Control method and regulation and control agent for parallel arterial road traffic volume coordinated regulation

Abstract

The invention relates to a control method and a regulation and control agent for coordinated regulation of traffic volume of parallel arterial roads, which are based on a relative saturation difference as a trigger condition of flow regulation between the parallel arterial roads. And synchronously determining a first signal control strategy for increasing the passing delay of the vehicle on the high-saturation trunk road and a first signal control strategy for increasing the passing delay of the vehicle on the high-saturation trunk road based on the target transfer flow, so as to realize balanced adjustment between the two parallel trunk roads. When the saturation degree of the parallel main road reaches the steady flow upper limit, vehicles not entering the congested main road can enter the parallel road with high admission ability and low traffic pressure, so that the traffic flow is uniformly distributed, and the utilization rate of the parallel road is improved while the congestion of the main road is relieved. According to the method, the path selection of the vehicle can be actively adjusted before the congestion is formed or after the congestion occurs, the bottleneck suppression is realized, and the road network traffic efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent traffic control, in particular to a control method and a regulation and control intelligent agent for parallel arterial traffic volume coordination, which can realize dynamic balance of road network traffic flow and active congestion prevention and control by intelligently adjusting parallel arterial green wave and red wave bandwidth. BACKGROUND

[0002] With the acceleration of urbanization, the pressure of urban road traffic is increasing. In the urban road network, there are often two or more parallel main roads connecting the same area and having similar functions. Due to the differences in road design standards, traffic capacity, driver preferences and real-time events, traffic flow is often unevenly distributed on these parallel main roads. The main road with higher traffic capacity tends to attract excessive traffic flow, leading to traffic saturation or even oversaturation at peak hours or key sections, forming a bottleneck congestion; while the other parallel main road may be in a low saturation state, and the road resources are not fully utilized. This uneven phenomenon not only reduces the traffic efficiency of the congested road, causing the risk of queue spread and regional traffic paralysis, but also causes waste of overall road network resources.

[0003] To deal with the problem of arterial traffic, the existing technology mainly focuses on the optimization control of a single road, which has the following forms:

[0004] 1) Arterial green wave coordination control: by optimizing the signal phase difference between consecutive intersections, vehicles can pass through multiple intersections in the form of "green wave", aiming to reduce the number of stops and travel delays. However, the traditional green wave control aims to maximize the green wave bandwidth of the designed arterial road, which essentially strengthens the attractiveness of the arterial road, which may exacerbate the flow loss of parallel roads and further worsen the imbalance of road network load. It lacks a mechanism to actively guide traffic to parallel roads. Moreover, green wave control is designed for a single main road, and the green wave band optimization is usually to maximize the green wave bandwidth of the designed road, ignoring the traffic volume distribution balance and coordination function of parallel roads. When one main road is overburdened, it is difficult to effectively guide vehicles to the parallel main road, and it lacks overall coordination of road network signal control.

[0005] 2) Red wave coordination control: As a reverse control method, it intentionally sets a phase difference to make vehicles encounter red lights successively, increasing the travel time of a specific path to suppress the traffic flow in that direction. However, it has the following defects: first, it lacks systematicity and may simply shift congestion points rather than eliminating them; second, it fails to coordinate with green wave control on parallel roads, making it impossible to achieve traffic redistribution by combining "suppression" and "diversion". That is, when the main road traffic is overloaded, this mechanism cannot guide part of the traffic to parallel paths. Although red wave control can limit local traffic flow, it is easy to cause local transfer of congestion without systematic coordination control.

[0006] 3) Adaptive regional control system (such as SCOOT, SCATS): Although it can dynamically adjust regional signal timing based on real-time detection data, its optimization goal is usually to minimize regional total delay or maximize traffic capacity, rather than load balancing between parallel roads. Its control model is complex and it is difficult to directly achieve cross-path coordination with the explicit goal of traffic migration.

[0007] 4) Independent path induction system (such as variable message boards, navigation APPs): It indirectly influences driver route selection by publishing traffic information. However, it is disconnected from signal control systems, and the induction recommendations may not match the actual traffic conditions provided by the signals, resulting in limited and lagging induction effects.

[0008] In summary, the existing technologies have the following obvious defects: first, the control target is isolated, i.e., focusing on single-path optimization without considering parallel roads as an organic whole that needs to be coordinated; second, there is a lack of active balancing mechanism, i.e., it cannot actively and quantitatively guide traffic migration between parallel roads through signal control before congestion occurs; third, the control means and induction target are disconnected, i.e., green wave / red wave as a control means is not directly and dynamically related to the induction target of "influencing route selection"; fourth, there is a lack of adaptability and robustness, i.e., most systems rely on fixed thresholds and empirical parameters, making it difficult to adapt to different road network characteristics and dynamic traffic changes.

[0009] That is, the core technical problem of existing technologies is: in a road network subsystem composed of two or more parallel roads, how to dynamically perceive and predict the traffic load state of each road through intelligent traffic signal coordination control, adjust the signal timing parameters of related intersections in real time to change the route selection behavior of vehicles between parallel paths, and achieve dynamic balanced distribution of traffic between parallel roads, avoid or alleviate local congestion caused by traffic concentration, and improve the operation efficiency and robustness of the overall road network.

