Accurate Monitoring and Early Warning System and Method for Vehicle and Passenger Flow in Highway Closed Service Areas

Abstract

The present application relates to the field of highway intelligent service area, mainly relates to the highway closed service area vehicle flow and passenger flow accurate monitoring and early warning service system and method, the system, including service area import and export ramp sensing module, fusion identification module, monitoring and judging module, early warning information release module and client service module, through the deployment of video camera and radar equipment in the closed service area import and export, using the deep learning algorithm based on realizing the accurate collection, identification, research, early warning and information release and client service of the vehicle number, license plate, vehicle type, license plate features and passengers in the vehicle of entering and exiting the service area, realizing the hierarchical classification early warning service of vehicle flow and passenger flow in the service area, strengthening the comprehensive management and control of the service area, providing the current information of the parking space, passenger flow congestion degree of the front service area for the traveler to make scientific decision on whether to enter, ensuring the safety and order of the vehicle and personnel in the service area and guaranteeing the service quality.

Description

Technical Field

[0001] This invention relates to the field of smart service areas on highways, and more particularly to a system and method for precise monitoring and early warning of vehicle and passenger flow in closed service areas on highways. Background Technology

[0002] With the rapid expansion of the highway network, service areas, as important locations providing services such as vehicle parking, refueling, dining, and rest along highways, are being used more and more frequently. Especially during holidays or peak traffic periods, the traffic and passenger flow in service areas increases dramatically, leading to problems such as parking shortages and overcrowding. This not only reduces the service level of service areas but also increases traffic safety hazards.

[0003] Traditional highway service area management relies primarily on manual patrols and simple camera monitoring, which makes it difficult to achieve real-time and efficient monitoring and management of traffic and passenger flow. Furthermore, service area managers often struggle to respond and adjust promptly to drastic fluctuations in traffic and passenger volume, easily leading to congestion, accidents, and other problems, thus impacting the operational efficiency and quality of travel services.

[0004] With the advancement of information technology, especially the widespread application of technologies such as the Internet of Things and big data analytics, intelligent management of highway service areas has become possible. Through precise monitoring and data analysis of traffic and passenger flow within service areas, real-time monitoring and early warning of parking space usage can be achieved, improving the operational management capabilities of service areas. However, most monitoring systems currently on the market involve numerous and costly installations, making it difficult to meet the widespread application needs of highway service areas nationwide. Therefore, developing a system with fewer installations, high cost-effectiveness, and real-time monitoring and early warning capabilities is of significant practical importance. Summary of the Invention

[0005] The purpose of this invention is to provide a system and method for accurate monitoring and early warning of vehicle and passenger flow in closed service areas of highways, thereby solving the aforementioned problems existing in the prior art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A precise monitoring and early warning system for vehicle and passenger flow in a closed highway service area includes a service area entrance and exit ramp sensing module, a fusion identification module, a vehicle and passenger flow monitoring and analysis module, an early warning information release module, and a client service module.

[0008] The service area entrance and exit ramp sensing module consists of an entrance ramp sensing unit, an exit ramp sensing unit, and a network transmission unit. The entrance ramp sensing unit is installed on the pole at the entrance ramp of the enclosed service area and includes an imported high-definition intelligent camera, an imported millimeter-wave radar, an imported thermal imaging camera, an imported infrared gas strobe light, and an imported LED fill light. The exit ramp sensing unit is installed on the pole at the exit ramp of the service area and includes an exit high-definition intelligent camera, an exit millimeter-wave radar, an exit thermal imaging camera, an exit infrared gas strobe light, and an exit LED fill light. The network transmission unit includes a network switch, a server, a storage terminal server, and a transmission network. The network transmission unit is used for system... The system transmits relevant data; the placement of the entrance ramp poles ensures that the imported high-definition intelligent cameras on them can capture panoramic photos of the front, license plate, and vehicle type of vehicles entering the area from the front and sides; after the photos are taken, they are uploaded to the fusion recognition module for identification; the imported thermal imaging camera counts the number of people inside the vehicles entering the area; the placement of the exit ramp poles ensures that the exit high-definition intelligent cameras on them can capture panoramic photos of the front, license plate, and vehicle type of vehicles leaving the area from the front and sides; after the photos are taken, they are uploaded to the fusion recognition module for identification; the exit thermal imaging camera counts the number of people inside the vehicles leaving the area; the license plate information is a unique identifier for each vehicle, used to identify the corresponding vehicle;

[0009] The fusion recognition module includes a vehicle front feature recognition unit, a vehicle side feature recognition unit, a key vehicle trajectory recognition unit, and an in-vehicle personnel recognition unit;

[0010] The vehicle and passenger flow monitoring and analysis module includes a vehicle flow early warning monitoring and analysis unit, a passenger flow early warning monitoring and analysis unit, and a key vehicle electronic fence unit;

[0011] The warning information publishing module is used to publish warning information; the warning information is published in three ways: ① by setting up an electronic display screen and broadcasting equipment 200-500 meters upstream of the service area entrance; ② by publishing the warning information through the variable message sign on the ETC gantry upstream of the service area; ③ by publishing the warning information to the client service module.

