Urban road bottleneck node congestion intelligent prediction method and device

Abstract

The invention discloses an urban road bottleneck node congestion intelligent prediction method and device. The device comprises an image acquisition module, a central processing module and a display module. The image acquisition module is used for acquiring vehicle information of urban road bottleneck nodes to obtain the number of vehicles; the central processing module is used for calculating the number of vehicles according to the number of vehicles and predicting the bottleneck node states of the urban road, including the congestion state, the predicted congestion duration and the congestion dissipation time; and the display module is used for displaying the operation state of the bottleneck node road section.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic management, and particularly to an intelligent prediction method and device for congestion at bottleneck nodes on urban roads. Background Art

[0002] In recent years, the problem of urban traffic congestion has become increasingly severe, greatly affecting the travel of citizens and the operation efficiency of cities. Solving the problem of urban traffic congestion is extremely urgent. In particular, recurrent traffic congestion has a great impact on citizens' travel. Recurrent traffic congestion mostly occurs at bottleneck traffic nodes. For example, due to the reduction of the number of lanes, signal control, and illegal parking, bottleneck nodes are formed. During peak hours, congestion at bottleneck nodes spreads rapidly upstream along the road, causing large-scale traffic congestion and paralyzing traffic. How to dynamically monitor and warn bottleneck nodes is the focus of current work in the field of urban traffic management.

[0003] The existing traffic management for bottleneck nodes mainly focuses on traditional traffic volume statistics and road running speed monitoring, and is unable to analyze and predict the process of congestion propagation and dissipation, cannot meet the requirements of high-precision navigation warning information, and lacks an intelligent prediction algorithm for congestion at bottleneck nodes and the corresponding integrated information system device.

[0004] Therefore, using existing information communication, data acquisition, and artificial intelligence technologies to conduct high-precision traffic monitoring, congestion propagation, and dissipation time estimation for bottleneck nodes is of great significance for strengthening the monitoring and management of bottleneck nodes. Summary of the Invention

[0005] Aiming at the technical problem that there is no intelligent prediction of congestion for bottleneck nodes in the prior art, the present invention proposes an intelligent prediction method and device for congestion at bottleneck nodes on urban roads.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] An intelligent prediction device for congestion at bottleneck nodes on urban roads, comprising an image acquisition module, a central processing module, and a display module;

[0008] The image acquisition module is used to collect vehicle information at bottleneck nodes on urban roads to obtain the number of vehicles;

[0009] The central processing module calculates the state of bottleneck nodes on urban roads according to the number of vehicles, including congestion status, predicted congestion duration, and congestion dissipation time;

[0010] The display module is used to display the running state of the bottleneck node section.

[0011] Preferably, it further includes a navigation module installed on the vehicle, which is used to provide users with information about bottleneck nodes on urban roads, including predicted congestion duration and congestion dissipation time.

[0012] Preferably, the central processing module is built-in with a storage unit for storing vehicle information and urban road network information.

[0013] Preferably, the image acquisition module and the display module are installed on the installation pole.

[0014] Preferably, the installation pole includes a vertical column, a transverse rod, and a connecting component; the vertical column is fixed in the foundation pit concrete, and the transverse rod is installed at the top of the vertical column through the connecting component.

[0015] Preferably, the installation poles are arranged at intervals of 50 m on urban roads.

[0016] The present invention also provides an intelligent prediction method for congestion at bottleneck nodes on urban roads, which specifically includes the following steps:

[0017] S1: Real-time collect the vehicle information of the bottleneck nodes on urban roads to obtain the number of vehicles, and send it to the central processing module;

[0018] S2: The central processing module predicts the status of the bottleneck nodes on urban roads according to the number of vehicles;

[0019] S3: Perform different displays according to the status of the bottleneck nodes.

[0020] Preferably, the S1 includes:

[0021] S1-1: Build a vehicle detection model and complete the training:

[0022] Search for "car" pictures in Baidu Images and download 1000 relevant pictures. Run the Labelimg software to calibrate and annotate the vehicle information, and establish an annotated training set U containing 1000 pictures;

[0023] Import the annotated training set U into the yolov5 model for image target detection training, assign a unique digital code "Y-M-D-S-h-n" to each vehicle, representing the year-month-day-image acquisition device id-hour-number of vehicles information, and obtain the trained vehicle detection model;

[0024] S1-2: Import the collected picture information into the trained vehicle detection model, add a unique digital code to the detected vehicles, and sum to calculate the number of vehicles M.