[0010] Therefore, there is an urgent need for a new control method that can intelligently perceive the state of parallel arterials, dynamically coordinate the bandwidth of green wave and red wave, and thus actively and accurately guide the distribution of traffic flow, so as to achieve load balancing and maximum traffic efficiency of the road network. SUMMARY

[0011] Therefore, there is an urgent need for a new control method that can intelligently perceive the state of parallel arterials, dynamically coordinate the bandwidth of green wave and red wave, and thus actively and accurately guide the distribution of traffic flow, so as to achieve load balancing and maximum traffic efficiency of the road network.

[0012] That is, in view of the defects of the existing signal adjustment mechanism and system, a parallel arterial flow induction control method based on arterial bottleneck intersection or section constraint is proposed. When the saturation of one of the parallel arterials reaches the upper limit of stable flow, the signal timing of the decision intersection upstream of the shunt arterial is adjusted: the bandwidth of the vehicles entering the congested arterial is reduced until the red wave is intercepted; the green wave bandwidth of the vehicles entering the parallel road is increased; and the vehicles not entering the congested arterial are guided to enter the parallel road with high acceptance capacity and low traffic pressure through variable signs or navigation, so as to achieve uniform distribution of traffic flow, alleviate arterial congestion, and improve the utilization rate of parallel roads. This method can actively adjust the route selection of vehicles before or after congestion occurs, and achieve bottleneck suppression and improve road network traffic efficiency.

[0013] Specifically, the first aspect of the present application discloses a control method for parallel arterial traffic volume coordination, which comprises the following steps: S1, acquiring the traffic state data of two parallel arterials connecting the same origin-destination region in real time, calculating the saturation and relative saturation difference of the parallel arterials based on the obtained data, and determining the arterial with higher saturation as the high-saturation arterial and the other arterial as the low-saturation arterial; S2, judging whether to start traffic flow adjustment based on the arterial saturation and relative saturation difference obtained in step S1; if yes, performing S3; S3, determining the target transfer flow that needs to be transferred from the high-saturation arterial to the low-saturation arterial, determining a first signal control strategy that increases the travel delay of vehicles on the high-saturation arterial based on the target transfer flow; at the same time, determining a second signal control strategy that reduces the travel delay of vehicles on the low-saturation arterial in real time based on the target transfer flow; wherein the first signal control strategy and the second signal control strategy act on the decision intersection of the starting region of the two parallel arterials or the coordinated intersection downstream thereof.

[0014] The parallel arterials refer to two parallel arterials with the same function connecting the same two areas, each of which has more intersections (referred to as coordinated intersections), and there is no connecting road between the intersections. If there is a connection, the number of paths that need to be decided increases, and the complexity of the agent increases. The present application takes two arterials without connecting roads as an example for illustration. Vehicles select paths at the decision intersection (the connecting intersection of the two parallel arterials in the starting area), and the real-time traffic of the road with large traffic capacity is also large, and vice versa.

[0015] In the prior art, the green wave control pursues the "smoothness" of the path itself, and the red wave control pursues the "limitation" of the congestion point. Both are fragmented, and neither considers the influence of their actions on the parallel competing paths. In the present application, the parallel arterials are regarded as a whole system, and through real-time monitoring of the load (saturation) state difference of the two paths, the two opposite control methods of "green wave" (reducing impedance, attracting traffic) and "red wave" (increasing impedance, repelling traffic) are actively and cooperatively used to act on different paths, so as to dynamically adjust the path selection behavior of the driver by using the lever of time impedance, and guide the system traffic distribution to tend to be balanced.

[0016] Moreover, the parameter adjustment (period, green ratio, phase difference) of the existing green wave / red wave control is based on historical schemes, simple congestion detection (such as on-off type), or optimization search with efficiency as the target. There has never been a complete control method that can start from the "expected traffic transfer target" and gradually derive the first signal control strategy and the second signal control strategy based on the "intersection green time" and the "path phase difference" respectively.

[0017] The overall architecture of the present application is a special solution tailored for the specific task of "parallel arterial flow regulation". It clearly defines the data flow, control flow and feedback flow, forming a complete, implementable and engineering-oriented technical system.

[0018] According to the method disclosed in the first aspect of the present application, in step S2, when the absolute value of the relative saturation difference is greater than the relative load threshold, and the saturation of the high-saturation arterial is greater than the high-load threshold, and the saturation of the low-saturation arterial is less than the low-load threshold, it is judged that the traffic flow regulation needs to be started. The detection and triggering in the prior art are usually based on single-point occupancy / flow, and there has never been a technical solution that uses the "difference between the saturations of the two parallel arterials" as the core control index, and uses the "high road warning, low road safety" double protection threshold to form a robust triggering logic. It enables the present application to identify the specific problem of "parallel road network imbalance", and ensures the safety and economy of the regulation action.

