Traffic signal control method based on generalization TUC and maximum pressure strategy of cv data

Abstract

The invention discloses a traffic signal control method based on a generalized TUC and a maximum pressure strategy of cv data, and the method comprises the steps: taking a #-shaped road network as a control region, carrying out the traffic signal control periodically, and carrying out the traffic signal control of all intersections in the control region through employing the maximum pressure strategy in a first period; starting from a second period, determining a supersaturated intersection based on video data collected by cv equipment arranged at each intersection and each entrance in the control area in the previous period, and determining an associated intersection of the supersaturated intersection; then, traffic signal control is carried out on all the supersaturated intersections and the associated intersections thereof by adopting a TUC strategy by adopting the associated intersections of all the supersaturated intersections in the current period, and traffic signal control is carried out on other intersections by adopting a maximum pressure strategy; the method has the advantages of being high in calculation speed, high in real-time performance, high in control performance under the supersaturated traffic condition and high in control precision.

Description

Technical Field

[0001] This invention relates to traffic signal control methods, and more particularly to a traffic signal control method based on generalized TUC and maximum pressure strategy using CV data. Background Technology

[0002] With the acceleration of urbanization in my country, urban road density is increasing, the distance between adjacent intersections is decreasing, and the road network is becoming more complex. Against this backdrop, urban road intersections, as core and critical nodes in urban road traffic operation, significantly influence the overall operation of the urban road network through efficient management and organization. How to improve the operational efficiency and capacity of the traffic system, alleviate road network congestion, and reduce the total travel time for pedestrians and vehicles through reasonable traffic signal control has become a crucial issue in current urban traffic management.

[0003] Currently, many roads in my country's urban road networks crisscross, forming a grid-like structure, also known as a "well"-shaped road network. This type of road network layout is regular and suitable for regional coordinated control. Most existing "well"-shaped road networks use a single-intersection timed control method. Its cycle is fixed and relies on historical traffic data, making it difficult to adjust signal duration according to real-time traffic dynamics. Its efficiency has reached its limit and it cannot adapt to the increasingly complex and ever-changing urban traffic environment.

[0004] To address this, the academic community has proposed various traffic signal control methods, such as TUC (Traffic-Responsive Urban Control) and Max-Pressure Control. TUC, developed by Diakaki, Papageorgiou, and others, is an adaptive signal control method for urban areas. By uniformly adjusting the green light time at each intersection within the road network, it minimizes the overall queue length and travel time in the entire area. It exhibits high control performance under oversaturated traffic conditions, but its calculation speed is slow and its real-time control performance is poor. Max-Pressure Control, developed by Varaiya, is a traffic signal control method whose core idea is to always prioritize the green light for the direction with the greatest pressure. It uses simple rules to achieve adaptive optimization control of complex road networks. It has a fast calculation speed and strong real-time control performance, but its control performance is limited and its control accuracy is not high under oversaturated traffic conditions. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a traffic signal control method based on CV data generalized TUC and maximum pressure strategy, which has fast calculation speed, strong real-time performance, high control performance under oversaturated traffic conditions, and high control accuracy.

[0006] The technical solution adopted by this invention to solve the above-mentioned technical problems is as follows: a traffic signal control method based on generalized TUC and maximum pressure strategy using CV data. Using a preset "grid"-shaped road network as the control area, traffic signal control is periodically executed. Specifically, in the first cycle, the maximum pressure strategy is used for traffic signal control at all intersections within the control area. Starting from the second cycle, based on the video data collected in the previous cycle by CV (video camera) devices installed at each intersection entrance within the control area, it is determined whether each intersection is in an oversaturated state in the current cycle. Then, for each intersection in the current cycle that is in an oversaturated state (i.e., an oversaturated intersection), its associated intersections are determined based on the road network topology and preset filtering rules. Next, the associated intersections of all oversaturated intersections in the current cycle constitute a set of intersections that require TUC coordinated control in the current cycle. In the current cycle, the TUC strategy is used for traffic signal control at all intersections within the set, while the maximum pressure strategy is maintained for the remaining intersections within the control area.