[0012] The client service module is installed on the relevant user's mobile phone and / or vehicle system, and is used to determine whether to actively pop up the warning information based on the current location of the device, and to passively display the warning information and basic information of the service area according to the user's instructions.

[0013] Preferably, the specific process of determining whether to actively pop up the warning information based on the current location of the device is as follows: after the client service module receives the warning information published by the warning information publishing module, it first determines whether the current navigation route of the device passes through the service area, and then determines whether the device is within the range of the warning information. If both determination results are yes, the warning information is automatically popped up through the device display.

[0014] Preferably, determining whether the local machine is within the range of the warning information specifically means determining whether the local machine's current location is within a certain distance upstream of the service area.

[0015] A method for precise monitoring and early warning of vehicle and passenger flow in closed service areas of highways, implemented based on the aforementioned system, specifically includes the following steps:

[0016] S1, collect real-time traffic flow information entering and exiting the closed service area, and calculate the total number of vehicles currently in the service area through the real-time traffic flow status monitoring algorithm of the closed service area;

[0017] Step S1 specifically involves: calculating the real-time traffic flow status of the closed service area using real-time data from the inlet and outlet ramp sensing units; setting the difference between the total number of vehicles entering and exiting the service area as the total number of vehicles currently in the service area, and assigning the device numbers of the inlet and outlet ramp sensing units as follows: the number of the i-th inlet vehicle, denoted as Cin. i Let i = 1 - N, and let Cout be the number of vehicles exiting at the j-th exit. j If j = 1 - M, then the total number of vehicles currently in the service area can be expressed by the formula:

[0018] S2, based on the data obtained by the entrance and exit ramp sensing unit in step S1, identify and obtain the existing vehicle types in the service area using a CNN-based vehicle side feature recognition algorithm;

[0019] The CNN-based vehicle side feature recognition algorithm in step S2 specifically includes: using a Convolutional Neural Network (CNN) as the core model in deep learning technology, and designing a FlowNet structure to reduce complexity; the first is FlowNetHalf, which reduces the number of convolutional kernels in each layer of FlowNet to half, reducing complexity to 1 / 4; the second is FlowNetInception, which adopts the Inception network structure, reducing complexity to 1 / 8. Through fast image processing, multi-layer convolution and pooling operations, it automatically learns the features in the image and automatically extracts and recognizes vehicle side features, thereby improving the accuracy and efficiency of recognition.

[0020] S3 uses a real-time passenger flow monitoring algorithm for closed parking lots to calculate the maximum number of existing passengers;

[0021] The real-time passenger flow monitoring algorithm in step S3 specifically includes: calculating the number of passengers in vehicles within the service area based on data measured by the entrance ramp sensing unit and the exit ramp sensing unit, which represents the real-time passenger flow in the service area; binding the vehicle information as a unique identifier, and simultaneously reducing the passenger flow corresponding to the vehicle when it is detected to have left, thus obtaining the maximum number of passengers in the service area at time t; the real-time passenger flow monitoring algorithm is expressed by the formula:

[0022]

[0023] Passengers t , representing the current number of passengers in the service area at time t;

[0024] This indicates the j-th vehicle in the sensing unit of the i-th ramp at time t;

[0025] This represents the number of passengers in the j-th vehicle within the sensing unit of ramp i at time t.

[0026] This indicates the j-th vehicle in the exit ramp sensing unit at time t;

[0027] S4, count the total number of parking spaces of each type in the service area, and then obtain the total number of parking spaces;

[0028] Step S4 specifically includes: the total number of parking spaces in the service area is denoted as Park. max The total number of parking spaces of various types in the service area was counted. The first type is parking spaces for small passenger vehicles, denoted as P. Car The second category includes parking spaces for large passenger vehicles and ordinary freight vehicles, denoted as P. Bus The third category is parking spaces for large trucks and hazardous materials transport vehicles, denoted as P. Van The fourth category is parking spaces for other types of vehicles, denoted as P. other The total number of parking spaces in the service area is Park. max The calculation formula is: Park max =Σ(P Van +P Car +P Bus +P other );

[0029] S5, calculate the upper limit of passenger flow capacity in each area of ​​the service area, and then obtain the maximum upper limit of passenger flow capacity of the service area;

[0030] Step S5 specifically includes: the maximum passenger flow capacity of the service area is denoted as Passengers. max This refers to the sum of the maximum passenger capacity of multiple areas within the service area, including the comprehensive operating area, public restrooms, and other passenger flow areas. The calculation formula is as follows:

[0031] Passengers max =∑(Passengers) Business +Passengers toilet +Passengers other )

[0032] In the formula: Passengers Business This indicates the maximum passenger capacity of the integrated operating area. toilet This indicates the maximum capacity of public restrooms. other This indicates the maximum passenger capacity for other passenger flow areas;

[0033] S6. Based on the above data, the current congestion status of the service area is calculated using the service area early warning analysis algorithm.