[0019] According to the method disclosed by the first aspect of the application, the relative load threshold is configured as a critical value for characterizing the balance of the operation state of the two parallel arterials, the high load threshold is configured as a critical value for characterizing the traffic flow entering an unstable state, and the low load threshold is configured as a critical value for characterizing the remaining capacity for accepting diverted vehicles. The relative load threshold is in the range of 0.15-0.25, the high load threshold is in the range of 0.65-0.75, and the low load threshold is in the range of 0.5-0.6.

[0020] According to the method disclosed by the first aspect of the application, the step of determining the target diversion flow rate to be diverted from the high-saturation arterial to the low-saturation arterial comprises: determining the target diversion flow rate based on the current hourly traffic volume and the preset traffic flow rate of the two parallel arterials, with the goal of achieving equal saturation.

[0021] According to the method disclosed by the first aspect of the application, the step of implementing the first signal control strategy of increasing the travel delay of vehicles on the high-saturation arterial comprises: determining the additional travel delay to be added to the high-saturation arterial based on the target diversion flow rate, the current traffic flow rate of the high-saturation arterial, and the current travel time of the high-saturation arterial; and adjusting the signal phase difference between coordinated intersections on the high-saturation arterial based on the additional travel delay, so as to increase the probability of vehicles entering from an upstream intersection encountering a red light at a downstream intersection, thereby achieving red wave control. The application creatively establishes a quantitative mapping relationship between the "flow diversion amount" and the "red wave phase difference adjustment amount", which closely combines the macroscopic traffic distribution theory with the microscopic signal control operation, is the "brain" and "execution algorithm" of the application, and embodies high intelligence and precision.

[0022] According to the method disclosed by the first aspect of the application, the additional travel delay to be added to the high-saturation arterial is configured to be proportional to the proportion of the target diversion flow rate to the current traffic flow rate of the high-saturation arterial and the current travel time of the high-saturation arterial.

[0023] According to the method disclosed by the first aspect of the application, the step of implementing the second signal control strategy of reducing the travel delay of vehicles on the low-saturation arterial in real time comprises: adjusting the signal phase difference between coordinated intersections on the low-saturation arterial, so that vehicles entering from an upstream intersection can continuously pass through downstream intersections at an expected speed, thereby achieving green wave control; and / or adjusting the green light time proportion of the travel phase allocated to the high-saturation arterial and the travel phase allocated to the low-saturation arterial at the decision intersection of the two parallel arterials, so as to reduce the green light time for the high-saturation arterial and correspondingly increase the green light time for the low-saturation arterial.

[0024] According to the method disclosed by the first aspect of the application, the step of adjusting the green light time ratio comprises: calculating a first predicted traffic volume going to the high-saturation arterial road and a second predicted traffic volume going to the low-saturation arterial road respectively based on the target diversion volume; determining a first traffic demand intensity corresponding to the first predicted traffic volume and a second traffic demand intensity corresponding to the second predicted traffic volume respectively; the traffic demand intensity is configured as the ratio of the predicted traffic volume to the saturation traffic capacity of the lane group in the direction thereof, and the saturation traffic capacity is configured as the product of the number of lanes in the corresponding direction and the saturation flow rate of a single lane; and reallocating the total green light time of the phases for going to the two parallel arterial roads in the decision intersection based on the ratio of the first traffic demand intensity to the second traffic demand intensity, and then determining the green light time of each phase.

[0025] According to the method disclosed by the first aspect of the application, the method further comprises: configuring a comprehensive performance evaluation function, wherein the value of the comprehensive performance evaluation function is configured to be negatively related to the absolute value of the relative saturation difference, the average delay of the two parallel arterial roads, and / or the queue length; and after step S3, the value of the comprehensive performance evaluation function is calculated based on the updated traffic state information, and the regulation result and / or the optimization parameter are evaluated by using the value.

[0026] According to the method disclosed by the first aspect of the application, the method further comprises step S4 of converting the first signal control strategy and the second signal control strategy determined in step S3 into signal control instructions.

[0027] Although the "uneven parallel arterial road flow" is an objective phenomenon, for a long time, the mainstream research and engineering practice in the field of traffic control have focused on the efficiency improvement of "point, line and surface". It is insightful to propose "flow balance between parallel roads" as an independent and explicit technical problem that needs to be actively realized by signal control. The skilled person in the art is used to doing signal optimization under the premise of given origin-destination and path flow, while the application essentially integrates the two traditionally separated fields of "signal control" and "dynamic traffic assignment", and makes the signal control actively assume the role of adjusting the path flow.

[0028] Moreover, the core idea of "implementing green wave for one arterial road to attract traffic flow, and implementing red wave for the other parallel arterial road to repel traffic flow, so as to achieve balance" in the present application is a breakthrough "reverse coordination" thinking. In the traditional cognition, green wave should be used on the most important road; red wave is a helpless measure used to limit the flow. No technical solution has ever proposed to actively and systematically apply the two means simultaneously and reversely to two competing roads, and to take it as a kind of normalized regulation strategy based on rules. The idea of the present application is not a simple improvement or combination of the existing green wave or red wave technology, but proposes a completely new control paradigm.