[0025] Preferably, the S2 includes:

[0026] S2-1: Calculate the vehicle difference according to the number of vehicles upstream and downstream of the bottleneck nodes on urban roads, and judge the status of the bottleneck nodes. If it is in a smooth state, send a "smooth" signal to the display module; if it is in a congested state, send a "congested" signal to the display module and enter S2-2;

[0027] q = |12 * (M 1 - M 0 )| (1)

[0028] In formula (1), q represents the vehicle difference between the first image acquisition module downstream of the bottleneck node and the first image acquisition module upstream of the bottleneck node; M 1 represents the number of vehicles in the cross-section of the first image acquisition module upstream of the bottleneck node; M 0 represents the number of vehicles in the first image acquisition module downstream of the bottleneck node;

[0029] When q ≤ q 预 , it represents a smooth state; when q > q 预 , it represents a congested state;

[0030] S2-2: Calculate the congestion propagation speed from the first image acquisition module upstream of the bottleneck node to the bottleneck node based on the vehicle difference:

[0031]

[0032] In formula (2), v w represents the congestion propagation speed from the first image acquisition module upstream of the bottleneck node to the bottleneck node; v f represents the free flow speed of urban roads. The free flow speeds of expressways, arterial roads, sub-arterial roads, and branch roads are 80 km / h, 60 km / h, 50 km / h, and 40 km / h respectively; ρ j represents the blocking density of the j-th lane;

[0033] S2-3: Calculate the congestion propagation time based on the congestion propagation speed:

[0034]

[0035] In formula (3), T s represents the congestion propagation time; L represents the distance from the bottleneck node to the first entrance / exit upstream; l j represents the distance from the j-th entrance / exit to the j-1-th entrance / exit; v j,w represents the congestion propagation speed within the j-th entrance / exit section; q i,main represents the number of vehicles in the cross-section within the i-th entrance / exit section; C , is the road capacity, C , = k * 1500, where k is the number of lanes;

[0036] S2-4: Finally, calculate the congestion dissipation time:

[0037]

[0038] In formula (4), T d represents the congestion dissipation time; L , represents the congestion mileage; v d represents the difference between the number of vehicles at the i-th inlet and outlet section and the number of vehicles M 0 at the first image acquisition module downstream of the bottleneck node;

[0039] S2-5: Calculate the predicted congestion duration according to the congestion propagation time and the congestion dissipation time:

[0040] T = T s + T d (6)

[0041] In formula (6), T represents the predicted congestion duration; T s represents the congestion propagation time; T d represents the congestion dissipation time.

[0042] Preferably, in S3, the display method is:

[0043] When the display module receives the "smooth" signal, "smooth" is displayed at the top and a green line is displayed at the bottom; when the "congestion" signal is received, "congestion" is displayed at the top, and at the same time, "predicted congestion duration T" is displayed, and a red line is displayed at the bottom.

[0044] In summary, due to the adoption of the above technical solutions, compared with the prior art, the present invention has at least the following beneficial effects:

[0045] The present invention provides an intelligent prediction algorithm and information device for urban road bottleneck node congestion. When the system is working, the image acquisition device collects traffic picture information of the road where it is located, and then imports the vehicle detection model M (yolov5) that has been trained. According to the road traffic flow calculation algorithm, the number of vehicles is counted, and then the vehicle number information is transmitted to the central processing device through the information transmission device. The central processing device calculates the vehicle number difference q. When q ≤ 20, it is determined that the road section is in a smooth state at this time, and a "smooth" signal will be sent to the display device; in step 3, when q > 20, it is determined that the road section is in a congested state at this time, and a "congestion" signal will be sent to the display device, and at the same time, the congestion propagation time is calculated, and the congestion dissipation time is calculated, and then the congestion duration is calculated. When the display device receives the "smooth" signal, "smooth" is displayed at the top of the LED display screen, and a green line is displayed at the bottom; when the "congestion" signal is received, "congestion" is displayed at the top of the LED display screen, and at the same time, "predicted congestion duration: T" is displayed, and a red line is displayed at the bottom. When the in-vehicle navigation device is clicked to enable, a voice prompt "predicted congestion duration: T" can be issued. Brief Description of the Drawings

[0046] Figure 1 Schematic diagram of an intelligent congestion prediction device for urban road bottleneck nodes according to an exemplary embodiment of the present invention.