[0034] The service area early warning analysis algorithm in step S6 includes: a service area parking space real-time status algorithm and a service area passenger flow real-time status algorithm.

[0035] In the real-time status algorithm for service area parking spaces, the real-time status is divided into three states: 0 (empty); 1 (normal); 2 (crowded). The algorithm is as follows:

[0036] R Vehicle idle =Car now / Park max R Vehicle idle <0.6 indicates idle status, and the real-time status is displayed as 0 (idle); 0.6≤R Vehicle idle <1.0 indicates normal operation; the real-time status is displayed as 1 (normal). Vehicle idle ≥1.0 indicates congestion, and the real-time status is displayed as 2 congestion.

[0037] In the real-time passenger flow status algorithm of the service area, the real-time passenger flow status is divided into three states: 0 (idle); 1 (normal); 2 (crowded). The algorithm is as follows:

[0038] R passenger idle =Passengers now / Passengers max R Passengers idle <0.6 indicates idle status, and the real-time status is displayed as 0 (idle); 0.6≤R Passengers idle<1.0 indicates normal operation; the real-time status is displayed as 1 (normal). Passengers idle ≥1.0 indicates congestion, and the real-time status is displayed as 2 congestion.

[0039] S7. Construct electronic fences for key vehicles and monitor the path and dwell time of the key vehicles in the service area. If relevant regulations are violated, trigger an early warning and take preset measures.

[0040] Step S7 specifically includes: constructing an electronic fence for hazardous chemical transport vehicles in the parking area of ​​the service area; issuing warnings for the parking location and stopping time of hazardous chemical transport vehicles entering the service area; binding the hazardous chemical transport vehicles with their license plate information and tracking the complete path trajectory of the hazardous chemical transport vehicles in the service area; and using GPS and Beidou positioning data of key operating vehicles for data fitting to realize electronic fence management in the service area. If a hazardous chemical transport vehicle enters the service area but does not enter the electronic fence corresponding to the hazardous chemical transport vehicle parking area, a warning is issued. If the vehicle stays in the electronic fence for more than half an hour, the service area staff are notified to conduct on-site inquiries. If it stays for more than one hour, the vehicle owner and the company to which the vehicle belongs are notified for an overdue warning. This realizes full-process monitoring of hazardous chemical transport vehicles from entry to exit, and actively pushes monitoring information to the driver, owner, and company of the vehicle. Overdue warnings are issued and telephone warnings are given. The full-process monitoring ends when the vehicle leaves the service area.

[0041] S8 displays service area warning information signs on electronic displays and variable message signs installed upstream of the service area entrance, indicating the warning information and basic information of the service area; and notifies relevant users through information push or broadcast.

[0042] Preferably, step S2 specifically includes:

[0043] ① Data preprocessing: The vehicle side images obtained and transmitted by the on-ramp perception units are preprocessed, including image scaling, normalization, and enhancement operations, to improve the robustness of the model.

[0044] ② Feature extraction: Use a pre-trained CNN model ResNet or VGG or a custom CNN model to extract features from the vehicle side image; these features include the vehicle outline, windows, and wheels;

[0045] ③ Feature Classification: The extracted features are input into a fully connected layer or classifier to classify vehicle types; a simple Softmax classifier is used to classify the vehicle types entering the service area. Among them: 1. Small passenger vehicles, 2. Large passenger vehicles, 3. Ordinary trucks, 4. Large trucks, 5. Hazardous materials transport vehicles, 6. Livestock transport vehicles, 7. Others;

[0046] ④ Vehicle occupant identification and counting: Thermal imaging cameras in the entrance and exit ramp sensing units take side photos of passing vehicles, and deep learning algorithms are used to identify and count the number of occupants. And it is linked to the license plate;

[0047] ⑤ Taking K vehicles entering the parking lot at time T as an example, the number of passengers on K vehicles is:

[0048] Preferably, the content displayed on the electronic display screen and variable message sign should comply with the relevant provisions of the current "Design Specifications for Highway Speed ​​Limit Signs" and "LED Variable Message Signs for Expressways".