[0007] Compared with existing technologies, the advantages of this invention lie in its phased and regional hybrid control strategy. In the first cycle, a maximum pressure strategy with lower computational load is uniformly applied to the entire control area to achieve rapid response. Starting from the second cycle, oversaturated intersections are identified based on video data collected by CV devices in the previous cycle, and their associated intersections are determined according to road network topology and screening rules. This precisely limits the computationally complex TUC coordinated control to a local set consisting of oversaturated intersections and their associated intersections. This significantly reduces the number of intersections requiring TUC calculation in each cycle, significantly improving calculation speed and real-time performance. Furthermore, the coordinated control of the TUC strategy within a local area effectively suppresses queue overflow and congestion spread. Meanwhile, the remaining intersections within the control area continue to use the maximum pressure strategy, forming a hierarchical control architecture of "local coordination and global speed". This allows the method of this invention to achieve both high control performance and high control accuracy under oversaturated traffic conditions.

[0008] Furthermore, the specific method for determining whether each intersection is in an oversaturated state in the current cycle is as follows: first, determine the various types of turning flows at each entrance of each intersection based on the video data, and then calculate the capacity of each turn at each entrance of each intersection. If the turning flow at any turn at a certain entrance of an intersection is greater than its capacity, then the intersection is determined to be in an oversaturated state in the current cycle.

[0009] Furthermore, the specific process for determining the various turning flows at a specific intersection and a specific entrance based on the video data is as follows: The number of left-turning vehicles, straight-ahead vehicles, and right-turning vehicles in the video data are counted. The number of left-turning vehicles is converted to the hourly unit to obtain the left-turn flow rate for that intersection and that entrance in the current cycle, where the unit of left-turn flow rate is vehicles per hour. The number of right-turning vehicles is converted to the hourly unit to obtain the right-turn flow rate for that intersection and that entrance in the current cycle, where the unit of right-turn flow rate is vehicles per hour. The number of straight-ahead vehicles is converted to the hourly unit to obtain the straight-ahead flow rate for that intersection and that entrance in the current cycle, where the unit of straight-ahead flow rate is vehicles per hour.

[0010] Furthermore, the traffic capacity of a certain turn at a certain intersection is determined based on the number of lanes for that turn, the lane saturation flow, the effective green light time of the intersection, and the signal cycle.

[0011] Furthermore, the specific process for calculating the traffic capacity of a certain turn at a certain entrance of an intersection is as follows: Step A1: Let the intersection be i. Define left turns at this intersection's entrance as Turn 1, straight turns as Turn 2, and right turns as Turn 3. Count the total number of lanes for turning j at this intersection's entrance i, and denote it as Q. ij The unit is a strip, j=1, 2, 3; and it is arranged according to 1 to Q. ij Let q be the lane number for turning j at the intersection i-th approach. Let lane number q be the q-th lane for turning j at the intersection i-th approach, and let θ be the width of this lane. ijq The unit is meters, and the saturation flow rate is denoted as S. ijq The unit is vehicles per hour, if θ ijq If S is less than or equal to 3.5m, then ijq =900 vehicles per hour, if θ ijq If it is greater than 3.5m, then S ijq =1200 vehicles per hour; Step A2: Calculate the turning capacity C of the intersection i at the turning point j using formula (1). ij The unit is vehicles per hour: (1) Among them, L i This is the signal period at intersection i, in seconds; g ij It is the effective green light time for turning at intersection i and entrance j, in seconds.

[0012] Furthermore, for each intersection that is in an oversaturated state in the current cycle, the specific method for determining its associated intersections based on the road network topology and preset filtering rules is as follows: First, construct the adjacency matrix of the oversaturated intersection in the current cycle based on the road network topology. Then, combine the adjacency matrix with preset path length thresholds and association strength thresholds to determine the associated intersections of the oversaturated intersection in the current cycle.