[0047] Figure 2 Schematic diagram of the installation positions of the image acquisition module and the display module according to an exemplary embodiment of the present invention.

[0048] Figure 3 Schematic diagram of an intelligent congestion prediction method for urban road bottleneck nodes according to an exemplary embodiment of the present invention. Detailed implementation manners

[0049] The present invention will be further described in detail below in conjunction with embodiments and specific implementation manners. However, it should not be understood that the scope of the above-mentioned subject matter of the present invention is limited to the following embodiments. All technologies implemented based on the content of the present invention belong to the scope of the present invention.

[0050] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention.

[0051] As Figure 1 shown, the present invention proposes an intelligent congestion prediction device for urban road bottleneck nodes, which includes an image acquisition module 1, a central processing module 2, a display module 3, and a navigation module 4.

[0052] The output end of the image acquisition module 1 is connected to the input end of the central processing module 2, and the output end of the central processing module 2 is respectively connected to the input ends of the display module 3 and the navigation module 4.

[0053] In this embodiment, the image acquisition module 1 can adopt a spherical infrared high-definition camera, which is used to collect vehicle information of urban road bottleneck nodes and obtain the number of vehicles.

[0054] The central processing module 2 calculates the state of urban road bottleneck nodes according to the number of vehicles, including congestion state, expected congestion duration, congestion dissipation time, etc.

[0055] In this embodiment, the central processing module is built-in with a storage unit (hard disk) for storing vehicle information and urban road network information.

[0056] The display module 3 can adopt an LED display screen, which is used to display the operation state of the bottleneck node section.

[0057] In this embodiment, when the display module 3 receives the "smooth" signal sent by the central processing module 2, the word "smooth" is displayed at the top of the display module 3 and shown with green lines at the bottom; when receiving the "congested" signal, the word "congested" is displayed at the top, and at the same time, "Estimated congestion duration: T" is displayed, and shown with red lines at the bottom.

[0058] The navigation module 4, which can be installed inside the vehicle, is used to provide users with information about the bottleneck nodes of urban roads, including congestion duration, congestion dissipation time, etc.; for example, the navigation module is a navigation software system, and clicking it can send out a voice prompt of "Estimated congestion duration".

[0059] In this embodiment, as Figure 2 shown, the image acquisition module 1 and the display module 3 are installed on the installation rod.

[0060] The installation rod includes a vertical column, a horizontal rod, and a connection component; the vertical column is fixed in the foundation pit concrete, and the foundation pit concrete is completed by concrete pouring and buried in the road foundation pit; the horizontal rod is installed at the top of the vertical column body through the connection component. Both the vertical column body and the horizontal rod are hollow steel circular materials. There is an installation slot in the middle of the vertical column body, which can be connected to the alarm prompt module; the image acquisition module 1 and the display module 3 are installed on the horizontal rod.

[0061] In this embodiment, the installation rods are arranged at intervals of 50m.

[0062] Based on the above device, as Figure 3 shown, the present invention also provides an intelligent prediction method for congestion at bottleneck nodes of urban roads, which specifically includes the following steps:

[0063] S1: Real-time collect the vehicle information of the bottleneck nodes of urban roads to obtain the number of vehicles, and send it to the central processing module.

[0064] S1-1: Build a vehicle detection model and complete the training.

[0065] In this embodiment, search for "car" pictures on Baidu Images and download 1000 relevant pictures. Run the Labelimg software to calibrate and label the vehicle information, and establish a labeled training set U containing 1000 pictures; import the training set U into the yolov5 model for picture target detection training, and assign a unique digital code "Y-M-D-S-h-n" to each vehicle, representing year-month-day-image acquisition device id-hour-vehicle number information, to obtain the trained vehicle detection model M (yolov5);

[0066] S1-2: Import the collected picture information (saved in JPEG format) into the trained vehicle detection model M (yolov5), add a unique digital code to the detected vehicles, take a 5-minute statistical interval, sum up the data with unique digital codes within the statistical period t to calculate the number of vehicles M, and then transmit the number of vehicles to the central processing module through the information transmission module.

[0067] S2: Predict the status of the bottleneck nodes on urban roads based on the number of vehicles.

[0068] S2-1: Calculate the vehicle difference based on the number of vehicles upstream and downstream of the bottleneck nodes on urban roads, and judge the status of the bottleneck nodes. If it is in a smooth state, send a "smooth" signal to the display module; if it is in a congested state, send a "congested" signal to the display module and enter S2-2.