[0049] The beneficial effects of this invention are:

[0050] The implementation of the technical solution of this invention can monitor and warn of the vacancy status of parking spaces in closed service areas of highways nationwide in real time, and can warn of congestion in service areas. This invention has the characteristics of requiring less equipment, high cost performance, and strong real-time performance, and can quickly improve the operation and management capabilities of service areas and the level of travel services. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the arrangement of the inlet ramp sensing unit in this invention;

[0052] Figure 2 This is a schematic diagram of the system architecture of the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0054] This invention relates to the field of smart service areas on highways, and is mainly applied to the implementation method of a precise monitoring and early warning service system for vehicle and passenger flow within closed smart service areas on highways. The system includes a service area entrance / exit ramp sensing module, a fusion identification module, a monitoring and analysis module, an early warning information dissemination module, and a client service module. By deploying video cameras and radar equipment at the entrances and exits of closed service areas, and employing deep learning algorithms, the system achieves precise collection, identification, analysis, early warning, and information dissemination of data on the number of vehicles entering and exiting the service area, license plates, vehicle types, license plate characteristics, and passengers inside the vehicles, as well as providing client services. This enables tiered and categorized early warning services for vehicle and passenger flow within the service area, strengthening the comprehensive management and control of the service area. It also provides travelers with information on the current status of parking spaces and passenger congestion in the service area ahead, allowing them to make informed decisions about whether to enter, ensuring the safety and orderliness of vehicles and personnel within the service area and guaranteeing service quality.

[0055] According to the "Basic Functional Requirements for Digital Construction of Highway Service Areas (Trial)," the digital construction of highway service areas nationwide needs to gradually realize the real-time monitoring, automatic collection, and external release of service area data information, so as to improve the service area operation and management capabilities and travel service levels.

[0056] The service area traffic and passenger flow precision monitoring and early warning service system based on the present invention is an important component of the construction of digital service areas. It can support the construction of provincial service area management information platforms and highway service area terminal systems (hereinafter referred to as terminal systems), and help improve the operational efficiency of service areas and enhance their safety management capabilities.

[0057] The present invention relates to a highway closed service area vehicle and passenger flow precision monitoring and early warning service system, comprising a service area entrance and exit ramp sensing module, a fusion recognition module, a vehicle and passenger flow monitoring and analysis module, an early warning information release module, and a client service module. The service area entrance and exit ramp sensing module consists of an entrance ramp sensing unit, an exit ramp sensing unit, and a network transmission unit. The entrance ramp sensing unit is installed on a pole at the entrance ramp of the closed service area and includes an imported high-definition intelligent camera, an imported millimeter-wave radar, an imported thermal imaging camera, an imported infrared gas strobe light, and an imported LED supplementary light. The exit ramp sensing unit is installed on a pole at the exit ramp of the service area and includes an exit high-definition intelligent camera, an exit millimeter-wave radar, an exit thermal imaging camera, an exit infrared gas strobe light, and an exit LED supplementary light. The network... The transmission unit includes a network switch, a server, a storage terminal server, and a transmission network. The network transmission unit is used for transmitting system-related data. The placement of the entrance ramp poles ensures that the imported high-definition intelligent cameras on them can capture panoramic photos of the front, license plate, and vehicle type of vehicles entering the area from both the front and side. After the photos are taken, they are uploaded to the fusion recognition module for identification. The imported thermal imaging camera counts the number of people inside the vehicles entering the area. The placement of the exit ramp poles ensures that the exit high-definition intelligent cameras on them can capture panoramic photos of the front, license plate, and vehicle type of vehicles leaving the area from both the front and side. After the photos are taken, they are uploaded to the fusion recognition module for identification. The exit thermal imaging camera counts the number of people inside the vehicles leaving the area. License plate information is a unique identifier for each vehicle, used to identify the corresponding vehicle.

[0058] The fusion recognition module includes a vehicle front feature recognition unit, a vehicle side feature recognition unit, a key vehicle trajectory recognition unit, and an in-vehicle personnel recognition unit;

[0059] The vehicle and passenger flow monitoring and analysis module includes a vehicle flow early warning monitoring and analysis unit, a passenger flow early warning monitoring and analysis unit, and a key vehicle electronic fence unit;

[0060] The warning information publishing module is used to publish warning information. Warning information is published in three ways: ① by setting up electronic displays and broadcasting equipment 200-500 meters upstream of the service area entrance; ② by publishing the warning information through variable message signs on the ETC gantry upstream of the service area; ③ by publishing the warning information to the client service module. The client service module includes mobile navigation software, service area apps, WeChat mini-programs, and other mobile client service modules. Warning information and basic information are published for travelers, especially drivers, using mobile navigation software, service area apps, WeChat mini-programs, and other mobile client service modules.

[0061] The client service module is installed on the relevant user's mobile phone and / or vehicle system, and is used to determine whether to actively pop up the warning information based on the current location of the device, and to passively display the warning information and basic information of the service area according to the user's instructions.

[0062] The specific process of determining whether to actively display the warning information based on the current location of the device is as follows: After receiving the warning information published by the warning information publishing module, the client service module first determines whether the current navigation route of the device passes through the service area, and then determines whether the device is within the range of the warning information. If both determinations are yes, the warning information is automatically displayed on the device's display device. Specifically, determining whether the device is within the range of the warning information means determining whether the device's current location is within a certain distance upstream of the service area.