[0013] Furthermore, the specific method for constructing the adjacency matrix of a certain oversaturated intersection in the current period based on the road network topology is as follows: Step B1: Count the number and location of all intersections in the control area, and node all intersections in the control area, that is, each intersection is represented by a node, and each two adjacent nodes are connected by a bidirectional arrow consisting of two unidirectional arrows in opposite directions. The bidirectional arrow represents the bidirectional road segment connecting the two adjacent intersections, thus obtaining the original road network of the control area. Step B2: Set the rules for valid road segments in two adjacent intersections. Specifically, refer to the two intersections as the first intersection and the second intersection. If the topological distance from the first intersection to the oversaturated intersection is greater than the topological distance from the second intersection to the oversaturated intersection, then the road segment from the first intersection to the second intersection is a valid road segment; otherwise, it is an invalid road segment. Step B3: In the original road network, retain all one-way arrows for valid road segments and delete all one-way arrows for invalid road segments to obtain a valid road network; Step B4: Denote the total number of all intersections within the control area as Y, and randomly number these Y intersections from 1 to Y. The intersection numbered y is called the y-th intersection, where y = 1, 2, ..., Y; Step B5: Construct the adjacency matrix M of the oversaturated intersection in the current period based on the effective road network. The expression for M is shown in formula (2): (2) In formula (2), m tr Let m be the element in the t-th row and r-th column of the adjacency matrix M, where t = 1, 2, ..., Y; r = 1, 2, ..., Y; m tr This represents the positional relationship between the t-th intersection and the r-th intersection; if the t-th intersection is an adjacent intersection to the r-th intersection and the one-way arrow points from the t-th intersection to the r-th intersection, then m tr =1; otherwise, m tr =0.

[0014] Furthermore, the specific process of determining the associated intersections of a certain oversaturated intersection based on preset path length thresholds and association strength thresholds is as follows: Step C1: Denote the path length threshold as u and the association strength threshold as v, where u is an integer greater than or equal to 1 and less than or equal to 10, and v is greater than or equal to 0 and less than or equal to 1. Step C2: Construct u intermediate matrices, and denote the k-th intermediate matrix as Mk, k=1,2,…,u. The expression of the k-th intermediate matrix Mk is shown in formula (3): (3) In formula (3), mk tr Let be the element in the t-th row and r-th column of the k-th intermediate matrix Mk; Step C3: Construct the correlation matrix P. The expression for the correlation matrix P is shown in formula (4): (4) In formula (4), p tr Let p be the element in the t-th row and r-th column of the correlation matrix P. tr The result is obtained using formula (5): (5) In formula (5), max(Mk) represents the maximum value of all elements in the k-th intermediate matrix Mk; Step C4: Denote the number of the oversaturated intersection as W, where W∈[1,Y]; determine the W-th column of the correlation matrix P, excluding p. WW If any of the other Y-1 elements is greater than the association strength threshold v, then the intersection with the number equal to the row number of that element is the associated intersection of the Wth intersection; otherwise, it is not the associated intersection of the Wth intersection. Attached Figure Description

[0015] Figure 1 The flowchart shows the traffic signal control method based on CV data generalized TUC and maximum pressure strategy of the present invention. Figure 2 This is a control area road network diagram used in the SUMO simulation of the traffic signal control method based on CV data and the maximum pressure strategy in Example 3. Figure 3 This is a diagram of nine intersections within the control area in the traffic signal control method based on CV data generalized TUC and maximum pressure strategy in Example 3; Figure 4 The original road network map containing bidirectional road segments and intersection nodes of adjacent intersections is shown in the traffic signal control method based on CV data generalized TUC and maximum pressure strategy in Example 3. Figure 5This is an effective road network map containing only valid road segments and intersection nodes in the traffic signal control method based on CV data generalized TUC and maximum pressure strategy in Example 3. Detailed Implementation

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0017] Example 1: As Figure 1 As shown, a traffic signal control method based on generalized TUC and maximum pressure strategy using CV data is presented. Using a pre-defined "grid" road network as the control area, traffic signal control is periodically executed. Specifically: In the first cycle, the maximum pressure strategy is used for traffic signal control at all intersections within the control area. Starting from the second cycle, based on the video data collected in the previous cycle by CV devices installed at each intersection entrance within the control area, it is determined whether each intersection is in an oversaturated state in the current cycle. Then, for each intersection in the current cycle that is in an oversaturated state (i.e., an oversaturated intersection), its associated intersections are determined based on the road network topology and pre-defined filtering rules. Next, the associated intersections of all oversaturated intersections in the current cycle constitute the set of intersections requiring TUC coordinated control in the current cycle. If an intersection is an associated intersection of multiple oversaturated intersections, it only needs to appear once in the intersection set. In the current cycle, the TUC strategy is used for traffic signal control at all intersections within the intersection set, while the maximum pressure strategy is maintained for traffic signal control at the remaining intersections within the control area.