[0069] q = |12*(M 1 -M 0 )| (1)

[0070] In formula (1), q represents the vehicle difference between the first image acquisition module downstream of the bottleneck node and the first image acquisition module upstream of the bottleneck node; M 1 represents the number of vehicles at the cross-section of the first image acquisition module upstream of the bottleneck node; M 0 represents the number of vehicles at the first image acquisition module downstream of the bottleneck node.

[0071] When q ≤ q 预 , it represents a smooth state, and a "smooth" signal will be sent to the display module; when q > q 预 , it represents a congested state, and a "congested" signal will be sent to the display module;

[0072] S2-2: Calculate the congestion propagation speed from the first image acquisition module upstream of the bottleneck node to the bottleneck node based on the vehicle difference:

[0073]

[0074] In formula (2), v w represents the congestion propagation speed from the first image acquisition module upstream of the bottleneck node to the bottleneck node; v f represents the free flow speed of urban roads. The free flow speeds of expressways, arterial roads, sub-arterial roads, and branch roads are 80 km / h, 60 km / h, 50 km / h, and 40 km / h respectively; ρ j represents the blocking density of the jth lane, for example, it is 200 vehicles / km;

[0075] S2-3: Calculate the congestion propagation time based on the congestion propagation speed:

[0076]

[0077] In formula (3), T s represents the congestion propagation time; L represents the distance from the bottleneck node to the first inlet / outlet upstream; l j represents the distance from the j-th inlet / outlet to the (j - 1)-th inlet / outlet, and i represents the total number of inlets / outlets; v j,w represents the congestion propagation speed within the j-th inlet / outlet section, which can be calculated according to formula (2); q i,main represents the number of vehicles in the cross-section within the i-th inlet / outlet section; C , is the road capacity, C , = k * 1500, where k is the number of lanes.

[0078] In this embodiment, a congestion propagation time data signal is sent to the display module and the navigation module.

[0079] S2-4: Finally, calculate the congestion dissipation time:

[0080]

[0081] In formula (4), T d represents the congestion dissipation time; L , represents the congestion mileage; v d represents the difference between the number of vehicles in the cross-section of the i-th inlet / outlet and the number of vehicles M 0 of the first image acquisition module downstream of the bottleneck node. When the difference is negative, the congestion propagation speed is the congestion dissipation speed, and in this case, there is:

[0082]

[0083] In formula (5), L represents the distance from the bottleneck node to the first inlet / outlet upstream; l j represents the distance from the j-th inlet / outlet to the (j - 1)-th inlet / outlet; q i,main represents the number of vehicles in the cross-section within the i-th inlet / outlet section; C , is the road capacity, C , = k * 1500, where k is the number of lanes.

[0084] S2-5: Calculate the predicted congestion duration based on the congestion propagation time and the congestion dissipation time, and send the predicted congestion duration to the display module and the navigation module, updating every 5 minutes:

[0085] T = T s + T d (6)

[0086] In formula (6), T represents the predicted congestion duration; T s represents the congestion propagation time; Td Indicates the congestion dissipation time.

[0087] S3: Make different displays according to the status of the bottleneck node to remind the user to pay attention.

[0088] In this embodiment, when the display module receives the "unobstructed" signal, "unobstructed" is displayed at the top and shown with green lines at the bottom; when the "congested" signal is received, "congested" is displayed at the top, and at the same time, "estimated congestion duration T" is displayed, and shown with red lines at the bottom.

[0089] Those of ordinary skill in the art can understand that the above embodiments are specific embodiments for implementing the present invention, and in actual applications, various changes can be made in form and details without departing from the spirit and scope of the present invention.

Claims

1. An intelligent prediction device for congestion at bottleneck nodes of urban roads, characterized in that: It includes image acquisition module, central processing module and display module; An image acquisition module is used to collect vehicle information at bottleneck nodes on urban roads to obtain the number of vehicles; The central processing module calculates the number of vehicles and predicts the status of bottleneck nodes on urban roads, including congestion status, predicted congestion duration, and congestion dissipation time; The display module is used to display the operating status of the bottleneck node section.

2. The intelligent prediction device for congestion at bottleneck nodes of urban roads according to claim 1, characterized in that: It also includes a navigation module installed on the vehicle, which is used to provide users with information on bottleneck nodes on urban roads, including the predicted duration of congestion and the time it takes for the congestion to dissipate.