[0029] The second aspect of the present application discloses a control agent for implementing the method disclosed in the first aspect of the present application, comprising:

[0030] a data acquisition unit configured to acquire traffic state data of two parallel arterials connecting the same origin-destination region in real time, calculate the saturation and relative saturation difference of the parallel arterials based on the acquired data, and determine the arterial with higher saturation as the high-saturation arterial and the other arterial as the low-saturation arterial;

[0031] a decision unit configured to determine whether to start traffic flow regulation based on the arterial saturation and relative saturation difference determined by the data acquisition unit, determine a target transfer flow to be transferred from the high-saturation arterial to the low-saturation arterial when the determination is YES, determine a first signal control strategy that increases the travel delay of vehicles on the high-saturation arterial based on the target transfer flow, and simultaneously determine a second signal control strategy that reduces the travel delay of vehicles on the low-saturation arterial in real time based on the target transfer flow;

[0032] wherein the first signal control strategy and the second signal control strategy act on the decision intersection at the starting region of the two parallel arterials or the coordinated intersection downstream thereof.

[0033] The control agent disclosed in the second aspect of the present application further comprises an execution unit configured to convert the first signal control strategy and the second signal control strategy determined by the decision unit into signal control instructions, and a feedback unit configured to evaluate the signal control instructions and implement updated signal control parameters, and feed back the updated parameters to the decision unit.

[0034] The control method and control agent for parallel arterial traffic flow coordination disclosed in the present application have the following beneficial effects: the relative saturation difference is used as the trigger condition for flow regulation between the parallel arterials, and the target transfer flow is used to achieve balanced regulation between the two parallel arterials, so that when the saturation of one of the parallel arterials reaches the upper limit of stable flow, the signal timing of the decision intersection upstream of the shunt arterial is adjusted to make the vehicles that do not enter the congested arterial enter the parallel road with large receiving capacity and small traffic pressure, thereby achieving uniform distribution of traffic flow, alleviating arterial congestion, and improving the utilization rate of the parallel road. The method can actively regulate the route selection of vehicles before congestion occurs or after congestion occurs, thereby inhibiting the bottleneck and improving the efficiency of the road network.

[0035] The control method and control agent for parallel arterial traffic flow coordination disclosed in the present application will be described in detail below with reference to the embodiments shown in the accompanying drawings and the reference numerals. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 A step flowchart of the control method in the present application is shown.

[0037] Figure 2 An example diagram of a scenario of parallel roads in the present application.

[0038] Figure 3 A framework flowchart of regulating an agent in the present application is shown. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0040] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative position relationship, movement condition, etc. between components in a certain attitude (as shown in the drawings), and if the certain attitude changes, the directional indications also change accordingly.

[0041] In addition, the description of “first”, “second” and the like in the present application is only for the purpose of description, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features defined with “first”, “second” can explicitly or implicitly include at least one of the features. In addition, the technical solutions of each embodiment can be combined with each other, but it must be based on the realization of a person of ordinary skill in the art, and when the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist, and is also not within the scope of protection required by the present application.

[0042] Figure 1 A step flowchart of the control method in the present application is shown. In combination with Figure 1As shown, the application discloses a control method for parallel arterial traffic volume coordination, which comprises the following steps: S1, real-time acquisition of traffic state data of two parallel arterials connecting the same origin-destination region, calculation of saturation and relative saturation difference of the parallel arterials based on the acquired data, and determination of the arterial with higher saturation as the high-saturation arterial and the other arterial as the low-saturation arterial; S2, judgment of whether to start traffic flow regulation based on the arterial saturation and relative saturation difference acquired in step S1; if yes, S3 is executed; S3, determination of a target transfer flow to be transferred from the high-saturation arterial to the low-saturation arterial, determination of a first signal control strategy for increasing the travel delay of vehicles on the high-saturation arterial based on the target transfer flow, and determination of a second signal control strategy for reducing the travel delay of vehicles on the low-saturation arterial in real time based on the target transfer flow; wherein the first signal control strategy and the second signal control strategy act on the decision intersection of the starting region of the two parallel arterials or the coordinated intersection downstream thereof.

[0043] Figure 2 The figure shows the scene of the parallel arterials in the application. Figure 2 As shown, the "parallel arterials" referred to in the application refer to two parallel roads with the same function connecting the same two regions, each road has multiple intersections (referred to as coordinated intersections), and there is no connecting road between the intersections. The parallel paths are two, if there is a connection, the number of paths to be decided increases, and the complexity of the agent construction increases. The application takes two arterials without connecting roads as an example for description. Vehicles make path selection at the decision intersection (the connecting intersection of the two parallel arterials in the starting region), and the real-time flow of the road with large capacity is also large, and vice versa.

[0044] In step S1, the traffic state data of two parallel arterials connecting the same origin-destination region are acquired in real time, including vehicle flow, speed, signal state, saturation and other parameters. The data can be collected in real time through the geomagnetic detector, video detector and signal machine interface arranged at the intersection, and a state matrix is created. The vector included in the state matrix can be expressed as: ; wherein is the A, B path flow, is the A, B path corresponding saturation, is the relative saturation difference of the two paths, is the average delay.

