A method and system for regional boundary control based on traffic congestion index

Abstract

This invention provides a regional boundary control method and system based on traffic congestion index, belonging to the field of traffic control technology. The method includes: collecting historical traffic data of the target traffic network according to time parameters and constructing road network topology maps for each time period; clustering each road network topology map using a spectral clustering algorithm; constructing a three-dimensional macroscopic basic graph model based on the segment length, cumulative vehicle count, and traffic completion rate of each traffic center area after clustering; collecting current traffic data of the target traffic network and constructing a target road network topology map; clustering the target road network topology map to obtain the target traffic center area using a spectral clustering algorithm; and regulating the traffic lights at the boundary intersections of the target traffic center area using a sliding mode variable structure control algorithm based on the three-dimensional macroscopic basic graph model. By controlling the boundary flow, the flow within the region is adjusted to ensure that the cumulative vehicle count of the regional road network is always maintained at the critical vehicle count, guaranteeing that the regional traffic completion flow reaches its maximum.

Description

Technical Field

[0001] This invention mainly relates to the field of traffic control technology, specifically to a regional boundary control method and system based on a traffic congestion index. Background Technology

[0002] The Macro-Based Map (MFD) describes the relationship between traffic flow and average density in a road network, reflecting the regional traffic conditions. Current traffic control technologies delineate congestion zones within a target traffic area based on the MFD, and implement boundary entry controls at intersections or road segments outside these zones to limit traffic flow into the congested areas, thus addressing congestion. However, this boundary control technology relies solely on historical data to determine the MFD, lacking the application of mobile internet to obtain real-time data. Furthermore, most studies statically divide the road network into appropriately sized and structured sub-zones, neglecting the fact that urban traffic is a complex, spatiotemporally changing nonlinear system. This means that congestion points within the road network exhibit time-varying traffic characteristics, and congestion spreads to adjacent areas.

[0003] Traditional static zoning control methods fail to reflect regional changes caused by traffic variations and are prone to congestion in non-target traffic areas, making it difficult for the entire traffic network to achieve optimal operation. Traditional dynamic zoning control methods only consider two factors: vehicle completion rate and cumulative vehicle count, without taking into account the impact of changes in traffic area size on the macroscopic basic map, making it difficult to implement precise traffic control strategies. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a regional boundary control method and system based on traffic congestion index, which addresses the shortcomings of the existing technology.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] A regional boundary control method based on traffic congestion index includes the following steps:

[0007] Historical traffic data of the target traffic network is collected according to time parameters to obtain the segment lengths of multiple roads and the traffic congestion index of each road in multiple different time periods.

[0008] Images of multiple roads and their corresponding traffic congestion indices are constructed according to multiple different time periods to obtain road network topology maps for each time period. The multiple road network topology maps are then clustered using a spectral clustering algorithm to obtain traffic center areas for multiple time periods. Each traffic center area consists of multiple central roads.

[0009] The cumulative number of vehicles and traffic completion rate of traffic center areas for multiple time periods are obtained respectively. A three-dimensional macro basic graph model is constructed based on the cumulative number of vehicles, the traffic completion rate and the road segment length of multiple central roads in each traffic center area.

[0010] Traffic data of the target traffic network at the current time is collected to obtain the target traffic congestion index of each road. An image is constructed by multiple roads and their corresponding target traffic congestion indices to obtain the target road network topology map. The target road network topology map is clustered by a spectral clustering algorithm to obtain the target traffic center area. The target traffic center area includes multiple boundary intersections equipped with traffic lights.

[0011] The sliding mode variable structure control algorithm calculates the number of vehicles in the target traffic center area based on the three-dimensional macroscopic basic graph model to obtain the number of controlled vehicles. Based on the number of controlled vehicles, the change duration of traffic lights at multiple boundary intersections is calculated to obtain the green light adjustment duration corresponding to multiple green lights. The display duration of the green lights at the corresponding boundary intersections is adjusted according to the multiple green light adjustment durations.

[0012] Another technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0013] A regional boundary control system based on a traffic congestion index, comprising:

[0014] The data acquisition module is used to collect historical traffic data of the target traffic network according to time parameters, and obtain the segment lengths of multiple roads and the traffic congestion index of each road in multiple different time periods.

[0015] The clustering module is used to construct images of multiple roads and their corresponding traffic congestion indices according to multiple different time periods, to obtain the road network topology map corresponding to each time period. The multiple road network topology maps are clustered by the spectral clustering algorithm to obtain the traffic center area for multiple time periods. The traffic center area is composed of multiple central roads.

[0016] The modeling module is used to obtain the cumulative number of vehicles and traffic completion rate corresponding to the traffic center area in multiple time periods, and to construct a three-dimensional macro basic graph model based on the multiple cumulative number of vehicles, the multiple traffic completion rates and the road segment lengths of multiple central roads in each traffic center area.

[0017] The data acquisition module is also used to collect traffic data of the target traffic network at the current time to obtain the target traffic congestion index for each road.

[0018] The clustering module is also used to construct images of multiple roads and their corresponding target traffic congestion indices to obtain a target road network topology map. The target road network topology map is then clustered using a spectral clustering algorithm to obtain a target traffic center area, which includes multiple boundary intersections equipped with traffic lights.

[0019] The control module is used to calculate the number of vehicles in the target traffic center area based on the three-dimensional macroscopic basic graph model using a sliding mode variable structure control algorithm, to obtain the number of controlled vehicles, and to calculate the change duration of traffic lights at multiple boundary intersections based on the number of controlled vehicles, to obtain the green light adjustment duration corresponding to multiple green lights, and to adjust the display duration of the green lights at the corresponding boundary intersections according to the multiple green light adjustment durations.