[0063] The method for precise monitoring and early warning of vehicle and passenger flow in closed service areas of highways of the present invention is based on the above system and specifically includes the following steps:

[0064] S1, collect real-time traffic flow information entering and exiting the closed service area, and calculate the total number of vehicles currently in the service area through the real-time traffic flow status monitoring algorithm of the closed service area;

[0065] Step S1 specifically involves: calculating the real-time traffic flow status of the closed service area using real-time data from the inlet and outlet ramp sensing units; setting the difference between the total number of vehicles entering and exiting the service area as the total number of vehicles currently in the service area, and assigning the device numbers of the inlet and outlet ramp sensing units as follows: the number of the i-th inlet vehicle, denoted as Cin. i Let i = 1 - N, and let Cout be the number of vehicles exiting at the j-th exit. j If j = 1 - M, then the total number of vehicles currently in the service area can be expressed by the formula:

[0066] S2, based on the data obtained by the entrance and exit ramp sensing unit in step S1, identify and obtain the existing vehicle types in the service area using a CNN-based vehicle side feature recognition algorithm;

[0067] By deploying cameras, infrared gas flash lights, LED fill lights, storage terminal servers, and other equipment on the front and sides of the service area entrance and exit ramps, deep learning algorithms are used to identify vehicle type, license plate, number of passengers (the maximum number of passengers can be verified by vehicle type), vehicle color, and other information.

[0068] The CNN-based vehicle side feature recognition algorithm in step S2 specifically includes: using a Convolutional Neural Network (CNN) as the core model in deep learning technology, and designing a FlowNet structure to reduce complexity; the first is FlowNetHalf, which reduces the number of convolutional kernels in each layer of FlowNet to half, reducing complexity to 1 / 4; the second is FlowNetInception, which adopts the Inception network structure, reducing complexity to 1 / 8. Through fast image processing, multi-layer convolution and pooling operations, it automatically learns the features in the image and automatically extracts and recognizes vehicle side features, thereby improving the accuracy and efficiency of recognition.

[0069] Step S2 specifically includes:

[0070] ① Data preprocessing: The vehicle side images obtained and transmitted by the on-ramp perception units are preprocessed, including image scaling, normalization, and enhancement operations, to improve the robustness of the model.

[0071] ② Feature extraction: Use a pre-trained CNN model ResNet or VGG or a custom CNN model to extract features from the vehicle side image; these features include the vehicle outline, windows, and wheels;

[0072] ③ Feature Classification: The extracted features are input into a fully connected layer or classifier to classify vehicle types; a simple Softmax classifier is used to classify the vehicle types entering the service area. Among them: 1. Small passenger vehicles, 2. Large passenger vehicles, 3. Ordinary trucks, 4. Large trucks, 5. Hazardous materials transport vehicles, 6. Livestock transport vehicles, 7. Others;

[0073] ④ Vehicle occupant identification and counting: Thermal imaging cameras in the entrance and exit ramp sensing units take side photos of passing vehicles, and deep learning algorithms are used to identify and count the number of occupants. And it is linked to the license plate;

[0074] ⑤ Taking K vehicles entering the parking lot at time T as an example, the number of passengers on K vehicles is:

[0075] S3 uses a real-time passenger flow monitoring algorithm for closed parking lots to calculate the maximum number of existing passengers;

[0076] The real-time passenger flow monitoring algorithm in step S3 specifically includes: calculating the number of passengers in vehicles currently operating within the service area based on data measured by the entrance ramp sensing unit and the exit ramp sensing unit, which represents the real-time passenger flow within the service area; binding passengers using license plate information as a unique identifier, and simultaneously deducting the passenger flow corresponding to a vehicle when it is detected to have left, thus obtaining the maximum current passenger flow in the service area; the real-time passenger flow monitoring algorithm is expressed by the formula:

[0077]

[0078] Passengers t , representing the current number of passengers in the service area at time t;

[0079] This indicates the j-th vehicle in the sensing unit of the i-th ramp at time t;

[0080] This represents the number of passengers in the j-th vehicle within the sensing unit of ramp i at time t.