[0045] In addition, in urban traffic management, road saturation is an important indicator of road operating state. Figure 2 As shown, it is assumed that the maximum capacity of the key node or section of the A path is , and the maximum capacity of the key node or section of the B path is The actual traffic flow of path A during a certain period is The actual traffic flow of path B during a certain period is So, the saturation of path A. B-path saturation .

[0046] Saturation S reflects the relationship between road supply and demand under certain conditions, specifically indicating the real-time pressure on the road. If S≈1, the road is close to saturation, vehicle speed and operation decrease significantly, and traffic efficiency drops rapidly; when S... At a value of 0.5, the road resource utilization rate is low, with a significant surplus of vehicle capacity. Therefore, to measure the load on two paths, the relative saturation difference parameter between parallel arterial roads is introduced: Ideally, the saturation of the two paths will be approximately equal, that is... At this point, the vehicle distribution along paths A and B reaches dynamic equilibrium, and the entire system achieves optimal resource allocation. However, in real-world road networks, traffic signal interference and the randomness of driver choices can negatively impact traffic flow. Inevitably, fluctuations will occur within a certain range. Therefore, it is necessary to set reasonable thresholds to determine when and under what conditions a heuristic mechanism is initiated.

[0047] Based on this, in step S2, a threshold condition based on saturation is set. That is, when the absolute value of the relative saturation difference is greater than the relative load threshold, and the saturation of the high-saturation arterial road exceeds the high load threshold, while the saturation of the low-saturation arterial road is lower than the low load threshold, it is determined that traffic flow regulation needs to be initiated. Existing detection and triggering technologies are usually based on single-point occupancy / flow. No previous technical solution has used the "difference in saturation between two parallel arterial roads" as the core control indicator, supplemented by a dual-protection threshold of "high-road warning and low-road safety" to form a robust triggering logic. This enables this application to identify the specific problem of "parallel road network imbalance" and ensures the safety and economy of the regulation action.

[0048] The relative load threshold is configured as a critical value to characterize the balanced operation of the two parallel arterial roads, the high load threshold is configured as a critical value to characterize the traffic flow entering an unstable state, and the low load threshold is configured as a critical value to characterize the remaining capacity to accommodate diverted vehicles.

[0049] In a specific embodiment, the relative load threshold ranges from 0.15 to 0.25, that is, When | When the saturation level of both paths A and B is greater than 0.5, it indicates a significant difference in saturation levels. If the saturation levels of both paths A and B are greater than 0.5, then the operational states of paths A and B are unbalanced, and traffic flow regulation should be considered. When the two paths are in a balanced state and the vehicle distribution reaches dynamic stability, the system is considered to be in a stable state.

[0050] In order to enhance the robustness of the system and avoid false triggering, the relative saturation error is taken as the main basis, supplemented by high load threshold and low load threshold. The value range of the high load threshold is 0.65-0.75, and the value range of the low load threshold is 0.5-0.6.

[0051] Taking the case of A path as a high-saturation arterial road, the high load threshold can be set to 0.7. According to the traffic flow fundamental diagram model, when the saturation is between 0.7 and 0.75, the traffic flow is in the upper limit of stable flow, and vehicles can still pass through efficiently. When the saturation exceeds 0.8, the traffic flow enters the unstable region, and occasional disturbances occur, leading to the rapid formation and spread of congestion. Therefore, 0.7 is chosen as the threshold to reserve control time before congestion forms, achieve "early intervention and early relief", and avoid induction after A path is completely paralyzed, reducing the overall road network efficiency. Similarly, the same applies to the case where B path is a high-saturation arterial road.

[0052] Taking the case of B path as a low-saturation arterial road, the low load threshold can be set to 0.55. As a diversion path, B path must have sufficient residual communication capacity to accommodate vehicles diverted from A path. A saturation of 0.55 means that B path currently uses only 55% of its capacity, providing a safe buffer for diverted vehicles to ensure a smooth and controllable diversion process. If B path itself is heavily loaded, forced induction may lead to secondary congestion, so the system should maintain the existing control strategy and not perform induction operations when B conditions are not met. Similarly, the same applies to the case where A path is a low-saturation arterial road.

[0053] When both 1) the absolute value of the relative saturation difference is greater than the relative load threshold, and 2) the saturation of the high-saturation arterial road exceeds the high load threshold, and 3) the saturation of the low-saturation arterial road is lower than the low load threshold, traffic flow regulation is determined to be started. Specifically, based on the real-time traffic state data collected in step S1, the relationship between and is compared. If and are satisfied at the same time, and or and are satisfied, traffic flow regulation is started. The present application designs an intelligent control method that can control traffic signals based on reinforcement learning and state perception to start the control steps of traffic flow regulation after determining that traffic flow regulation needs to be started, specifically based on step S3.

[0054] The present application designs an intelligent control method that can control traffic signals based on reinforcement learning and state perception to start the control steps of traffic flow regulation after determining that traffic flow regulation needs to be started, specifically based on step S3.