[0020] The beneficial effects of this invention are as follows: By constructing a road network topology map based on the road lengths and traffic congestion indices obtained from the target traffic network, roads with similar congestion conditions are connected in the road network topology map. A spectral clustering algorithm is used to cluster roads according to their congestion conditions, identifying the traffic center area composed of the most congested roads. A three-dimensional macroscopic basic graph model is constructed based on vehicle operation data of the traffic center area at different time periods to reflect the congestion situation at each time period. Congested road segments at the current stage are calculated based on real-time congestion data of the target traffic network. The signal light durations at the intersections of the most congested road segments at the current stage are adjusted based on historical congestion data from the three-dimensional macroscopic basic graph model, allowing as many vehicles as possible to pass through the congested road segments while limiting the number of vehicles entering the congested road segments, thereby alleviating road congestion.

[0021] Based on a three-dimensional macroscopic basic graph model, the critical flow rate for the number of vehicles within a congested area is calculated. By adjusting the display duration of traffic lights in the congested area, precise allocation and optimized management of traffic flow are achieved. This rapidly reduces the number of vehicles in the congested area and keeps them within the critical flow range, alleviating congestion problems in traditional traffic management, thereby improving the overall operational efficiency of the road network and reducing environmental pollution. Dynamic adjustments to real-time traffic data offer strong practicality and flexibility. By dynamically controlling the display duration of traffic lights at the boundary intersections of congested road sections in the target traffic center, the number of congested road sections is controlled, enhancing adaptability and robustness under complex traffic conditions and preventing localized excessive congestion caused by the spread of traffic congestion. Attached Figure Description

[0022] Figure 1 A flowchart of a regional boundary control method based on traffic congestion index provided in an embodiment of the present invention;

[0023] Figure 2 A flowchart of a region boundary control method provided in an embodiment of the present invention;

[0024] Figure 3 This is a schematic diagram of the control principle of the three-dimensional macroscopic basic graphical model provided in the embodiments of the present invention;

[0025] Figure 4 A schematic diagram of the target traffic network provided in an embodiment of the present invention;

[0026] Figure 5 A schematic diagram of a vehicle guidance strategy provided in an embodiment of the present invention;

[0027] Figure 6 A block diagram of a regional boundary control system based on a traffic congestion index provided in an embodiment of the present invention. Detailed Implementation

[0028] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0029] Nowadays, travelers are accustomed to using the internet and mobile internet terminals to obtain traffic information for navigation and make travel decisions based on the directional routes provided by the navigation. However, the directional routes provided are based on the user's optimal guidance strategy. The provided directional guidance strategy is not coordinated with the city's traffic signal control scheme, which will aggravate the delay at intersections and fail to guide the balanced distribution of traffic flow. It is easy for traffic congestion to occur due to the optimal route recommended by the navigation software.

[0030] Therefore, a traffic network boundary control strategy that considers changes in vehicle completion rate, cumulative number of vehicles, and area size of traffic zones is needed to improve the traffic flow completion rate within the controlled traffic network in order to implement precise traffic control strategies.

[0031] like Figure 1 and Figure 2 As shown in the figure, an embodiment of the present invention provides a regional boundary control method based on a traffic congestion index, which includes the following steps:

[0032] Historical traffic data of the target traffic network is collected according to time parameters to obtain the segment lengths of multiple roads and the traffic congestion index of each road in multiple different time periods.

[0033] Images of multiple roads and their corresponding traffic congestion indices are constructed according to multiple different time periods to obtain road network topology maps for each time period. The multiple road network topology maps are then clustered using a spectral clustering algorithm to obtain traffic center areas for multiple time periods. Each traffic center area consists of multiple central roads.

[0034] The cumulative number of vehicles and traffic completion rate of traffic center areas for multiple time periods are obtained respectively. A three-dimensional macro basic graph model is constructed based on the cumulative number of vehicles, the traffic completion rate and the road segment length of multiple central roads in each traffic center area.

[0035] Traffic data of the target traffic network at the current time is collected to obtain the target traffic congestion index of each road. An image is constructed by multiple roads and their corresponding target traffic congestion indices to obtain the target road network topology map. The target road network topology map is clustered by a spectral clustering algorithm to obtain the target traffic center area. The target traffic center area includes multiple boundary intersections equipped with traffic lights.

[0036] The sliding mode variable structure control algorithm calculates the number of vehicles in the target traffic center area based on the three-dimensional macroscopic basic graph model to obtain the number of controlled vehicles. Based on the number of controlled vehicles, the change duration of traffic lights at multiple boundary intersections is calculated to obtain the green light adjustment duration corresponding to multiple green lights. The display duration of the green lights at the corresponding boundary intersections is adjusted according to the multiple green light adjustment durations.

[0037] It should be understood that data on the target traffic network structure is collected through the API interface of navigation map software to obtain the connection structure of multiple roads and the corresponding road segment lengths. The time collection parameter is set to 5 minutes. Traffic data such as traffic congestion index, cumulative vehicle count, and traffic completion rate are obtained through the traffic status query API interface of navigation map software. All traffic data are obtained through statistics from users of navigation map software. The scope of the target traffic network is typically as large as a city-level area and as small as a hotspot area, such as the intersection around a train station.

[0038] The Traffic Situation Query API is provided via HTTP and is used to query traffic conditions within a specified route, circular area, or rectangular area. It can return the desired traffic situation based on user input. This service is suitable for scenarios such as real-time traffic information query, traffic data analysis, and intelligent transportation systems.

