramp control on the toll station according to the ramp control strategy.

[0102] Specifically, in an embodiment of the present application, determining the congestion level corresponding to the predicted traffic condition after optimization, and based on a preset ramp control library, determining the ramp control strategy corresponding to the congestion level specifically includes:

[0103] Based on a preset congestion level division standard and according to the predicted traffic condition after optimization, determine the congestion level corresponding to the section to be controlled;

[0104] According to the congestion level, determine the corresponding ramp control strategy in the preset ramp control library, and according to the traffic data collected in real time for the section to be controlled, determine the actual situation corresponding to the road to be controlled;

[0105] According to the actual situation, adjust the ramp control strategy and generate the target ramp control strategy corresponding to the section to be controlled.

[0106] In an embodiment, based on the output of the prediction model, perform real-time dynamic control on the ramp and the toll station. If the system detects that the traffic flow is about to reach the congestion threshold, automatically adjust the traffic signal duration of the ramp and appropriately extend the green light time to improve the traffic efficiency.

[0107] It should be noted that the key indicators output by the model in the embodiment of the present application include predicted traffic flow, vehicle speed distribution, queue length, traffic flow peak period, and lane utilization rate. The predicted traffic flow is used to represent the estimated number of vehicles passing through in the next few minutes or hours; the vehicle speed distribution is used to represent the vehicle speed changes at different time periods on the ramp and can reflect the smoothness of the traffic flow; the queue length is used to represent the number and length of queuing vehicles predicted on the ramp and measure the congestion degree; the traffic flow peak period is used to represent the time and duration of the future flow peak; the lane utilization rate is used to represent the utilization rate of each lane at the ramp and the toll station entrance and measure whether the vehicles are evenly distributed.

[0108] The adjustment of the signal duration is for the average waiting time, maximum queue length, passing rate, alternating control, and emergency response.

[0109] Average waiting time: Reducing the average waiting time of vehicles on the ramp is one of the main objectives. If the model predicts an increase in traffic flow, the green light duration of the signal can be appropriately extended so that more vehicles can pass through in one cycle.

[0110] Maximum queue length: Control the queue length by adjusting the green light duration to avoid the queue exceeding the ramp or affecting the traffic on the main road.

[0111] Passing rate: The ratio of the number of vehicles passing through the ramp to the time, which measures the traffic flow through the ramp per unit time. Adjusting the signal light duration can increase this ratio.

[0112] Alternating control: If the traffic flow at some ramp entrances is large, the system can preferentially release the lanes with high traffic flow according to the prediction results, or temporarily close the lanes with low traffic flow to optimize the overall traffic efficiency.

[0113] Emergency response: In case of an accident or emergency, the adjustment of the signal light duration can be dynamically adjusted based on the severity of the accident or its impact on the traffic flow.

[0114] The server controls the opening time of the ramp entrance barrier according to the traffic flow and vehicle speed conditions to prevent too many vehicles from entering the ramp and causing congestion. When necessary, the standby channel is opened to relieve the traffic pressure on the main ramp.

[0115] Through image recognition technology and sensor data, traffic accidents, vehicle breakdowns or other emergencies are automatically identified. When an emergency is detected, the system will quickly calculate the impact of the accident on the traffic and take actions according to the preset emergency plan. It can automatically close some ramp entrances or exits to prevent more vehicles from entering the accident area, and adjust the traffic lights to preferentially guide vehicles to detour. At the same time, detour suggestions and traffic condition information will be real-time released to drivers through highway information boards and highway radios to help them choose a more suitable driving route.

[0116] In addition, during the operation of this application, the implementation effects will be continuously recorded, and the traffic flow data and emergency handling results after implementation will be collected. Through the continuous accumulation of data and the self-learning of the AI model, the traffic control decision-making ability of the system is optimized, and the accuracy of prediction and the efficiency of emergency handling are improved. The hardware equipment of the system is regularly maintained and upgraded to ensure the long-term stable operation of the system.