[0055] Based on step S3, the signal control strategy of the induction system in the current state is calculated, and the action space is defined as: ; wherein, respectively represent the phase green light time of the A path and the B path entering the decision intersection, represent the red light time increment of the decision intersection. The following steps are specifically implemented:

[0056] Step S31, determine the target transfer flow that needs to be transferred from the high saturation arterial road to the low saturation arterial road, which includes: based on the current hourly traffic volume and the preset traffic flow of the two parallel arterial roads, the target transfer flow is determined to achieve the goal of equal saturation.

[0057] Specifically, first, real-time flow conversion and saturation calculation is performed based on the following formula:

[0058] ;

[0059] ;

[0060] );

[0061] );

[0062] wherein , is the traffic capacity of the A path and the B path, is the real-time traffic volume of the intersection detected by the road detector.

[0063] Then, the calculation of the target transfer flow is performed. The specific implementation process is as follows:

[0064] Let the arrival flow rate of the west entrance of the decision intersection be Q, and the traffic flow ratio before induction is left turn: straight a:b, then the distribution ratio is:

[0065] b;

[0066] The traffic volume that needs to be transferred is , if the saturation balance is to be achieved, it is required that:

[0067] (1) , then:

[0068] ;

[0069] (2) , then:

[0070] ;

[0071] Thus, the target flow that needs to be transferred can be solved.

[0072] or .

[0073] Step S32, based on the target transfer flow, determine the first signal control strategy that increases the vehicle travel delay on the high saturation arterial; at the same time, based on the target transfer flow, determine the second signal control strategy that reduces the vehicle travel delay on the low saturation arterial in real time.

[0074] The step of implementing the first signal control strategy of increasing the vehicle travel delay on the high saturation arterial includes: based on the target transfer flow, the current traffic flow of the high saturation arterial and the current travel time of the high saturation arterial, determining the additional travel delay to be added to the high saturation arterial; based on the additional travel delay, adjusting the signal phase difference between the coordinated intersections on the high saturation arterial to increase the probability of the vehicle entering from the upstream intersection encountering a red light at the downstream intersection, thereby realizing red wave control. The present application creatively establishes a quantitative mapping relationship between the "flow transfer amount" and the "red wave phase difference adjustment amount", which closely combines the macroscopic traffic distribution theory with the microscopic signal control operation, is the "brain" and "execution algorithm" of the present application, and embodies high intelligence and precision.

[0075] Regarding the first signal control strategy, it can be realized through the following calculation process:

[0076] 1) Implement green wave for the path with low saturation, ;

[0077] 2) Implement red wave for the path with high saturation, let the current travel time of the path be , let the travel time of the path need to be increased to in order to transfer flow . According to the impedance distribution principle, the impedances of the two paths should satisfy the flow distribution relationship under the new balance. And, where the additional travel delay to be added to the high saturation arterial is configured to be proportional to the proportion of the target transfer flow to the current traffic flow of the high saturation arterial, and the current travel time of the high saturation arterial, then the following formula is calculated: ; Where k=2.

[0078] Assume that at the decision intersection N1, the originally scheduled green light start time for the path is T, in order to realize the average delay, the green light time needs to be delayed , the formula is set as , =0.9.

[0079] Finally, red wave control is implemented for the vehicle entering the path, and the phase difference adjustment amount is: ; wherein is the original coordinated phase difference between the intersection i,j.

[0080] This is the steps of the second signal control strategy for reducing the vehicle delay on the low-saturation arterial road in real time, including: adjusting the signal phase difference between the coordinated intersections on the low-saturation arterial road, so that the vehicles entering from the upstream intersection can pass through the downstream intersection continuously at the desired speed, thereby realizing green wave control; and / or at the decision intersection of the two parallel arterial roads, adjusting the green time proportion of the traffic phase allocated to the high-saturation arterial road and the traffic phase allocated to the low-saturation arterial road to reduce the green time of the high-saturation arterial road and correspondingly increase the green time of the low-saturation arterial road.

[0081] Specifically, the step of adjusting the green time proportion includes: based on the target transfer flow, respectively calculating the first predicted traffic volume expected to go to the high-saturation arterial road and the second predicted traffic volume expected to go to the low-saturation arterial road after the transfer is implemented; respectively determining the first traffic demand intensity corresponding to the first predicted traffic volume and the second traffic demand intensity corresponding to the second predicted traffic volume; the traffic demand intensity is configured as the ratio of the predicted traffic volume to the saturated traffic capacity of the lane group in the direction thereof, and the saturated traffic capacity is configured as the product of the number of lanes in the corresponding direction and the saturated flow rate of a single lane; based on the proportion of the first traffic demand intensity and the second traffic demand intensity, reallocating the total green time of the phases for going to the two parallel arterial roads in the decision intersection, and then determining the green time of each phase.

[0082] The specific implementation of the green time allocation can be realized by the following process:

[0083] 1) , then: ;

[0084] 2) , then: ;

[0085] wherein n is the number of lanes, is the saturated flow rate;

[0086] The green light allocation time is represented by the following formula: ; , that is, the green time of each phase is equal to the total time multiplied by the proportion of its traffic demand intensity in the total intensity.

[0087] wherein, is the target transfer flow, the first predicted traffic volume is represented by , and the second predicted traffic volume is represented by , is the traffic demand intensity, and is the saturated traffic capacity, The sum of the green light of the left turn and the straight at the intersection is decided, i.e., The total green light time is decided.

