Multi-traffic-subarea distributed collaborative optimization method and system based on double-layer architecture

Abstract

The invention provides a multi-traffic subarea distributed collaborative optimization method and system based on a double-layer architecture, and relates to the technical field of traffic control and automatic control, and the method comprises the steps: building a weighted directed graph through a road network model, and dividing the weighted directed graph into a plurality of traffic subareas; based on the macroscopic basic diagram of each traffic sub-region, constructing an upper-layer multi-sub-region collaborative optimization model, and obtaining an optimal accumulated vehicle number of each traffic sub-region under collaborative control; optimizing the timing of each intersection in the traffic subareas by taking the optimal accumulated vehicle number of each traffic subarea as the constraint of the lower-layer single subarea optimization control model to obtain an optimal timing scheme of the controlled road network; according to the method, the overall traffic efficiency of each sub-region and a road network and flow distribution balance among the sub-regions are considered, based on a double-layer cascade architecture, the upper layer obtains the optimal accumulated vehicle number of each sub-region under cooperative control, through sub-region boundary control, cascading with the lower layer is carried out on the conditions of intersections in the sub-regions, parameters are continuously updated, and multi-sub-region cooperative optimization is carried out.

Description

Technical Field

[0001] The present disclosure relates to the technical field of traffic control and automatic control, and in particular to a distributed collaborative optimization method and system for multiple traffic sub-areas based on a double-layer architecture. Background Art

[0002] With the increasing number of cars, traffic congestion is becoming more and more serious, expanding from single intersections and single routes to regional congestion. Therefore, control over traffic areas is of great practical significance for alleviating urban congestion.

[0003] The current regional optimization control is often implemented in one or several smaller-scale areas based on the characteristic curves of their macro basic graphs when they are running alone, and by regulating their regional boundary signals. However, in the regulation process, each area only relies on the optimal cumulative number of vehicles in its own macro basic graph, without considering the mutual influence of traffic flow in the coupled area and the balanced distribution of traffic flow, which has limitations.

[0004] In fact, when optimizing traffic control in multiple areas, the influence between the areas makes it impossible for them to reach the optimal cumulative number of vehicles in their macro basic diagram at the same time. When optimizing control in multiple areas, while ensuring that the cumulative number of vehicles in each sub-area is as close as possible to its own theoretical optimal cumulative number of vehicles, it is also necessary to consider the coordination and differences between the sub-areas, the operation of vehicles in the sub-areas and other factors to achieve an asymmetric balance among the sub-areas.

[0005] Therefore, in traditional regional optimization control, there is a lack of consideration of combining the sub-interval collaborative optimization strategy with the signal optimization control strategy within the sub-interval and insufficient consideration of the asymmetric balance problem of the sub-interval, resulting in low control efficiency. Summary of the invention

[0006] In order to solve the above problems, the present invention proposes a multi-traffic sub-area distributed collaborative optimization method and system based on a double-layer architecture, taking into account the traffic efficiency of each sub-area itself and the overall traffic network, and the balanced traffic distribution between sub-areas. Based on the double-layer cascade architecture, the upper layer obtains the optimal cumulative number of vehicles in each sub-area under collaborative control, and through the sub-area boundary control, cascades with the lower layer on the conditions of each intersection inside the sub-area and continuously updates the parameters to perform multi-sub-area collaborative optimization.

[0007] According to some embodiments, the present disclosure adopts the following technical solutions:

[0008] The multi-traffic sub-area distributed collaborative optimization method based on a two-layer architecture includes:

[0009] According to the topological structure of the controlled road network and the collected traffic flow data, a road network model is established, and the traffic flow data in the road network model is improved through simulation;

[0010] The road network model is used to construct a weighted directed graph. By dividing the region of the weighted directed graph, several traffic sub-areas are obtained, and the macro basic graph of each traffic sub-area is drawn;

[0011] Based on the macro basic graph of each traffic sub-area, the road network optimization problem is decomposed into multiple sub-area collaborative optimization problems, and the upper-level multi-sub-area collaborative optimization model is constructed. The distributed optimization algorithm is used to solve the global optimal solution, and the optimal cumulative number of vehicles in each traffic sub-area under collaborative control is obtained.

[0012] The optimal cumulative number of vehicles in each traffic sub-area is used as the constraint of the lower-level single-sub-area optimization control model, and the timing of each intersection within the traffic sub-area is optimized to obtain the optimal timing plan for the controlled road network.

[0013] According to some embodiments, the present disclosure adopts the following technical solutions:

[0014] The multi-traffic sub-area distributed collaborative optimization system based on a two-layer architecture includes:

[0015] The model building module is configured to: build a road network model according to the topological structure of the controlled road network and the collected traffic flow data, and improve the traffic flow data in the road network model through simulation;

[0016] The area division module is configured to: construct a weighted directed graph using a road network model, obtain a number of traffic sub-areas by dividing the weighted directed graph into regions, and draw a macro basic graph of each traffic sub-area;

[0017] The collaborative optimization module is configured to: decompose the road network optimization problem into multiple sub-area collaborative optimization problems based on the macro basic graph of each traffic sub-area, construct an upper-level multi-sub-area collaborative optimization model, and use a distributed optimization algorithm to solve the global optimal solution to obtain the optimal cumulative number of vehicles in each traffic sub-area under collaborative control;

[0018] The timing optimization module is configured to use the optimal cumulative number of vehicles in each traffic sub-area as a constraint of the lower-level single-sub-area optimization control model, optimize the timing of each intersection within the traffic sub-area, and obtain the optimal timing plan for the controlled road network.