[0117] In one embodiment, in order to more effectively manage the sections of the urban highway to be controlled and relieve traffic congestion, the traffic management department developed an intelligent ramp control system. Based on the preset congestion level classification criteria, the optimized estimated traffic conditions, and the real-time collected traffic data, the system dynamically adjusts and generates the target ramp control strategy. First, according to the preset congestion level classification criteria, the optimized estimated traffic conditions are evaluated. These criteria may include indicators such as vehicle flow, average speed, congestion length, and density. Through the evaluation, the system determines the congestion level corresponding to the section to be controlled, such as mild congestion, moderate congestion, or severe congestion, etc. This helps to provide a basis for the subsequent selection of ramp control strategies.

[0118] According to the determined congestion level, the system searches for corresponding ramp control strategies in the preset ramp control library. These strategies may include adjusting the signal timing of ramps, restricting vehicle entry, opening emergency lanes, etc. The system selects the ramp control strategy that best suits the current congestion level to relieve traffic pressure to the greatest extent and improve road traffic efficiency. To ensure the effectiveness of the ramp control strategy, the system also needs to consider the actual situation of the section to be controlled. Therefore, the system collects real-time traffic data of the section to be controlled, including vehicle flow, speed, vehicle type distribution, etc. By analyzing this data, the system can more accurately understand the real-time traffic conditions of the section and provide a basis for adjusting the ramp control strategy.

[0119] According to the actual situation, the system adjusts the selected ramp control strategy. The adjustments may include modifying the signal timing plan, adjusting the conditions for restricting vehicle entry, or optimizing the opening strategy of the emergency lane, etc. The adjusted ramp control strategy is generated as the target ramp control strategy and sent to relevant traffic control devices or personnel through wireless communication. These strategies will be used to guide the actual ramp control operations. Implement the target ramp control strategy and continuously monitor the traffic conditions of the section to be controlled. Through real-time feedback and adjustment, the system can ensure the effectiveness and adaptability of the ramp control strategy. If the traffic conditions change or new congestion occurs, the system will immediately trigger a new estimation and strategy adjustment process to maintain the safety and smoothness of the road.

[0120] The above is the method embodiment proposed in this application. Based on the same inventive concept, the embodiments of this application also provide an artificial intelligence-based toll station ramp control device, and its structure is as Figure 2 shown.

[0121] Figure 2 This is the internal structure schematic diagram of an artificial intelligence-based toll station ramp control device provided by the embodiment of this application. As Figure 2 shown, the device includes:

[0122] At least one processor;

[0123] And a memory communicatively connected to at least one processor;

[0124] Wherein, the memory stores instructions executable by at least one processor, and the instructions are executed by at least one processor so that at least one processor can:

[0125] Based on the roadside devices preset for the section to be controlled, collect the video stream and sensor data of the section to be controlled, and preprocess the video stream to determine the key frame images;

[0126] Analyze the key-frame image in combination with the sensor data, determine the traffic data and traffic impact factor values corresponding to the section to be controlled, and input the traffic data and traffic impact factor values into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to the preset time period;

[0127] Generate a recommended planned path for the corresponding driving vehicle based on the real-time obtained driving requirements of the vehicle and the estimated traffic conditions, and optimize the estimated traffic conditions based on the confirmation status of the driving vehicle for the corresponding recommended planned path;

[0128] Determine the congestion level corresponding to the optimized estimated traffic conditions, and determine the ramp control strategy corresponding to the congestion level based on the preset ramp control library, so as to perform ramp control on the toll station according to the ramp control strategy.