3. The intelligent prediction device for urban road bottleneck node congestion according to claim 1, characterized in that: The central processing module is equipped with a storage unit for storing vehicle information and urban road network information.

4. The intelligent prediction device for urban road bottleneck node congestion according to claim 1, characterized in that: The image acquisition module and the display module are installed on the installation rod.

5. The intelligent prediction device for urban road bottleneck node congestion as claimed in claim 4, characterized in that: The installation rod comprises a column, a transverse rod and a connecting assembly; the column is fixed in the foundation pit concrete, and the transverse rod is installed on the top of the column through the connecting assembly.

6. The intelligent prediction device for urban road bottleneck node congestion according to claim 4, characterized in that: The installation poles are arranged every 50 m on the urban roads.

7. An intelligent prediction method for urban road bottleneck node congestion based on the device according to any one of claims 1 to 6, characterized in that: The specific steps include: S1: Collect vehicle information at bottleneck nodes of urban roads in real time to obtain the number of vehicles and send it to the central processing module; S2: The central processing module predicts the status of bottleneck nodes on urban roads based on the number of vehicles; S3: Display differently according to the status of the bottleneck node.

8. The method for intelligent prediction of urban road bottleneck node congestion as claimed in claim 7, characterized in that: The S1 includes: S1-1: Build a vehicle detection model and complete training: Search for "car" in Baidu pictures and download 1,000 related pictures. Run Labelimg software to calibrate vehicle information and create a labeled training set U containing 1,000 pictures. Import the labeled training set U into the yolov5 model for image target detection training, assign a unique digital code "YMDShn" to each vehicle, representing year-month-day-image acquisition device id-hour-number of vehicles, and obtain the trained vehicle detection model; S1-2: Import the collected image information into the trained vehicle detection model, add a unique digital code to the detected vehicles, and calculate the number of vehicles M by summing them up.

9. The intelligent prediction method for urban road bottleneck node congestion according to claim 7, characterized in that: The S2 includes: S2-1: Calculate the vehicle difference according to the number of vehicles upstream and downstream of the bottleneck node of the urban road, and judge the state of the bottleneck node. If it is unobstructed, a "unobstructed" signal will be sent to the display module; if it is congested, a "congested" signal will be sent to the display module and enter S2-2; q=|12*M1-M0)|(1) In formula (1), q represents the difference between the first image acquisition module downstream of the bottleneck node and the first image acquisition module upstream of the bottleneck node; M1 represents the number of vehicles in the cross section of the first image acquisition module upstream of the bottleneck node; M0 represents the number of vehicles in the first image acquisition module downstream of the bottleneck node; When q≤q 预 When q>q 预 When , it indicates congestion; S2-2: Calculate the congestion propagation speed from the first image acquisition module upstream of the bottleneck node to the bottleneck node based on the vehicle difference: In formula (2), v w represents the congestion propagation speed from the first image acquisition module upstream of the bottleneck node to the bottleneck node; v f represents the free flow speed of urban roads. The free flow speeds of expressways, trunk roads, secondary trunk roads and branch roads are 80km / h, 60km / h, 50km / h and 40km / h respectively; ρ j represents the blocking density of the jth lane; S2-3: Calculate the congestion propagation time based on the congestion propagation speed: In formula (3), T s represents the congestion propagation time; L represents the distance between the bottleneck node and the first upstream entrance and exit; l j represents the distance from the jth entrance to the j-1st entrance; v j,w represents the congestion propagation speed within the jth import and export section; q i,main represents the number of vehicles in the i-th import and export section; C , is the road capacity, C , =k*1500, k is the number of lanes; S2-4: Finally calculate the congestion dissipation time: In formula (4), T d Indicates the congestion dissipation time; L , represents congestion mileage; v d It represents the difference between the number of vehicles at the i-th import and export section and the number of vehicles M0 of the first image acquisition module downstream of the bottleneck node; S2-5: Calculate the predicted congestion duration based on the congestion dissipation time and congestion dissipation time: T=T s +T d (6) In formula (6), T represents the predicted congestion duration; T s represents the congestion propagation time; T d Indicates the time it takes for congestion to dissipate.

10. The intelligent prediction method for urban road bottleneck node congestion according to claim 7, characterized in that: In S3, the display method is: When the display module receives a "smooth" signal, "smooth" is displayed at the top and displayed with green lines at the bottom; when a "congested" signal is received, "congested" is displayed at the top and "estimated congestion duration T" is displayed at the bottom, and displayed with red lines at the bottom.

Images