[0081] This indicates the j-th vehicle in the exit ramp sensing unit at time t;

[0082] S4, count the total number of parking spaces of each type in the service area, and then obtain the total number of parking spaces;

[0083] Step S4 specifically includes: the total number of parking spaces in the service area is denoted as Park. max The total number of parking spaces of various types in the service area was counted. The first type is parking spaces for small passenger vehicles, denoted as P. Car The second category includes parking spaces for large passenger vehicles and ordinary freight vehicles, denoted as P. Bus The third category is parking spaces for large trucks and hazardous materials transport vehicles, denoted as P. Van The fourth category is parking spaces for other types of vehicles, denoted as P. other The total number of parking spaces in the service area is Park. max The calculation formula is: Park max =∑(P Van +P Car +P Bus +P other );

[0084] S5, calculate the upper limit of passenger flow capacity in each area of ​​the service area, and then obtain the maximum upper limit of passenger flow capacity of the service area;

[0085] Step S5 specifically includes: the maximum passenger flow capacity of the service area is denoted as Passengers. max This refers to the sum of the maximum passenger capacity of multiple areas within the service area, including the comprehensive operating area, public restrooms, and other passenger flow areas. The calculation formula is as follows:

[0086] Passengers max =Σ(Passengers Business +Passengers toilet +Passengers other )

[0087] In the formula: Passengers Business This indicates the maximum passenger capacity of the integrated operating area. toilet This indicates the maximum capacity of public restrooms. other This indicates the maximum passenger capacity for other passenger flow areas;

[0088] S6. Based on the above data, the current congestion status of the service area is calculated using the service area early warning analysis algorithm.

[0089] The service area early warning analysis algorithm in step S6 includes: a service area parking space real-time status algorithm and a service area passenger flow real-time status algorithm.

[0090] In the real-time status algorithm for service area parking spaces, the real-time status is divided into three states: 0 (empty); 1 (normal); 2 (crowded). The algorithm is as follows:

[0091] R Vehicle idle =Car now / Park max R Vehicle idle <0.6 indicates idle status, and the real-time status is displayed as 0 (idle); 0.6≤R Vehicle idle <1.0 indicates normal operation; the real-time status is displayed as 1 (normal). Vehicle idle ≥1.0 indicates congestion, and the real-time status is displayed as 2 congestion.

[0092] In the real-time passenger flow status algorithm of the service area, the real-time passenger flow status is divided into three states: 0 (idle); 1 (normal); 2 (crowded). The algorithm is as follows:

[0093] R passenger idle =Passengers now / Passengers max R Passengers idle <0.6 indicates idle status, and the real-time status is displayed as 0 (idle); 0.6≤R Passengers i dle<1.0 indicates normal operation; the real-time status is displayed as 1 (normal). Passengers idle ≥1.0 indicates congestion, and the real-time status is displayed as 2 congestion.

[0094] S7. Construct electronic fences for key vehicles and monitor the path and dwell time of the key vehicles in the service area. If relevant regulations are violated, trigger an early warning and take preset measures.

[0095] Step S7 specifically includes: constructing an electronic fence for hazardous chemical transport vehicles in the parking area of ​​the service area; issuing warnings for the parking location and stopping time of hazardous chemical transport vehicles entering the service area; binding the hazardous chemical transport vehicles with their license plate information and tracking the complete path trajectory of the hazardous chemical transport vehicles in the service area; and using GPS and Beidou positioning data of key operating vehicles for data fitting to realize electronic fence management in the service area. If a hazardous chemical transport vehicle enters the service area but does not enter the electronic fence corresponding to the hazardous chemical transport vehicle parking area, a warning is issued. If the vehicle stays in the electronic fence for more than half an hour, the service area staff are notified to conduct on-site inquiries. If it stays for more than one hour, the vehicle owner and the company to which the vehicle belongs are notified for an overdue warning. This realizes full-process monitoring of hazardous chemical transport vehicles from entry to exit, and actively pushes monitoring information to the driver, owner, and company of the vehicle. Overdue warnings are issued and telephone warnings are given. The full-process monitoring ends when the vehicle leaves the service area.

[0096] S8. Service area warning information signs are displayed on electronic displays and variable message signs located upstream of the service area entrance, indicating the warning information and basic details of the service area; this information is also pushed to relevant users via push notifications or broadcasts. The content displayed on the electronic displays and variable message signs should comply with the relevant provisions of the current "Design Specifications for Highway Speed ​​Limit Signs" and "Electric LED Variable Message Signs for Expressways". The aforementioned basic information can also be provided to the public through roadside broadcasting equipment or smart terminals such as mobile navigation software, reasonably guiding the public in choosing service areas and improving user satisfaction.

[0097] By adopting the above-disclosed technical solution of this invention, the following beneficial effects are obtained:

[0098] The implementation of the technical solution of this invention can monitor and warn of the vacancy status of parking spaces in closed service areas of highways nationwide in real time, and can warn of congestion in service areas. This invention has the characteristics of requiring less equipment, high cost performance, and strong real-time performance, and can quickly improve the operation and management capabilities of service areas and the level of travel services.