[0018] The traffic signal control method based on CV data and the maximum pressure strategy in this embodiment has the following advantages: I. Significantly improved computational efficiency and real-time response capability of the control algorithm. In the first control cycle, a computationally less computationally intensive maximum pressure strategy is uniformly applied to all intersections within the "grid"-shaped control area, ensuring rapid control initiation. Starting from the second cycle, oversaturated intersections are dynamically identified based on video data collected by CV devices from the previous cycle. TUC (Transmission Control Override) calculations and implementation are triggered only for the set consisting of oversaturated intersections and their associated intersections determined based on road network topology and filtering rules. This method, through periodic filtering, strictly limits complex TUC calculations to a necessary local scope, significantly reducing the number of intersections requiring TUC calculations in each cycle, thereby effectively reducing the overall computational load and improving the computational speed and real-time response capability of the signal control method.

[0019] Second, it effectively improves control performance and accuracy under oversaturated traffic conditions. Oversaturated intersections are not treated in isolation, but rather their associated intersections are incorporated into a coordinated control set based on the road network topology. Within this set, a Transmission Control (TUC) strategy is employed for signal control. This directional coordination mechanism enables traffic flow coordination and proactive suppression of queue overflow in oversaturated areas. Simultaneously, other intersections within the control area maintain a maximum pressure strategy, forming a hybrid control architecture combining "local TUC coordination" and "global maximum pressure rapid response." This hybrid control architecture ensures precise handling of local congestion under oversaturated conditions, preventing congestion spread and thus achieving higher control performance and accuracy under oversaturated traffic conditions.

[0020] Example 2: This example is basically the same as Example 1, except that: In this example, the specific way to determine whether each intersection is in an oversaturated state in the current cycle is as follows: First, determine the various types of turning flows at each entrance of each intersection based on video data, and then calculate the capacity of each turn at each entrance of each intersection. If the turning flow of any turn at a certain entrance of an intersection is greater than its capacity, it is determined that the intersection is in an oversaturated state in the current cycle.

[0021] In this embodiment, the specific process of determining the various types of turning traffic flow at a certain intersection and a certain entrance based on video data is as follows: The number of left-turning vehicles, the number of straight-going vehicles, and the number of right-turning vehicles in the video data are counted. The number of left-turning vehicles is converted into the left-turn traffic flow for the current period at the intersection and the entrance, with the unit being vehicles per hour. The number of right-turning vehicles is converted into the right-turn traffic flow for the current period at the intersection and the entrance, with the unit being vehicles per hour. The number of straight-going vehicles is converted into the straight-going traffic flow for the current period at the intersection and the entrance, with the unit being vehicles per hour.

[0022] In this embodiment, the traffic capacity of a certain turn at a certain intersection is determined based on the number of lanes for that turn, the lane saturation flow, the effective green light time of the intersection, and the signal cycle. The specific process for calculating the traffic capacity of a certain turn at a certain intersection is as follows: Step A1: Let the intersection be i. Define left turns at this intersection's entrance as Turn 1, straight turns as Turn 2, and right turns as Turn 3. Count the total number of lanes for turning j at this intersection's entrance i, and denote it as Q. ij The unit is a strip, j=1, 2, 3; and it is arranged according to 1 to Q. ij Let q be the lane number for turning j at the intersection i-th approach. Let lane number q be the q-th lane for turning j at the intersection i-th approach, and let θ be the width of this lane. ijq The unit is meters, and the saturation flow rate is denoted as S. ijq The unit is vehicles per hour, if θijq If S is less than or equal to 3.5m, then ijq =900 vehicles per hour, if θ ijq If it is greater than 3.5m, then S ijq =1200 vehicles per hour; Step A2: Calculate the turning capacity C of the intersection i at the turning point j using formula (1). ij The unit is vehicles per hour: (1) Among them, L i This is the signal period at intersection i, in seconds; g ij It is the effective green light time for turning at intersection i and entrance j, in seconds.