[0039] In this embodiment of the invention, historical traffic data is accurately analyzed, and congested road segments in the traffic flow are calculated. A three-dimensional macroscopic basic graph model is constructed based on vehicle operation data in congested road segments corresponding to each historical time period to reflect the congestion situation at different times and for different road segment lengths. Real-time traffic operation status data is obtained. Combining sliding mode variable structure control theory, the real-time traffic data of the target traffic network is regulated and analyzed based on the three-dimensional macroscopic basic graph model. Traffic flow in congested road segments is reasonably controlled through traffic light duration control, effectively improving the operational efficiency of the macroscopic road network and providing accurate analysis, prediction, and decision-making basis for urban traffic management and control.

[0040] Preferably, the step of constructing images of multiple roads and their corresponding traffic congestion indices according to multiple different time periods to obtain road network topology maps for each time period includes:

[0041] S11. Calculate the similarity between traffic congestion indices for any given time period using a similarity function to obtain the congestion similarity between multiple roads within that time period. The similarity function is:

[0042]

[0043] Among them, w ij Let x be the congestion similarity between the i-th road and the j-th road. i Let x be the traffic congestion index of the i-th road. j Let σ be the traffic congestion index of the j-th road. 2 Let be the variance of the traffic congestion index, and ||·|2 be the L2 norm.

[0044] S12. Multiple roads are treated as multiple road nodes, and the congestion similarity of the multiple roads is used as the weight value of multiple congestion edges. Based on the weight value, the multiple congestion edges are connected to the corresponding road nodes to construct a road network topology graph.

[0045] S13. Repeat S11 to S12 to process multiple roads and their corresponding traffic congestion indices for multiple different time periods to obtain the road network topology map for each time period.

[0046] Specifically, each road in the target traffic network is treated as a node V in an undirected graph, and the traffic congestion index of each road is represented by the node value x. The weight w of the edge E of the undirected graph is calculated based on the node values ​​of each node, which is the congestion similarity between road segments, and is expressed as:

[0047]

[0048] Among them, w ij Let V = {1, 2, ..., N} be the congestion similarity between the i-th road and the j-th road. This is equivalent to X = (x1, x2, ..., xj). n ), E = W = [w ij ], 1≤i, j≤N, (i,j)∈E indicates that the i-th node and the j-th node are connected. This indicates that the i-th node and the j-th node are not connected;

[0049] By connecting the nodes according to their weights, a weighted undirected graph (i.e., a road network topology graph) is obtained.

[0050] In this embodiment of the invention, by calculating the similarity of congestion conditions of each road segment, road segments with similar and adjacent congestion conditions are grouped into one category based on the similarity, so as to facilitate the clustering of congested road segments.

[0051] Preferably, the step of clustering multiple road network topology maps using a spectral clustering algorithm to obtain traffic center areas for multiple time periods includes:

[0052] S21. Sum the edges corresponding to any node in any road network topology to obtain the connectivity value. Repeat this process for all nodes to obtain the connectivity value of each node.

[0053] S22. The road network topology is segmented according to multiple connectivity values ​​using a spectral clustering calculation expression to obtain initial segmentation lines. These initial segmentation lines are then minimized using an objective function to obtain further segmentation lines. The road network topology is then divided according to these segmentation lines to obtain the traffic center area in the road network topology for any given time period. The spectral clustering calculation expression is:

[0054]

[0055] Where Ncut(V) is the initial segmentation line, A k For the road node of the kth region category, A k For A k The complement, d i Let K be the connectivity value of the i-th road node, and K be the number of region categories.

[0056] S23. Repeat S21 to S22 to process all road network topology maps and obtain traffic center areas for multiple time periods.

[0057] Specifically, the target traffic network is divided into control sub-regions based on the similarity between road segments, including:

[0058] Calculate the connectivity value of each node (i.e., the weight of the edges connected to that node). The connectivity calculation expression is:

[0059]

[0060] Where, d i w represents the connectivity value of the i-th node (i.e., the road). ij Let N be the congestion similarity between the i-th road and the j-th road, and N be the number of roads.

[0061] The road network topology is segmented based on the connectivity values ​​of each node and the number of unconnected road nodes in the topology map using spectral clustering calculation expressions. This results in the traffic center area and the corresponding peripheral area. The spectral clustering calculation expression is as follows:

[0062]

[0063] The road network topology diagram is represented by the following segmentation:

[0064]

[0065] As should be understood, the spectral clustering algorithm treats all data as points in space, which can be connected by lines. The edge weight between two points that are far apart is lower, while the weight between two points that are close together is higher. By dividing the graph composed of all data points, the algorithm aims to minimize the sum of edge weights between different subgraphs, i.e., to make the congestion situation different, and maximize the sum of edge weights within the subgraphs, i.e., to make the congestion situation similar.

[0066] In this embodiment of the invention, congested roads in various time periods of historical data throughout the day are clustered to identify congested central road segments. A control model is then constructed based on traffic flow data for roads with different congestion lengths, which facilitates subsequent vehicle volume adjustment for congested road segments.

[0067] Preferably, the step of constructing a three-dimensional macroscopic basic graph model based on multiple cumulative vehicle counts, multiple traffic completion rates, and multiple central road segment lengths in each of the traffic center areas includes:

[0068] The lengths of the road segments of multiple central roads in each of the aforementioned traffic center areas are summed to obtain the total length of the central road segments corresponding to each of the aforementioned traffic center areas;

[0069] A three-dimensional coordinate system is constructed, with the cumulative number of vehicles as the horizontal axis, the total length of the central road segment as the vertical axis, and the traffic completion rate as the vertical axis. In this way, the cumulative number of vehicles, traffic completion rate, and total length of the central road segment corresponding to each traffic center area are converted into three-dimensional coordinates, resulting in multiple control points. These control points are then imported into the three-dimensional coordinate system to complete the construction of the three-dimensional macroscopic basic map model.