[0099] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A highway closed service area vehicle flow and passenger flow precision monitoring and early warning service system, characterized in that, The service area entrance and exit ramp sensing module, the fusion identification module, the vehicle flow and passenger flow monitoring and judgment module, the early warning information publishing module and the client service module are included. The service area entrance and exit ramp sensing module is composed of an entrance ramp sensing unit, an exit ramp sensing unit and a network transmission unit; the entrance ramp sensing unit is arranged on an entrance ramp stand of a closed service area, and includes an entrance high-definition intelligent camera, an entrance millimeter wave radar, an entrance thermal imaging camera, an entrance infrared gas flash lamp and an entrance LED light supplementing lamp. The exit ramp sensing unit is arranged on an exit ramp stand of the service area, and includes an exit high-definition intelligent camera, an exit millimeter wave radar, an exit thermal imaging camera, an exit infrared gas flash lamp and an exit LED light supplementing lamp; the network transmission unit includes a network switch, a server, a storage terminal server and a transmission network; the network transmission unit is used for transmission of system related data; the setting position of the entrance ramp stand can ensure that the entrance high-definition intelligent camera thereon can take photos of the vehicle head, license plate and vehicle model of the vehicle entering the service area from the front and side panoramic view; the photos are uploaded to the fusion identification module for identification after being taken; the number of people in the vehicle entering the service area is counted by the entrance thermal imaging camera; The setting position of the exit ramp stand can ensure that the exit high-definition intelligent camera thereon can take photos of the vehicle head, license plate and vehicle model of the vehicle exiting the service area from the front and side panoramic view; the photos are uploaded to the fusion identification module for identification after being taken; The number of people in the vehicle exiting the service area is counted by the exit thermal imaging camera; the license plate information is the unique identification of each vehicle, and is used for identifying the corresponding vehicle; The fusion recognition module comprises a vehicle front feature recognition unit, a vehicle side feature recognition unit, a key vehicle trajectory recognition unit and an in-vehicle personnel recognition unit; the key vehicle trajectory recognition unit binds the key vehicle with its license plate information and tracks the complete path trajectory of the key vehicle in the service area; the in-vehicle personnel recognition unit takes a side photo of the passing vehicle by a thermal imaging camera in the entrance and exit ramp sensing unit, identifies the number of vehicle passengers by using a deep learning algorithm, and binds the number of vehicle passengers with the license plate . The vehicle flow and passenger flow monitoring and judgment module includes a vehicle flow early warning monitoring and judgment unit, a passenger flow early warning monitoring and judgment unit and a key vehicle electronic fence unit; the key vehicle electronic fence unit constructs a vehicle electronic fence for key vehicles, and monitors the path trajectory and stay time of the key vehicles in the service area; if the key vehicles violate relevant regulations, an early warning is triggered and a preset measure is taken; the data fitting is performed by using the GPS and Beidou positioning data of the key operating vehicles, so that the electronic fence management in the service area is realized; The early warning information publishing module is used for publishing early warning information; the early warning information is published in three ways: ① an electronic display screen and a broadcasting device are arranged at a position 200-500 meters upstream of the service area entrance to publish the early warning information; ② the early warning information is published on a variable information board on an ETC gantry upstream of the service area; and ③ the early warning information is published to the client service module; The client service module is installed on a related user's mobile phone and / or vehicle machine, and is used for actively popping up the early warning information according to the current position of the local machine, and passively displaying the early warning information and the basic information of the service area according to the user's instruction.

2. The highway closed service area vehicle flow and passenger flow precision monitoring and early warning service system according to claim 1, characterized in that, The specific process of determining whether to actively pop up the early warning information according to the current position of the machine is that after the client service module receives the early warning information published by the early warning information publishing module, it first determines whether the current navigation route of the machine passes through the service area, and then determines whether the machine is in the early warning information coverage range, and if both determination results are yes, the early warning information is automatically popped up through the machine display device.

3. The highway closed service area vehicle flow and passenger flow precision monitoring and early warning service system according to claim 2, characterized in that, The specific process of determining whether the machine is in the early warning information coverage range is to determine whether the current position of the machine is within a certain distance on the upstream side of the service area.