[0088] In step S3, the comprehensive performance evaluation function is configured, and the value of the comprehensive performance evaluation function is configured to be negatively related to the absolute value of the relative saturation difference, the average delay of the two parallel arterials, and / or the queue length; after step S3, the value of the comprehensive performance evaluation function is calculated based on the updated traffic state information, and the regulation result and / or the optimization parameter are evaluated by using the value.

[0089] Specifically, in order to make the system better intervene, a reward function balancing the flow distribution and the traffic efficiency is defined:

[0090]

[0091] Wherein, α, β, γ are weight coefficients, and the values are respectively , The average delay is The queue length is. The agent realizes the flow balance and the congestion suppression by minimizing the value of the reward function.

[0092] In addition, the method of the application further includes step S4 of converting the first signal control strategy and the second signal control strategy determined in step S3 into signal control instructions. That is, according to the vector output by step S3, the calculation result is converted into specific signal control instructions:

[0093] (1) Adjust the green light time of the A and B paths ;

[0094] (2) Implement red wave delay for the path with high saturation at the decision intersection ;

[0095] (3) Implement green wave coordination for the path with low saturation, and the phase difference is set to

[0096] The execution layer guarantees that the received adjustment instructions are compatible with the signal machine control logic, and needs to have real-time and safety constraints.

[0097] In addition, after step S4, a feedback step is further included, which feeds back the traffic operation effect according to the execution scheme to evaluate the advantages and disadvantages of the scheme. The system uses a sliding time window mechanism to evaluate the traffic indicators in a 5-minute to 15-minute interval, such as the average delay, the queue length, and the traffic saturation, to update the network parameters in real time and feed back to the decision layer to realize continuous monitoring and optimization.

[0098] ​In the present application, although "uneven parallel arterial traffic flow" is an objective phenomenon, for a long time, the mainstream research and engineering practice in the field of traffic control have focused on the efficiency improvement of "point, line and surface". It is insightful to propose "traffic flow balance between parallel roads" as an independent and explicit technical problem that needs to be actively realized through signal control. The skilled person in the art is used to optimizing signals under the premise of given starting point and path flow, while the present application essentially integrates "signal control" and "dynamic traffic assignment", which are traditionally separated fields, and lets the signal control actively assume the role of adjusting path flow.

[0099] Moreover, in the present application, the core idea of "implementing green wave on one arterial road to attract traffic flow, while implementing red wave on another parallel arterial road to repel traffic flow, so as to achieve balance" is a breakthrough "reverse coordination" thinking. In the traditional cognition, green wave should be used on the most important road; red wave is a helpless measure used to limit traffic flow. No technical solution has ever proposed to actively and systematically apply these two means simultaneously and reversely to two competing roads, and to use it as a kind of normalized and rule-based regulation strategy. The idea of the present application is not a simple improvement or combination of existing green wave or red wave technology, but proposes a completely new control paradigm.

[0100] Figure 3 The framework flowchart of the regulation agent in the present application is shown. In combination with Figure 3 The present application also discloses a regulation agent for implementing the foregoing method, which can also be called a system, comprising:

[0101] a data acquisition unit configured to acquire traffic state data of two parallel arterial roads connecting the same origin-destination area in real time, calculate the saturation and relative saturation difference of the parallel arterial roads based on the acquired data, and determine the arterial road with higher saturation as the high-saturation arterial road and the other arterial road as the low-saturation arterial road;

[0102] a decision unit configured to determine whether to start traffic flow regulation based on the arterial road saturation and relative saturation difference determined by the data acquisition unit, determine the target transfer flow that needs to be transferred from the high-saturation arterial road to the low-saturation arterial road when the determination is yes, determine the first signal control strategy that increases the travel delay of vehicles on the high-saturation arterial road based on the target transfer flow, and simultaneously determine the second signal control strategy that reduces the real-time travel delay of vehicles on the low-saturation arterial road based on the target transfer flow;

[0103] wherein the first signal control strategy and the second signal control strategy act on the decision intersection at the starting area of the two parallel arterial roads or the coordinated intersection downstream thereof.

[0104] The regulation intelligent agent further comprises an execution unit configured to convert the first signal control strategy and the second signal control strategy determined by the decision unit into signal control instructions; and a feedback unit configured to evaluate the signal control instructions, update the signal control parameters, and feed back the updated parameters to the decision unit.

[0105] In combination Figure 3 As shown in the figure, the perception layer represents the data acquisition unit, the decision layer represents the decision unit, the execution layer represents the execution unit, and the feedback layer represents the feedback unit.

[0106] All the functions, effects and implementation processes described in the control method in the application are also applicable to the regulation intelligent agent in the application. To avoid repetition, they will not be described here.