[0023] Example 3: This example is basically the same as Example 2, except that: In this example, for each intersection that is in an oversaturated state in the current cycle, the specific method for determining its associated intersections based on the road network topology and preset filtering rules is as follows: First, construct the adjacency matrix of the oversaturated intersection in the current cycle based on the road network topology, and then combine the adjacency matrix to determine the associated intersections of the oversaturated intersection in the current cycle according to the preset path length threshold and association strength threshold.

[0024] In this embodiment, the specific method for constructing the adjacency matrix of a certain oversaturated intersection in the current period based on the road network topology is as follows: Step B1: Count the number and location of all intersections in the control area, and nod all intersections in the control area. That is, each intersection is represented by a node, and each two adjacent nodes are connected by a bidirectional arrow consisting of two unidirectional arrows in opposite directions. The bidirectional arrow represents the bidirectional road segment connecting the two adjacent intersections, thus obtaining the original road network of the control area. Step B2: Set the rules for valid road segments in two adjacent intersections. Specifically, refer to the two intersections as the first intersection and the second intersection. If the topological distance from the first intersection to the oversaturated intersection is greater than the topological distance from the second intersection to the oversaturated intersection, then the road segment from the first intersection to the second intersection is a valid road segment; otherwise, it is an invalid road segment. Step B3: In the original road network, retain all one-way arrows for valid road segments and delete all one-way arrows for invalid road segments to obtain the valid road network; Step B4: Denote the total number of all intersections within the control area as Y, and randomly number these Y intersections from 1 to Y. The intersection numbered y is called the y-th intersection, where y = 1, 2, ..., Y; Step B5: Construct the adjacency matrix M of the oversaturated intersection in the current period based on the effective road network. The expression for M is shown in formula (2): (2) In formula (2), m tr Let m be the element in the t-th row and r-th column of the adjacency matrix M, where t = 1, 2, ..., Y; r = 1, 2, ..., Y; m tr This represents the positional relationship between the t-th intersection and the r-th intersection; if the t-th intersection is an adjacent intersection to the r-th intersection and the one-way arrow points from the t-th intersection to the r-th intersection, then m tr =1; otherwise, m tr =0.

[0025] In this embodiment, the specific process of determining the associated intersections of an oversaturated intersection based on a preset path length threshold and an association strength threshold is as follows: Step C1: Denote the path length threshold as u and the association strength threshold as v, where u is an integer greater than or equal to 1 and less than or equal to 10, and v is greater than or equal to 0 and less than or equal to 1. Step C2: Construct u intermediate matrices, and denote the k-th intermediate matrix as Mk, k=1,2,…,u. The expression of the k-th intermediate matrix Mk is shown in formula (3): (3) In formula (3), mk tr Let be the element in the t-th row and r-th column of the k-th intermediate matrix Mk; Step C3: Construct the correlation matrix P. The expression for the correlation matrix P is shown in formula (4): (4) In formula (4), p tr Let p be the element in the t-th row and r-th column of the correlation matrix P. tr The result is obtained using formula (5): (5) In formula (5), max(Mk) represents the maximum value of all elements in the k-th intermediate matrix Mk; Step C4: Denote the number of the oversaturated intersection as W, where W∈[1,Y]; determine the W-th column of the correlation matrix P, excluding p... WW If any of the other Y-1 elements is greater than the association strength threshold v, then the intersection with the number equal to the row number of that element is the associated intersection of the Wth intersection; otherwise, it is not the associated intersection of the Wth intersection.

[0026] To verify the performance of the traffic signal control method based on CV data and the maximum pressure strategy of the present invention, the traffic signal control method based on CV data and the maximum pressure strategy of the present invention was implemented in ArcGIS software.