[0070] It should be understood that a common macroscopic fundamental graph model (MFD model) is constructed from two parameters: the cumulative number of vehicles (veh) and the traffic completion rate (veh / s). For a road network with a fixed area, the characteristics of the road network determine the fixed relationship between the cumulative number of vehicles and the traffic completion rate. Therefore, by adjusting the number of vehicles entering and leaving the road network, the maximum traffic completion rate can be achieved. In traffic control, in order to continuously reduce the congestion area, the road network area needs to be used as the third dimension of the MFD model. However, the distribution of road segments in different areas of the same area is different. Therefore, the total length of road segments is used to measure the road network area to construct a three-dimensional macroscopic fundamental graph model (3D-MFD model).

[0071] In this embodiment of the invention, the total length of road segments in the road network is used as the third-dimensional data of the MFD model to construct a three-dimensional macroscopic basic graph model. In this way, when using the sliding mode variable structure control algorithm to control traffic flow, the congested road segments of the road network are combined to gradually reduce the congested area of ​​the target region.

[0072] Preferably, such as Figure 3 As shown, the calculation of the number of vehicles in the target traffic center area using the sliding mode variable structure control algorithm based on the three-dimensional macroscopic basic graph model to obtain the number of controlled vehicles includes:

[0073] The traffic core area is selected from the three-dimensional macro basic map model, and the increment of the target traffic center area relative to the traffic core area is calculated to obtain the road length to be controlled and the number of vehicles to be controlled.

[0074] A sliding surface state function is constructed based on the road length to be controlled, the number of vehicles to be controlled, and the set control gain. The number of vehicles to be controlled is obtained by solving the sliding surface state function using the constant velocity approach law.

[0075] Specifically, the cumulative number of vehicles n in the target traffic center area after traffic control measures are implemented. i The vehicle control threshold should be met. Based on the vehicle control threshold, a sliding surface is constructed on the three-dimensional macroscopic basic graph model. Then, when solving the state function of the sliding surface, the number of controlled vehicles Δn is calculated. i The vehicle control threshold is met. The vehicle control threshold is expressed as:

[0076]

[0077] in, n is the critical cumulative number of vehicles. i ε represents the cumulative number of vehicles in the current area. i For the fluctuation range of traffic control, ε i Take N * 1% to 3%;

[0078] The sliding surface state function is expressed as:

[0079] S(q)=S l (L(q))+βS v (q),

[0080] Where S(q) is the sliding surface state function, S l (L(q)) is the length of the path to be controlled, S v (q) represents the number of vehicles to be controlled, and β represents the control gain;

[0081] Using the constant velocity approach law, the number of vehicles controlled reaches the sliding surface, i.e., S=0. The constant velocity approach law is expressed as:

[0082]

[0083] in, Let sgn(·) be the approach value, sgn(·) be the step function, and μ be the approach rate;

[0084] The process of solving the sliding surface state function using the constant velocity approach law can be expressed as follows:

[0085]

[0086] Where, n i (q) represents the cumulative number of vehicles in the current region, and q represents the number of cycles in the constant-speed approach law process; Find the number of controlled vehicles Δn i .

[0087] It should be understood that the constant velocity approach law mainly addresses the sliding surface by ensuring that the control process can approach the sliding surface at a constant speed, so that the deviation between the state variables and the sliding surface (i.e., the length of the road to be controlled and the number of vehicles to be controlled) can be reduced to zero quickly and stably.

[0088] The goal of the controller (i.e., the sliding mode variable structure control algorithm) is to control the length of the road to be controlled and the number of vehicles to be controlled to simultaneously approach the sliding surface. After the control target is achieved (i.e., the number of vehicles controlled is within the control vehicle threshold range), the traffic flow entering and leaving the target traffic center area reaches a dynamic equilibrium. At this time, the traffic flow in the target traffic center area reaches its maximum, and the traffic operation state of the target traffic center area reaches its optimal level.

[0089] There is a critical cumulative number of vehicles N within the target traffic center area. * Complete the traffic flow G, so that the target traffic center area has the maximum completed traffic flow G. * When the cumulative number of vehicles n i <N * At that time, the central roads of the target traffic center area are in the free flow and steady flow stages, i.e., in a non-congested state; when n i Continue to increase until n i =N * When, G = G * The central roads of the target traffic center area are in an unstable flow stage, i.e., saturated; when n i >N * At that time, G follows n i As traffic volume increases, the central roads in the target traffic center area are in a forced flow phase, i.e., congested. Therefore, by regulating the flow within the area through boundary flow control, the cumulative number of vehicles on the regional road network can be kept near the critical number, thus ensuring maximum traffic flow for the region.

[0090] In this embodiment of the invention, by solving the state function of the constructed sliding surface, the current cumulative number of vehicles is controlled within the optimal critical cumulative number of vehicles, and the control number of vehicles that can eliminate the difference between the cumulative number of vehicles and the optimal critical cumulative number of vehicles is obtained. Then, the duration of traffic lights is adjusted according to the control number of vehicles to reduce congested road sections and alleviate traffic congestion.

[0091] Preferably, the step of selecting the traffic core area from the three-dimensional macroscopic basic map model and calculating the increment of the target traffic center area relative to the traffic core area to obtain the road length to be controlled and the number of vehicles to be controlled includes:

[0092] From the three-dimensional macro basic map model, the traffic center area with the smallest total length of the central road segment is selected as the traffic core area, and the total length of the core road segment of the traffic core area is obtained. The road segment lengths of multiple central roads in the target traffic center area are summed to obtain the total length of the target road segment. The total length of the target road segment is subtracted from the total length of the core road segment to obtain the length of the road to be controlled.

[0093] Obtain the target cumulative number of vehicles in the target traffic center area at the current time, and subtract the target cumulative number of vehicles from the preset critical cumulative number of vehicles to obtain the number of vehicles to be controlled.