4. A highway closed service area vehicle flow and passenger flow precision monitoring and early warning service method, characterized in that, The system implementation according to any one of claims 1-3 specifically comprises the following steps: S1, real-time collection of vehicle flow information entering and exiting the closed service area, and calculation of the total number of existing vehicles in the service area by a closed service area real-time vehicle flow state monitoring algorithm; The step S1 is specifically: calculating the real-time traffic state of the closed service area through the real-time data of the import ramp sensing unit and the export ramp sensing unit; setting the total number difference of the import and export vehicles measured by the service area import and export ramp sensing unit as the total number of the existing vehicles in the service area, and the device numbers of the service area import and export ramp sensing unit are respectively: the first import vehicle number, denoted as , i=1-N, the first export vehicle number, denoted as , j=1-M, and the total number of the existing vehicles in the service area is expressed by the formula: . S2, according to the data obtained by the entrance and exit ramp sensing unit in step S1, a CNN-based vehicle side feature recognition algorithm is used to identify and obtain the types of existing vehicles in the service area; The CNN-based vehicle side feature recognition algorithm of step S2 specifically includes: using a convolutional neural network CNN as the core model in deep learning technology, reducing complexity by designing a FlowNet structure; the first is FlowNetHalf, which reduces the number of convolution kernels in each layer of FlowNet to half, and the complexity is reduced to 1 / 4. The second is FlowNetInception, which uses an Inception network structure to reduce the complexity to 1 / 8. Through fast image processing, multi-layer convolution and pooling operations, and automatic learning of image features, the vehicle side feature recognition algorithm automatically extracts and identifies vehicle side features, thereby improving the accuracy and efficiency of recognition; S3, calculating the maximum number of existing passenger flows by using a closed parking lot real-time passenger flow state monitoring algorithm; The real-time passenger flow state monitoring algorithm of the step S3 specifically comprises: calculating the existing number of people in the vehicle in the service area through the data measured by the import ramp sensing unit and the export ramp sensing unit, so as to obtain the real-time passenger flow number in the service area; through binding according to the license plate information as a unique identifier, when it is monitored that the vehicle leaves, the passenger flow number corresponding to the vehicle is also synchronously reduced, so that the maximum number of existing passenger flow at the moment in the service area is obtained The real-time passenger flow state monitoring algorithm is represented by a formula. ； , indicates the current passenger flow in the service area at the moment , indicates moment import Ramp sensing unit Vehicle; , indicates time entry / exit loop sensing unit number of passengers in the vehicle , indicates moment exit Ramp sensing unit Vehicle; S4, counting the total number of each type of parking space in the service area to obtain the total amount of all parking spaces; The step S4 specifically comprises: total parking spaces of the service area are recorded as The total number of each type of parking space in the service area is counted, the first type is small passenger car parking space, recorded as The second type includes large passenger car and ordinary truck parking space, recorded as The third type is large truck and dangerous goods transport vehicle parking space, recorded as The fourth type is other types of vehicle parking space, recorded as The total number of parking spaces in the service area The calculation formula is: ​ S5, counting the upper limit of the passenger flow capacity of each area in the service area to obtain the highest upper limit of the passenger flow capacity of the service area; The step S5 specifically comprises: recording the passenger flow maximum capacity of the service area as is the sum of the passenger flow upper limit capacities of the comprehensive operation area, the public toilet, and other passenger flow areas in the service area, and the calculation formula is: In the formula: represents the upper limit of the passenger flow capacity of the comprehensive management area, represents the upper limit of the passenger flow capacity of the public toilet, represents the upper limit of the passenger flow capacity of other passenger flow areas; S6, calculating the current congestion of the service area by a service area early warning analysis algorithm based on the above data; The service area early warning analysis algorithm of step S6 includes a service area parking space real-time state algorithm and a service area passenger flow real-time state algorithm; In the service area parking space real-time state algorithm, the service area parking space real-time state is divided into three states: 0 idle; 1 normal; 2 crowded, and the algorithm is as follows: , , indicates idle, and displays the real-time status as 0 idle; , indicates normal, and displays the real-time status as 1 normal, , indicates congestion, and displays the real-time status as 2 congestion; In the service area passenger flow real-time state algorithm, the service area passenger flow real-time state is divided into three states: 0 idle; 1 normal; 2 crowded, and the algorithm is as follows: , , indicates idle, and displays the real-time status as 0 idle; , indicates normal, and displays the real-time status as 1 normal, , indicates congestion, and displays the real-time status as 2 congestion; S7, constructing a vehicle electronic fence for key vehicles and monitoring the path trajectory and stay time of the key vehicles in the service area, and if the relevant provisions are violated, triggering an early warning and taking preset measures; The step S7 specifically comprises: constructing a vehicle electronic fence in the dangerous goods transport vehicle parking area in the service area, warning the parking position and parking time of the dangerous goods transport vehicle entering the service area; binding the dangerous goods transport vehicle with its license plate information, and tracking the complete path trajectory of the dangerous goods transport vehicle in the service area; and using the key operating vehicle GPS Beidou positioning data for data fitting to realize the electronic fence management in the service area, if the dangerous goods transport vehicle enters the service area but does not enter the electronic fence corresponding to the dangerous goods transport vehicle parking area, a warning is given, or if the vehicle is in the electronic fence for more than half an hour, the service area staff is notified to conduct on-site inquiry, and if it is more than 1 hour, the vehicle owner and the vehicle enterprise are notified for overtime warning, so as to realize the whole process monitoring of the dangerous goods transport vehicle from entering to leaving the service area, and the monitoring information is actively pushed to the driver, owner and enterprise of the vehicle, the overtime warning and telephone warning are given, and the whole process monitoring is ended when the vehicle leaves the service area; S8, through the electronic display screen and variable information board set upstream of the service area entrance, the service area warning information board is displayed, and the warning information and basic information of the service area are marked; and the information is pushed or broadcasted to the related users.

Images