[0027] The SUMO traffic simulation platform was used to construct the simulation environment, and a system was built for example... Figure 2 The illustrated "grid"-shaped road network with nine intersections is the control area. Traffic signal control is implemented using the generalized TUC based on CV data and the maximum pressure strategy of this invention. A cycle of 30 minutes (0.5 hours) is set, with a path length threshold of 2 and an association strength threshold of 0.2. In practical applications, the cycle, path length threshold, and association strength threshold are set according to actual control requirements. Under the same simulation environment, TUC control and maximum pressure control are applied to all intersections in the control area for comparison.

[0028] When performing traffic signal control using the generalized TUC and maximum pressure strategy based on CV data of the present invention, in the first cycle, the method performs the following: Figure 2 All nine intersections within the area shown employ a maximum pressure strategy for traffic signal control.

[0029] Starting from the second cycle, with Figure 3 The intersection numbered 9 is shown as an example. Intersection 9 has 4 entrances, each with only one straight-through and one-turn lane. The width of each lane is greater than 3.5 meters, so the saturation flow rate of each lane is 1200 vehicles per hour. The signal cycle of this intersection is 80 seconds. The effective green light time for straight-through traffic at the south and north entrances is 30 seconds, and the effective green light time for straight-through traffic at the east and west entrances is 40 seconds.

[0030] In a certain period, video data of each entrance to intersection No. 9 in the previous period is obtained and the traffic flow of each turn is counted. According to the video data, the number of vehicles going straight at the east entrance in the previous period is 350, the number of vehicles going straight at the west entrance is 400, the number of vehicles going straight at the south entrance is 250, and the number of vehicles going straight at the north entrance is 350. Since a cycle is preset to 0.5 hours, the number of vehicles in the above single cycle is converted into the flow rate per hour, resulting in a straight-through flow rate of 700 vehicles per hour at the east entrance, 800 vehicles per hour at the west entrance, 500 vehicles per hour at the south entrance, and 700 vehicles per hour at the north entrance. Using formula (1), the straight-through capacity of each entrance of intersection 9 is calculated as follows: East entrance straight-through capacity = 600 vehicles per hour; West entrance straight-through capacity = 600 vehicles per hour; South entrance straight-through capacity = 450 vehicles per hour; North entrance straight-through capacity = 450 vehicles per hour. By comparing the straight-through capacity and flow rate of each entrance of intersection 9, it is determined that the straight-through capacity of each entrance is less than the flow rate of each entrance. Therefore, it is judged that intersection 9 is in an oversaturated state in the current cycle. Figure 2 The original road network of the controlled area is obtained as follows: Figure 4 As shown, the effective road network is as follows: Figure 5 As shown. The adjacency matrix M of the 9th intersection in the current cycle is calculated using formula (2): ; Construct two intermediate matrices, the first intermediate matrix M1=M, and calculate the second intermediate matrix M2 using formula (3): ; The correlation matrix P is calculated using formulas (4) and (5): ; In the 9th column of the correlation matrix P, P 59 P 69 P 89 The values ​​are all greater than the association strength threshold of 0.2. Therefore, intersections No. 5, No. 6, and No. 8 are associated intersections of intersection No. 9, while the other intersections are not associated intersections of intersection No. 9.

[0031] After all the associated intersections of the oversaturated intersections are determined, the associated intersections of all the oversaturated intersections in the current cycle are used to form the set of intersections that need to be coordinated and controlled by TUC in the current cycle. In the current cycle, the TUC strategy is used for traffic signal control of all intersections in the set of intersections, while the remaining intersections continue to use the maximum pressure strategy for traffic signal control.

[0032] During the above simulation, the simulation lasted for 3600 seconds, and the timing was recorded throughout the process. The average calculation time of the present invention, the TUC control method (all intersections use TUC control), and the maximum pressure control method (all intersections use maximum pressure control) in a single cycle were statistically obtained, as well as the maximum queue length of the roads in the control area during the entire control process.

[0033] The comparison data of average calculation time for a single cycle is shown in Table 1: Table 1: Comparison of Average Calculation Time for a Single Period

[0034] As shown in Table 1, the average calculation time per cycle of the method of the present invention is significantly shorter than that of the TUC control method, and is on the same order of magnitude as that of the maximum pressure control method, which verifies its advantages of fast calculation speed and strong real-time performance.