[0094] Specifically, the difference in road segment length between the real-time central area and the core area of ​​the road network is calculated. The total length of road segments within the dynamic congestion zone and the number of vehicles within the regional road network are expressed as follows:

[0095] S l (L(q))=L d (q)-L c ,

[0096] S v (q)=n d (q)-N * ,

[0097] Among them, S l (L(q)) is the length of the path to be controlled, L d (q) represents the total length of the target road segment, L c S is the total length of the core road section. v (q) represents the number of vehicles to be controlled, n d (q) represents the target cumulative vehicle count, and N* represents the critical cumulative vehicle count. A corresponding critical cumulative vehicle count is pre-set according to the total length of each road segment.

[0098] In this embodiment of the invention, the deviation between the state variables and the sliding surface is calculated based on the three-dimensional macroscopic basic graph model.

[0099] To clarify the boundary control strategy, the number of phases, phase difference, cycle, and green light loss time of the traffic lights at the intersection in the target traffic center area are set to fixed values. One cycle t of the control cycle T is equal to the effective duration C of the traffic light cycle (e.g., 120 seconds).

[0100] Preferably, the boundary control strategy calculates the change duration of traffic lights at multiple boundary intersections based on the number of controlled vehicles to obtain the green light adjustment duration corresponding to multiple green lights, including:

[0101] S31. Obtain the green light duration corresponding to multiple boundary intersections in the current control cycle and the predicted number of vehicles expected to enter the target traffic center area in the next control cycle. Divide the number of controlled vehicles by the predicted number of vehicles to obtain the control rate.

[0102] S32. Multiply the control rate by any green light duration to obtain the green light adjustment amount at any boundary intersection. Add the green light duration and the green light adjustment amount to obtain the green light adjustment duration.

[0103] S33. Repeat S31 to S32 to process the green light duration of all boundary intersections and obtain the green light adjustment duration corresponding to multiple green lights.

[0104] Specifically, the boundary control strategy constructed in this invention is based on the number of controlled vehicles Δn. i The green light adjustment amount is calculated for the green light duration of roads waiting to enter the target traffic center area from the w-th intersection at the boundary of the target traffic center area during the (t+1)-th control cycle. The expression for calculating the green light adjustment amount is as follows:

[0105]

[0106] Wherein, Δg w,in (t+1) represents the green light adjustment amount, Δn i To control the number of vehicles, n in (t+1) represents the predicted number of vehicles in the (t+1)th control cycle, g w,in (t) represents the green light duration, t represents the statistical time period, t∈kT, k=1,2,3,···,K, T represents the control time cycle duration, K represents the number of times the statistics are performed, w∈W, and W represents the set of intersections that need to be controlled.

[0107] Summing the green light duration and green light adjustment amount, we obtain the green light adjustment duration g of the traffic lights facing the road entering the traffic center area of ​​the control target at the boundary intersection w in the (t+1)th control cycle. w,in (t+1), the expression for calculating the green light adjustment time is:

[0108]

[0109] The system determines whether the green light adjustment duration for any road entering the target traffic center area at any boundary intersection is greater than the set maximum green light duration. If so, the green light adjustment duration is modified to the maximum green light duration. If not, the system determines whether the green light adjustment duration is less than the set minimum green light duration. If so, the green light adjustment duration is modified to the minimum green light duration. If not, the green light adjustment duration remains unchanged (i.e., it is not modified). This process is repeated for all intersections in the target traffic center area to determine the green light adjustment duration for roads entering the target traffic center area.

[0110] It should be understood that the boundary intersection of the target traffic center area is a four-way intersection. Entering the target traffic center area involves either driving straight from the west road onto the east road or turning left from the north road onto the east road. Therefore, vehicles expected to enter the target traffic center area in the next control cycle are located on the west road (for straight traffic) and the north road (for left-turn traffic). The expected number of vehicles entering the target traffic center area in the next control cycle is obtained based on navigation data used by users in the navigation map software. Since some vehicles on the road do not use navigation, the expected number of vehicles is divided by the preset navigation usage rate to obtain the predicted number of vehicles expected to enter the target traffic center area in the next control cycle. The initial value for the traffic light duration is the initial green light time set at the intersection; the control process increases or decreases this initial value.

[0111] Another way to control the duration of traffic lights at the boundary intersections of the target traffic center area is:

[0112] Multiple peak-hour traffic flow directions are defined to obtain the corresponding inbound and outbound ends for each peak period. The traffic lights at multiple boundary intersections corresponding to the inbound ends of each peak period are then set as the controllable traffic lights. For example... Figure 4 As shown, the target traffic center area has 7 boundary intersections. If the main flow of vehicles in the target traffic network during the morning rush hour is from west to east, then the traffic lights at the 1st, 2nd, 3rd, and 4th boundary intersections on the west side will be set as controllable traffic lights. The control rate will be multiplied by the duration of each controllable traffic light to obtain multiple traffic light adjustment values. Each adjustment value will be added to the corresponding duration of the controllable traffic light to obtain the adjustment duration of each controllable traffic light. The traffic lights at the entering boundary intersections will then be adjusted according to the adjustment duration of each controllable traffic light.

[0113] In this embodiment of the invention, by calculating the green light duration at each intersection at the boundary of the main flow direction of vehicle entry in the next control cycle, the signal light timing at the boundary intersection of the target traffic center area is adjusted, thereby controlling the vehicle flow in the target traffic center area.