[0129] The embodiment of the present application also provides a non-volatile computer storage medium, storing computer-executable instructions, which can be executed to:

[0130] Collect the video stream and sensor data of the section to be controlled based on the roadside equipment preset for the section to be controlled, and preprocess the video stream to determine the key-frame image;

[0131] Analyze the key-frame image in combination with the sensor data, determine the traffic data and traffic impact factor values corresponding to the section to be controlled, and input the traffic data and traffic impact factor values into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to the preset time period;

[0132] Generate a recommended planned path for the corresponding driving vehicle based on the real-time obtained driving requirements of the vehicle and the estimated traffic conditions, and optimize the estimated traffic conditions based on the confirmation status of the driving vehicle for the corresponding recommended planned path;

[0133] Determine the congestion level corresponding to the optimized estimated traffic conditions, and determine the ramp control strategy corresponding to the congestion level based on the preset ramp control library, so as to perform ramp control on the toll station according to the ramp control strategy.

[0134] Each embodiment in the present application is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiments.

[0135] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0136] The devices and media provided by the embodiments of the present application correspond one-to-one with the methods. Therefore, the devices and media also have beneficial technical effects similar to those of their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be elaborated here.

[0137] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0138] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0139] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 steps of the functions specified in one block or multiple blocks.

[0141] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0142] The memory may include non-permanent memory in the computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

[0143] Computer-readable media includes permanent and non-permanent, removable and non-removable media and can be implemented by any method or technology for information storage. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0144] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the existence of additional identical elements in the process, method, commodity or device comprising the element.

[0145] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A toll station ramp control method based on artificial intelligence, characterized in that: The method comprises: Based on the preset roadside equipment of the road section to be controlled, the video stream and sensor data of the road section to be controlled are collected, and the video stream is pre-processed to determine the key frame image; Analyze the key frame image in combination with the sensor data to determine the traffic data and traffic influencing factor values ​​corresponding to the road section to be controlled, and input the traffic data and the traffic influencing factor values ​​into a trained traffic flow prediction model to output an estimated traffic condition corresponding to a preset time period; Based on the real-time acquired vehicle driving demand and the estimated traffic conditions, a recommended planned path for the corresponding driving vehicle is generated, and based on the confirmation status of the driving vehicle on the corresponding recommended planned path, the estimated traffic conditions are optimized; Determine the congestion level corresponding to the optimized estimated traffic condition, and determine the ramp control strategy corresponding to the congestion level based on a preset ramp control library, so as to perform ramp control on the toll station according to the ramp control strategy; Based on the preset roadside equipment of the road section to be controlled, the video stream and sensor data of the road section to be controlled are collected, and the video stream is preprocessed to determine the key frame image, specifically including: Through the preset cameras and sensors of the road section to be controlled, the video stream and sensor data corresponding to the road section to be controlled are collected in real time, and the video stream and sensor data uploaded by the cameras and the sensors are received; Analyze the video stream to identify the latter image of two adjacent images in the video stream whose preceding and following information are inconsistent, so as to determine the key frame image corresponding to the video stream; The traffic data and the traffic influencing factor values ​​are input into the trained traffic flow prediction model to output the estimated traffic conditions corresponding to the preset time period, specifically including: Determine whether the data format corresponding to the traffic data and the traffic influencing factor value is consistent with the input format corresponding to the traffic flow prediction model, and if not, convert the traffic data and the traffic influencing factor value into the input format; Input the traffic data and the traffic influencing factor values ​​in the input format into the trained traffic flow prediction model, and output the estimated traffic conditions corresponding to the road section to be controlled within a preset time period; The estimated traffic condition is verified, and if the verification fails, the estimated traffic condition is corrected based on a preset expert database to obtain a target estimated traffic condition.