[0035] The present invention, the TUC control method (all intersections use TUC control), and the maximum pressure control method (all intersections use maximum pressure control) are compared in Table 2. The maximum queue length data for roads within the control area during the entire control process are shown in Table 2. Table 2: Comparison of Maximum Queue Lengths

[0036] As shown in Table 2, the maximum queue length of the method of the present invention is significantly smaller than that of the maximum pressure control method, and is on the same order of magnitude as that of the TUC control method, which verifies its advantages of high control performance and high control accuracy under oversaturated traffic conditions.

[0037] In summary, the traffic signal control method based on generalized TUC and maximum pressure strategy using CV data of the present invention can easily acquire relevant traffic data by utilizing the CV devices that are already densely distributed at intersections of various road segments in the road network. Through coordinated control of the TUC strategy and the maximum pressure strategy, it has the characteristics of fast calculation speed, strong real-time performance, high control performance under oversaturated traffic conditions, and high control accuracy, which can realize real-time high-precision control of the current traffic signal and improve traffic efficiency.

Claims

1. A traffic signal control method based on generalized TUC and maximum pressure strategy using CV data, characterized in that, Using a pre-defined grid-shaped road network as the control area, traffic signal control is periodically implemented. Specifically: In the first cycle, the maximum pressure strategy is used for traffic signal control at all intersections within the control area; starting from the second cycle, based on the video data collected in the previous cycle by CV devices installed at each entrance of each intersection within the control area, it is determined whether each intersection is in an oversaturated state in the current cycle; then, for each intersection in the current cycle that is in an oversaturated state, i.e., an oversaturated intersection, its associated intersections are determined based on the road network topology and pre-defined filtering rules; next, the associated intersections of all oversaturated intersections in the current cycle are used to form a set of intersections that require TUC coordinated control in the current cycle; in the current cycle, the TUC strategy is used for traffic signal control at all intersections within the set of intersections, while the maximum pressure strategy is maintained for traffic signal control at the remaining intersections within the control area.

2. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data as described in claim 1, characterized in that, The specific method for determining whether an intersection is in an oversaturated state in the current cycle is as follows: First, determine the various types of turning flows at each entrance of each intersection based on the video data, and then calculate the capacity of each turn at each entrance of each intersection. If the turning flow of any turn at a certain entrance of an intersection is greater than its capacity, it is determined that the intersection is in an oversaturated state in the current cycle.

3. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data according to claim 2, characterized in that, The specific process for determining the various turning flows at a specific intersection and a specific entrance based on the video data is as follows: The number of left-turning vehicles, straight-ahead vehicles, and right-turning vehicles in the video data are counted. The number of left-turning vehicles is converted to the hourly unit to obtain the left-turn flow rate for that intersection and that entrance in the current cycle, where the unit is vehicles per hour. The number of right-turning vehicles is converted to the hourly unit to obtain the right-turn flow rate for that intersection and that entrance in the current cycle, where the unit is vehicles per hour. The number of straight-ahead vehicles is converted to the hourly unit to obtain the straight-ahead flow rate for that intersection and that entrance in the current cycle, where the unit is vehicles per hour.

4. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data according to claim 3, characterized in that, The traffic capacity of a certain turn at a certain entrance of an intersection is determined based on the number of lanes for that turn, the lane saturation flow, the effective green light time of the intersection, and the signal cycle.

5. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data according to claim 4, characterized in that, The specific process for calculating the traffic capacity of a certain turn at a certain entrance of an intersection is as follows: Step A1: Let the intersection be i, and define the left turn at this intersection as Turn 1, the straight turn as Turn 2, and the right turn as Turn 3. Count the total number of turning lanes j at the intersection i-th approach, denoted as Q. ij The unit is strips, j=1, 2, 3; And in accordance with 1 to Q ij Let q be the lane number for turning j at the intersection i-th approach. Let lane number q be the q-th lane for turning j at the intersection i-th approach, and let θ be the width of this lane. ijq The unit is meters, and the saturation flow rate is denoted as S. ijq The unit is vehicles per hour, if θ ijq If S is less than or equal to 3.5m, then ijq =900 vehicles per hour, if θ ijq If it is greater than 3.5m, then S ijq =1200 vehicles per hour; Step A2: Calculate the turning capacity C of the intersection i at the turning point j using formula (1). ij The unit is vehicles per hour: （1） Among them, L i This is the signal period at intersection i, in seconds; g ij It is the effective green light time for turning at intersection i and entrance j, in seconds.

6. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data according to claim 1, characterized in that, For each intersection that is in an oversaturated state in the current cycle, the specific method for determining its associated intersections based on the road network topology and preset filtering rules is as follows: First, construct the adjacency matrix of the oversaturated intersection in the current cycle based on the road network topology. Then, combine the adjacency matrix with preset path length thresholds and association strength thresholds to determine the associated intersections of the oversaturated intersection in the current cycle.

7. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data as described in claim 6, characterized in that, The specific method for constructing the adjacency matrix of an oversaturated intersection in the current period based on the road network topology is as follows: Step B1: Count the number and location of all intersections in the control area, and node all intersections in the control area, that is, each intersection is represented by a node, and each two adjacent nodes are connected by a bidirectional arrow consisting of two unidirectional arrows in opposite directions. The bidirectional arrow represents the bidirectional road segment connecting the two adjacent intersections, thus obtaining the original road network of the control area. Step B2: Set the rules for valid road segments in two adjacent intersections. Specifically, refer to the two intersections as the first intersection and the second intersection. If the topological distance from the first intersection to the oversaturated intersection is greater than the topological distance from the second intersection to the oversaturated intersection, then the road segment from the first intersection to the second intersection is a valid road segment; otherwise, it is an invalid road segment. Step B3: In the original road network, retain all one-way arrows for valid road segments and delete all one-way arrows for invalid road segments to obtain a valid road network; Step B4: Denote the total number of all intersections within the control area as Y, and randomly number these Y intersections from 1 to Y. The intersection numbered y is called the y-th intersection, where y = 1, 2, ..., Y; Step B5: Construct the adjacency matrix M of the oversaturated intersection in the current period based on the effective road network. The expression for M is shown in formula (2): （2） In formula (2), m tr Let Y be the element in the t-th row and r-th column of the adjacency matrix M, where t = 1, 2, ..., Y; r = 1, 2, ..., Y. m tr This represents the positional relationship between the t-th intersection and the r-th intersection; if the t-th intersection is an adjacent intersection to the r-th intersection and the one-way arrow points from the t-th intersection to the r-th intersection, then m tr =1; otherwise, m tr =0.

8. The traffic signal control method based on generalized TUC and maximum pressure strategy using CV data according to claim 7, characterized in that, The specific process for determining the associated intersections of an oversaturated intersection based on preset path length thresholds and association strength thresholds is as follows: Step C1: Denote the path length threshold as u and the association strength threshold as v, where u is an integer greater than or equal to 1 and less than or equal to 10, and v is greater than or equal to 0 and less than or equal to 1. Step C2: Construct u intermediate matrices, and denote the k-th intermediate matrix as Mk, k=1,2,…,u. The expression of the k-th intermediate matrix Mk is shown in formula (3): (3) In formula (3), mk tr Let be the element in the t-th row and r-th column of the k-th intermediate matrix Mk; Step C3: Construct the correlation matrix P. The expression for the correlation matrix P is shown in formula (4): （4） In formula (4), p tr Let p be the element in the t-th row and r-th column of the correlation matrix P. tr The result is obtained using formula (5): （5） In formula (5), max(Mk) represents the maximum value of all elements in the k-th intermediate matrix Mk; Step C4: Denote the number of the oversaturated intersection as W, where W∈[1,Y]; Determine the W-th column of the correlation matrix P, excluding p. WW If any of the other Y-1 elements is greater than the association strength threshold v, then the intersection with the number equal to the row number of that element is the associated intersection of the Wth intersection; otherwise, it is not the associated intersection of the Wth intersection.

Images