[0114] Preferably, such as Figure 5 As shown, after the step of adjusting the display duration of the green light at the corresponding boundary intersection according to the multiple green light durations, the boundary control strategy further includes:

[0115] The predicted number of vehicles is calculated based on the set prediction parameters for each green light adjustment duration and the number of roads waiting to enter the target traffic center area. The predicted number of controlled vehicles is obtained by subtracting the predicted number of controlled vehicles from the predicted number of controlled vehicles. The remaining number of vehicles is then divided by a preset guidance compliance rate to obtain the guided number of vehicles. Based on the guided number of vehicles, vehicles waiting to enter the target traffic center area in the next control cycle are guided to roads outside the target traffic center area.

[0116] Specifically, in the boundary control strategy, when it is impossible to fully regulate all vehicle flow expected to enter the target traffic center area in the next control cycle, a set vehicle guidance strategy is activated. The predicted number of controlled vehicles is obtained by calculating the green light duration and the number of roads expected to enter the target traffic center area according to the set prediction parameters. The vehicle prediction calculation expression is as follows:

[0117]

[0118] Among them, c w (t+1) represents the predicted number of vehicles entering the target traffic center area from the w-th intersection at the boundary of the target traffic center area, m represents the number of roads waiting to enter the target traffic center area, gw,in(t+1) represents the green light adjustment duration, t0 represents the time expected for the first vehicle to start and cross the stop line after the green light turns on in the next control cycle, and t i The average time it takes for a vehicle to cross the stop line is denoted by γ, which is a reduction factor. The prediction parameters include t0 and t... i And γ, t0 is set to 2.3 seconds, t i With the time set to 2.5 seconds and γ set to 0.9, the vehicle prediction calculation expression can be represented as:

[0119]

[0120] If driving straight from the west side road into one lane on the east side and turning left from the north side road into one lane on the east side are both ways to enter the target traffic center area, then the number of roads m to enter the target traffic center area is 2.

[0121] The remaining vehicle count is obtained by subtracting the predicted control vehicle count from the predicted vehicle count for all intersections at the target traffic center boundary and all roads leading into the target traffic center. Then, the remaining vehicle count at intersections expected to enter the target traffic center but unable to pass through its boundary in the next control cycle is calculated using the induction calculation expression. This yields the induction vehicle count. Vehicles expected to enter the target traffic center are then guided to detour through the outer area (i.e., the area outside the target traffic center) according to this induction vehicle count. The induction calculation expression is as follows:

[0122] Y(t+1)=n′(t+1) / ρ(t+1),

[0123] Where Y(t+1) is the number of induced vehicles in the (t+1)th period, n′(t+1) is the number of remaining vehicles in the (t+1)th period, ρ(t+1) is the induced compliance rate in the (t+1)th period, and 0<ρ<1.

[0124] It should be understood that the remaining number of vehicles is calculated by subtracting the number of vehicles that have already passed the green light and adjusted (i.e., the number of vehicles that have passed the green light and entered the target traffic center area from the boundary intersection) from the number of vehicles expected to travel from the original route in the navigation software to the target traffic center area. The guidance strategy uses an advanced route planning API interface configured with the navigation map software to send detour reminders to traffic participants using the navigation map software who expect to enter the target traffic center area in the next control cycle of the traffic lights, based on the number of guided vehicles. This involves displaying road change prompts and traffic guidance routes on the navigation map software user's end. The advanced route planning API interface is provided via HTTP and is suitable for scenarios requiring real-time traffic information and route planning to help optimize routes and improve travel efficiency. Since some users of the navigation map software may not accept detour reminders, a guidance compliance rate is set to calculate the remaining number of vehicles to be guided. If the number of vehicles is greater than the remaining number, a detour reminder greater than the remaining number is sent to the corresponding number of traffic participants. If the remaining number of vehicles is 5 and the guidance compliance rate is set to 0.5, the number of guided vehicles is calculated to be 10 using the guidance calculation expression.

[0125] If the total signal light cycle is 120 seconds and the peak traffic period is 2 hours, then the control threshold is set to 60 (that is, the cycle of the boundary control process of the target traffic center area is set to 60 times).

[0126] In this embodiment of the invention, a traffic guidance strategy is activated for vehicles located at the boundary of the target traffic center area and waiting to enter the target traffic center area to guide vehicles to detour and avoid the target traffic center area in order to alleviate traffic congestion. By guiding traffic participants, the traffic pressure in the target traffic center area is alleviated, so that the urban traffic flow is as evenly distributed as possible, and the operational efficiency of the macro network is effectively improved.

[0127] like Figure 6 As shown in the figure, an embodiment of the present invention provides a regional boundary control system based on a traffic congestion index, comprising:

[0128] The data acquisition module is used to collect historical traffic data of the target traffic network according to time parameters, and obtain the segment lengths of multiple roads and the traffic congestion index of each road in multiple different time periods.

[0129] The clustering module is used to construct images of multiple roads and their corresponding traffic congestion indices according to multiple different time periods, to obtain the road network topology map corresponding to each time period. The multiple road network topology maps are clustered by the spectral clustering algorithm to obtain the traffic center area for multiple time periods. The traffic center area is composed of multiple central roads.

[0130] The modeling module is used to obtain the cumulative number of vehicles and traffic completion rate corresponding to the traffic center area in multiple time periods, and to construct a three-dimensional macro basic graph model based on the multiple cumulative number of vehicles, the multiple traffic completion rates and the road segment lengths of multiple central roads in each traffic center area.

[0131] The data acquisition module is also used to collect traffic data of the target traffic network at the current time to obtain the target traffic congestion index for each road.

[0132] The clustering module is also used to construct images of multiple roads and their corresponding target traffic congestion indices to obtain a target road network topology map. The target road network topology map is then clustered using a spectral clustering algorithm to obtain a target traffic center area, which includes multiple boundary intersections equipped with traffic lights.