2. The method for toll station ramp control based on artificial intelligence according to claim 1 is characterized in that: Analyzing the key frame image in combination with the sensor data to determine the traffic data and traffic influencing factor values ​​corresponding to the road section to be controlled, specifically including: Obtaining the historical ramp control records corresponding to the road section to be controlled, and determining the traffic influencing factors corresponding to the road section to be controlled based on the historical ramp control records; wherein the traffic influencing factors include time periods, holidays, and weather, and the time periods include peak periods, trough periods, and regular periods; Segmenting the key frame image to determine the moving vehicle in the key frame image, and identifying the vehicle model, vehicle color and license plate number corresponding to the moving vehicle; Determine the speed of the traveling vehicle according to the sensor data corresponding to the traveling vehicle on the adjacent road section of the road section to be controlled, so as to obtain the traffic data corresponding to the road section to be controlled; The data collection time corresponding to the traffic data is obtained, so as to determine the traffic influencing factor value corresponding to the road section to be controlled according to the data collection time and in combination with the sensor data.

3. The method for toll station ramp control based on artificial intelligence according to claim 1 is characterized in that: Before inputting the traffic data and the traffic influencing factor values ​​into a trained traffic flow prediction model to output an estimated traffic condition corresponding to a preset time period, the method further includes: Based on the deep learning algorithm, a traffic flow prediction model corresponding to the road section to be controlled is constructed; Obtain the historical traffic data corresponding to the road section to be controlled and the historical traffic influencing factor value corresponding to the historical traffic data, and mark the traffic conditions of the next time period corresponding to the historical traffic data and the historical traffic influencing factor value; The historical traffic data and the historical traffic influencing factor values ​​are input into the traffic flow prediction model, and the corresponding estimated traffic conditions are output until the estimated traffic conditions match the marked traffic conditions for the next time period, thereby completing the training of the traffic flow prediction model.

4. The method for controlling ramps at toll booths based on artificial intelligence according to claim 1, characterized in that: Based on the real-time acquired vehicle driving demand and the estimated traffic conditions, a recommended planning path for the corresponding driving vehicle is generated, specifically including: Determine whether there is a potential risk in the road section to be controlled according to the estimated traffic conditions, and if so, obtain driving demand information corresponding to vehicles traveling in the road section to be controlled; wherein the driving demand information includes the destination, expected arrival time, and preferred route; Based on the driving demand information and in combination with the estimated traffic conditions, generating at least one corresponding recommended planning route for the driving vehicle; wherein the recommended planning route includes an estimated duration, an estimated cost, and congestion conditions; The recommended planning path is sent to the corresponding traveling vehicle through wireless communication for the vehicle occupants to select and confirm.

5. The method for toll station ramp control based on artificial intelligence according to claim 1 is characterized in that: Based on the confirmation status of the driving vehicle on the corresponding recommended planned path, optimizing the estimated traffic condition specifically includes: Receive confirmation of the corresponding recommended planned path from all vehicles in the road section to be controlled, and dynamically update the traffic influencing factor value corresponding to the road section to be controlled based on the feedback data; According to the feedback data and in combination with the updated traffic influencing factor values, the model parameters of the traffic flow prediction model are iteratively optimized to generate a target estimated traffic condition through the optimized traffic flow prediction model.

6. The method for toll station ramp control based on artificial intelligence according to claim 1 is characterized in that: Determine the congestion level corresponding to the optimized estimated traffic conditions, and determine the ramp control strategy corresponding to the congestion level based on the preset ramp control library, specifically including: Based on the preset congestion level classification standard and according to the optimized estimated traffic conditions, determine the congestion level corresponding to the road section to be controlled; According to the congestion level, a corresponding ramp control strategy is determined in a preset ramp control library, and the actual situation corresponding to the road section to be controlled is determined according to the traffic data collected in real time from the road section to be controlled; According to the actual situation, the ramp control strategy is adjusted, and a target ramp control strategy corresponding to the road section to be controlled is generated.

7. A toll station ramp control device based on artificial intelligence, characterized in that: The device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; In which, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an artificial intelligence-based toll station ramp control method as described in any one of claims 1-6.

8. A non-volatile computer storage medium storing computer executable instructions, characterized in that: When the computer executable instructions are executed, an artificial intelligence-based toll station ramp control method as described in any one of claims 1 to 6 is implemented.

Images