[0107] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the purpose and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A control method for coordinated regulation of traffic volume on parallel arterial roads, characterized in that, Includes the following steps: S1. Real-time acquisition of traffic status data of two parallel arterial roads connecting the same origin and destination areas. Based on the acquired data, calculate the saturation and relative saturation difference of the parallel arterial roads, and determine that the arterial road with higher saturation is a high-saturation arterial road and the other arterial road is a low-saturation arterial road. S2, based on the saturation of the main road and the difference in relative saturation obtained in step S1, determine whether to initiate traffic flow regulation; If so, then execute S3; S3, determine the target transfer flow that needs to be transferred from the high-saturation arterial road to the low-saturation arterial road, and based on the target transfer flow, determine a first signal control strategy that increases the vehicle travel delay on the high-saturation arterial road; at the same time, based on the target transfer flow, determine a second signal control strategy that reduces the vehicle travel delay on the low-saturation arterial road in real time. The first signal control strategy and the second signal control strategy apply to the decision intersection at the starting area of ​​the two parallel arterial roads or the coordination intersection downstream of them.

2. The method according to claim 1, characterized in that, In step S2, when the absolute value of the relative saturation difference is greater than the relative load threshold, and the saturation of the high-saturation arterial road exceeds the high load threshold while the saturation of the low-saturation arterial road is lower than the low load threshold, it is determined that traffic flow regulation needs to be initiated.

3. The method according to claim 1, characterized in that, The steps for determining the target transfer flow that needs to be shifted from a high-saturation arterial road to a low-saturation arterial road include: determining the target transfer flow based on the current hourly traffic volume and preset traffic flow of the two parallel arterial roads with the goal of achieving equal saturation.

4. The method according to claim 1, characterized in that, The steps for implementing a first signal control strategy to increase vehicle delays on high-saturation arterial roads include: Based on the target diversion flow, the current traffic flow on the high-saturation arterial road, and the current travel time on the high-saturation arterial road, determine the additional travel delay that will be added to the high-saturation arterial road; Based on the additional travel delay, the signal phase difference between coordinated intersections on high-saturation arterial roads is adjusted to increase the probability that vehicles entering from upstream intersections will encounter red lights at downstream intersections, thereby achieving red wave control.

5. The method according to claim 4, characterized in that, The additional travel delay added to high-saturation arterial roads will be configured to be proportional to the proportion of the target transfer flow to the current traffic flow on the high-saturation arterial road, as well as the current travel time on the high-saturation arterial road.

6. The method according to claim 1, characterized in that, The steps of a second signal control strategy to reduce vehicle delays on low-saturation arterial roads in real time include: Adjusting the signal phase difference between coordinated intersections on low-saturation arterial roads allows vehicles entering from upstream intersections to pass through downstream intersections continuously at the desired speed, thereby achieving green wave control. And / or, at decision intersections of two parallel arterial roads, adjust the ratio of green light time allocated to the traffic phase going to the high-saturation arterial road and the traffic phase going to the low-saturation arterial road to reduce the green light time going to the high-saturation arterial road and correspondingly increase the green light time going to the low-saturation arterial road.

7. The method according to claim 6, characterized in that, The steps for adjusting the green light time ratio include: based on the target transfer flow, calculating the first predicted traffic volume expected to go to the high-saturation arterial road and the second predicted traffic volume expected to go to the low-saturation arterial road after the transfer is implemented; The first traffic demand intensity corresponding to the first predicted traffic volume and the second traffic demand intensity corresponding to the second predicted traffic volume are determined respectively. The traffic demand intensity is configured as the ratio of the predicted traffic volume to the saturation capacity of the lane group in its direction, and the saturation capacity is configured as the product of the number of lanes in the corresponding direction and the saturation flow rate of a single lane. Based on the ratio of the first traffic demand intensity to the second traffic demand intensity, the total green light time for the phases at the decision intersection used to travel to the two parallel arterial roads is reallocated, thereby determining the green light time for each phase.

8. The method according to any one of claims 1-7, characterized in that, The method further includes step S4, which converts the first signal control strategy and the second signal control strategy determined in step S3 into signal control commands.

9. A regulatory agent for implementing the method according to any one of claims 1-8, characterized in that, include: The data acquisition unit is configured to acquire traffic status data of two parallel arterial roads connecting the same origin and destination areas in real time, calculate the saturation and relative saturation difference of the parallel arterial roads based on the acquired data, and determine that the arterial road with higher saturation is a high-saturation arterial road and the other arterial road is a low-saturation arterial road. The decision-making unit is constructed based on the arterial road saturation and relative saturation difference determined by the data acquisition unit to determine whether to initiate traffic flow regulation. When the determination is yes, the target transfer flow that needs to be transferred from the high-saturation arterial road to the low-saturation arterial road is determined. Based on the target transfer flow, a first signal control strategy that increases the vehicle travel delay on the high-saturation arterial road is determined. At the same time, based on the target transfer flow, a second signal control strategy that reduces the vehicle travel delay on the low-saturation arterial road in real time is determined. The first signal control strategy and the second signal control strategy apply to the decision intersection at the starting area of ​​the two parallel arterial roads or the coordination intersection downstream of them.

10. The regulatory agent according to claim 9, characterized in that, The control agent further includes an execution unit, which is configured to convert the first signal control strategy and the second signal control strategy determined by the decision unit into signal control instructions. The feedback unit is configured to evaluate the signal control commands, update the signal control parameters, and feed the updated parameters back to the decision unit.

Images