[0133] The control module is used to calculate the number of vehicles in the target traffic center area based on the three-dimensional macroscopic basic graph model using a sliding mode variable structure control algorithm, to obtain the number of controlled vehicles, and to calculate the change duration of traffic lights at multiple boundary intersections based on the number of controlled vehicles, to obtain the green light adjustment duration corresponding to multiple green lights, and to adjust the display duration of the green lights at the corresponding boundary intersections according to the multiple green light adjustment durations.

[0134] The aforementioned regional boundary control system based on traffic congestion index can be found in the above description of the implementation details and beneficial effects of a regional boundary control method based on traffic congestion index, which will not be repeated here.

[0135] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0136] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0137] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0138] The modules described as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the objectives of the embodiments of the present invention, depending on actual needs.

[0139] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A regional boundary control method based on traffic congestion index, characterized in that, Includes the following steps: Historical traffic data of the target traffic network is collected according to time parameters to obtain the segment lengths of multiple roads and the traffic congestion index of each road in multiple different time periods. Images of multiple roads and their corresponding traffic congestion indices are constructed according to multiple different time periods to obtain road network topology maps for each time period. The multiple road network topology maps are then clustered using a spectral clustering algorithm to obtain traffic center areas for multiple time periods. Each traffic center area consists of multiple central roads. The cumulative number of vehicles and traffic completion rate of traffic center areas for multiple time periods are obtained respectively. A three-dimensional macro basic graph model is constructed based on the cumulative number of vehicles, the traffic completion rate and the road segment length of multiple central roads in each traffic center area. Traffic data of the target traffic network at the current time is collected to obtain the target traffic congestion index of each road. An image is constructed by multiple roads and their corresponding target traffic congestion indices to obtain the target road network topology map. The target road network topology map is clustered by a spectral clustering algorithm to obtain the target traffic center area. The target traffic center area includes multiple boundary intersections equipped with traffic lights. The sliding mode variable structure control algorithm calculates the number of vehicles in the target traffic center area based on the three-dimensional macroscopic basic graph model to obtain the number of controlled vehicles. Based on the number of controlled vehicles, the change duration of traffic lights at multiple boundary intersections is calculated to obtain the green light adjustment duration corresponding to multiple green lights. The display duration of the green lights at the corresponding boundary intersections is adjusted according to the multiple green light adjustment durations. The method involves clustering multiple road network topology maps using a spectral clustering algorithm to obtain traffic center areas for multiple time periods, including: S21. Sum the edges corresponding to any node in any road network topology to obtain the connectivity value. Repeat this process for all nodes to obtain the connectivity value of each node. S22. The road network topology is segmented according to multiple connectivity values ​​using a spectral clustering calculation expression to obtain initial segmentation lines. These initial segmentation lines are then minimized using an objective function to obtain further segmentation lines. The road network topology is then divided according to these segmentation lines to obtain the traffic center area in the road network topology for any given time period. The spectral clustering calculation expression is: Where Ncut(V) is the initial segmentation line, A k Let k be the set of road nodes for the k-th region category. d is the complement of the set of road nodes for the k-th region category. i Let K be the connectivity value of the i-th road node, and K be the number of region categories. For A k and Congestion similarity between roads; S23. Repeat S21 to S22 to process all road network topology maps and obtain traffic center areas for multiple time periods.

2. The regional boundary control method according to claim 1, characterized in that, The process involves constructing images of multiple roads and their corresponding traffic congestion indices for different time periods to obtain road network topology maps for each time period, including: S11. Calculate the similarity between traffic congestion indices for any given time period using a similarity function to obtain the congestion similarity between multiple roads within that time period. The similarity function is: Among them, w ij Let x be the congestion similarity between the i-th road and the j-th road. i Let x be the traffic congestion index of the i-th road. j Let σ be the traffic congestion index of the j-th road. 2 Let be the variance of the traffic congestion index, and ||·|2 be the L2 norm. S12. Multiple roads are treated as multiple road nodes, and the congestion similarity of the multiple roads is used as the weight value of multiple congestion edges. Based on the weight value, the multiple congestion edges are connected to the corresponding road nodes to construct a road network topology graph. S13. Repeat S11 to S12 to process multiple roads and their corresponding traffic congestion indices for multiple different time periods to obtain the road network topology map for each time period.

3. The regional boundary control method according to claim 1, characterized in that, The construction of a three-dimensional macroscopic basic graph model based on multiple cumulative vehicle counts, multiple traffic completion rates, and multiple central road segment lengths in each traffic center area includes: The lengths of the road segments of multiple central roads in each of the aforementioned traffic center areas are summed to obtain the total length of the central road segments corresponding to each of the aforementioned traffic center areas; A three-dimensional coordinate system is constructed, with the cumulative number of vehicles as the horizontal axis, the total length of the central road segment as the vertical axis, and the traffic completion rate as the vertical axis. In this way, the cumulative number of vehicles, traffic completion rate, and total length of the central road segment corresponding to each traffic center area are converted into three-dimensional coordinates, resulting in multiple control points. These control points are then imported into the three-dimensional coordinate system to complete the construction of the three-dimensional macroscopic basic map model.

4. The regional boundary control method according to claim 3, characterized in that, The calculation of the number of vehicles in the target traffic center area using the sliding mode variable structure control algorithm based on the three-dimensional macroscopic basic graph model to obtain the number of controlled vehicles includes: The traffic core area is selected from the three-dimensional macro basic map model, and the increment of the target traffic center area relative to the traffic core area is calculated to obtain the road length to be controlled and the number of vehicles to be controlled. A sliding surface state function is constructed based on the road length to be controlled, the number of vehicles to be controlled, and the set control gain. The number of vehicles to be controlled is obtained by solving the sliding surface state function using the constant velocity approach law.

5. The regional boundary control method according to claim 4, characterized in that, The process of selecting the traffic core area from the three-dimensional macroscopic basic map model and calculating the increment of the target traffic center area relative to the traffic core area to obtain the road length to be controlled and the number of vehicles to be controlled includes: From the three-dimensional macro basic map model, the traffic center area with the smallest total length of the central road segment is selected as the traffic core area, and the total length of the core road segment of the traffic core area is obtained. The road segment lengths of multiple central roads in the target traffic center area are summed to obtain the total length of the target road segment. The total length of the target road segment is subtracted from the total length of the core road segment to obtain the length of the road to be controlled. Obtain the target cumulative number of vehicles in the target traffic center area at the current time, and subtract the target cumulative number of vehicles from the preset critical cumulative number of vehicles to obtain the number of vehicles to be controlled.

6. The regional boundary control method according to claim 1, characterized in that, The step of calculating the switching duration of traffic lights at multiple boundary intersections based on the number of controlled vehicles to obtain the green light adjustment duration corresponding to multiple green lights includes: S31. Obtain the green light duration corresponding to multiple boundary intersections in the current control cycle and the predicted number of vehicles expected to enter the target traffic center area in the next control cycle. Divide the number of controlled vehicles by the predicted number of vehicles to obtain the control rate. S32. Multiply the control rate by any green light duration to obtain the green light adjustment amount at any boundary intersection. Add the green light duration and the green light adjustment amount to obtain the green light adjustment duration. S33. Repeat S31 to S32 to process the green light duration of all boundary intersections and obtain the green light adjustment duration corresponding to multiple green lights.

7. The regional boundary control method according to claim 6, characterized in that, After the step of adjusting the display duration of the green light at the corresponding boundary intersection according to the multiple green light adjustment durations, the method further includes: The predicted number of vehicles is calculated based on the set prediction parameters for each green light adjustment duration and the number of roads waiting to enter the target traffic center area. The predicted number of controlled vehicles is obtained by subtracting the predicted number of controlled vehicles from the predicted number of controlled vehicles. The remaining number of vehicles is then divided by a preset guidance compliance rate to obtain the guided number of vehicles. Based on the guided number of vehicles, vehicles waiting to enter the target traffic center area in the next control cycle are guided to roads outside the target traffic center area.

8. A regional boundary control system based on a traffic congestion index, characterized in that, include: The data acquisition module is used to collect historical traffic data of the target traffic network according to time parameters, and obtain the segment lengths of multiple roads and the traffic congestion index of each road in multiple different time periods. The clustering module is used to construct images of multiple roads and their corresponding traffic congestion indices according to multiple different time periods, to obtain the road network topology map corresponding to each time period. The multiple road network topology maps are clustered by the spectral clustering algorithm to obtain the traffic center area for multiple time periods. The traffic center area is composed of multiple central roads. The modeling module is used to obtain the cumulative number of vehicles and traffic completion rate corresponding to the traffic center area in multiple time periods, and to construct a three-dimensional macro basic graph model based on the multiple cumulative number of vehicles, the multiple traffic completion rates and the road segment lengths of multiple central roads in each traffic center area. The data acquisition module is also used to collect traffic data of the target traffic network at the current time to obtain the target traffic congestion index for each road. The clustering module is also used to construct images of multiple roads and their corresponding target traffic congestion indices to obtain a target road network topology map. The target road network topology map is then clustered using a spectral clustering algorithm to obtain a target traffic center area, which includes multiple boundary intersections equipped with traffic lights. The control module is used to calculate the number of vehicles in the target traffic center area based on the three-dimensional macroscopic basic graph model using a sliding mode variable structure control algorithm to obtain the number of controlled vehicles, and to calculate the change duration of traffic lights at multiple boundary intersections based on the number of controlled vehicles to obtain the green light adjustment duration corresponding to multiple green lights, and to adjust the display duration of the green lights at the corresponding boundary intersections according to the multiple green light adjustment durations respectively. The method involves clustering multiple road network topology maps using a spectral clustering algorithm to obtain traffic center areas for multiple time periods, including: a. Sum the edges corresponding to any node in any road network topology to obtain the connectivity value. Repeat this process for all nodes to obtain the connectivity value of each node. b. The road network topology is segmented according to multiple connectivity values ​​using a spectral clustering calculation expression to obtain initial segmentation lines. These initial segmentation lines are then minimized using an objective function to obtain further segmentation lines. The road network topology is then divided according to these segmentation lines to obtain the traffic center area in any given time period. The spectral clustering calculation expression is: Where Ncut(V) is the initial segmentation line, A k Let k be the set of road nodes for the k-th region category. d is the complement of the set of road nodes for the k-th region category. i Let K be the connectivity value of the i-th road node, and K be the number of region categories. For A k and Congestion similarity between roads; c. Repeat steps a to b to process all road network topology maps and obtain traffic center areas for multiple time periods.

9. The regional boundary control system according to claim 8, characterized in that, In the clustering module, images of multiple roads and their corresponding traffic congestion indices are constructed according to multiple different time periods to obtain road network topology maps for each time period, including: The similarity between traffic congestion indices for any given time period is calculated using a similarity function, resulting in the congestion similarity between multiple roads within that time period. The similarity function is as follows: Among them, w ij Let x be the congestion similarity between the i-th road and the j-th road. i Let x be the traffic congestion index of the i-th road. j Let σ be the traffic congestion index of the j-th road. 2 Let be the variance of the traffic congestion index, and ||·|2 be the L2 norm. Multiple roads are treated as multiple road nodes, and the congestion similarity of the multiple roads is used as the weight value of multiple congestion edges. Based on the weight value, the multiple congestion edges are connected to the corresponding road nodes to construct a road network topology graph. This process is used to process multiple roads and their corresponding traffic congestion indices at different time periods to obtain the road network topology map for each time period.

Images