[0019] According to some embodiments, the present disclosure adopts the following technical solutions:

[0020] A computer program product includes a computer program, which, when executed by a processor, implements the multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture.

[0021] According to some embodiments, the present disclosure adopts the following technical solutions:

[0022] A non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the distributed collaborative optimization method of multiple traffic sub-areas based on a double-layer architecture is implemented.

[0023] According to some embodiments, the present disclosure adopts the following technical solutions:

[0024] An electronic device comprises: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device implements the multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] Compared with traditional small-scale regional control, the present invention not only considers the macro-basic map of each region, but also takes into account the coordination between regions and the differences in size between regions, and adopts a road network composed of multiple regions on a larger scale; on the basis of using the Louvain algorithm to divide the large-scale road network into sub-regions, the Bregman Alternating Direction Method of Multipliers with Consensus variable (CBADMM) is used to solve the optimization target under collaborative control, which solves the lack of consideration of combining the sub-interval collaborative optimization strategy with the signal optimization control strategy within the sub-region in traditional regional optimization control and the lack of consideration of the asymmetric balance problem of the sub-interval. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings constituting a part of the present disclosure are used to provide a further understanding of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure.

[0028] Figure 1 This is a flow chart of the method of Example 1.

[0029] Figure 2 This is an example diagram of the road network model of Example 1.

[0030] Figure 3 This is an example diagram of sub-area division in Example 1. DETAILED DESCRIPTION

[0031] The present disclosure is further described below in conjunction with the accompanying drawings and embodiments.

[0032] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present disclosure belongs.

[0033] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0034] Example 1

[0035] In one embodiment of the present disclosure, a multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture is provided, including:

[0036] Step S1: establishing a road network model according to the topological structure of the controlled road network and the collected traffic flow data, and improving the traffic flow data in the road network model through simulation;

[0037] Step S2: construct a weighted directed graph using the road network model, obtain several traffic sub-areas by dividing the regions of the weighted directed graph, and draw a macro basic graph of each traffic sub-area;

[0038] Step S3: Based on the macro basic graph of each traffic sub-area, the road network optimization problem is decomposed into multiple sub-area collaborative optimization problems, an upper-level multi-sub-area collaborative optimization model is constructed, and a distributed optimization algorithm is used to solve the global optimal solution to obtain the optimal cumulative number of vehicles in each traffic sub-area under collaborative control;

[0039] Step S4: Taking the optimal cumulative number of vehicles in each traffic sub-area as the constraint of the lower-level single-sub-area optimization control model, the timing of each intersection within the traffic sub-area is optimized to obtain the optimal timing plan for the controlled road network.

[0040] As an embodiment, the disclosed multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture considers the traffic efficiency of each sub-area itself and the overall traffic network, and the balance of traffic distribution between sub-areas. Based on the double-layer cascade architecture, the upper layer obtains the optimal cumulative number of vehicles in each sub-area under collaborative control, and through the sub-area boundary control, cascades with the lower layer on the conditions of each intersection within the sub-area and continuously updates the parameters to perform multi-sub-area collaborative optimization, such as Figure 1 The specific implementation process is as follows:

[0041] Step 1: Collect the topological information and traffic flow data of the controlled road network, and use traffic simulation modeling tools to establish Figure 2The road network model shown in the figure is used, and the traffic flow data in the road network model is improved through simulation.

[0042] The traffic data here refers to the unit traffic volume collected through pre-set data collection points and collection cycles; the traffic simulation modeling tool can use existing software, such as Vissim, etc. Of course, technical personnel in this field can choose the corresponding traffic simulation modeling tool according to their own situation and needs, and are not limited to the above examples, and are not limited here.

[0043] Furthermore, data collection points are set up at 5 meters from the stop line of the intersection, 30 meters from the exit road of the intersection, and in the middle of the road section to make the acquired traffic data more reasonable and accurate; further, in terms of traffic flow data, the number of vehicles collected continuously for 15 minutes multiplied by 4 is taken as the hourly flow.

[0044] Step 2: Use the Louvain community detection algorithm for weighted directed graphs based on simulation data to divide the entire road network into multiple traffic sub-areas.

[0045] Based on the correlation of traffic density, traffic volume and signal cycle, a correlation model is established, and the correlation between adjacent intersections is used as the edge weight to construct a weighted directed graph of the road network model; the Louvain community detection algorithm is used to maximize the modularity increment, and the nodes in the weighted directed graph are merged into communities until the communities where the nodes are located no longer change, thereby dividing the weighted directed graph of the road network model into multiple traffic sub-areas. The specific steps are as follows:

[0046] Step 2.1: Establish association model and modularity model

[0047] The calculation formula of traffic flow correlation is:

[0048]

[0049] Among them, I q(i→j) represents the traffic flow correlation from the upstream intersection i to the downstream intersection j, a represents the traffic flow dispersion coefficient, T i→j Usually, 0.8 times the travel time of traffic from intersection i to intersection j is taken, n i→j represents the number of vehicle flow branches from the upstream intersection i to the downstream intersection j. For a standard intersection, n i→j =3, is the maximum flow rate of the entrance road from intersection i to intersection j, q i→j is the unit vehicle flow of the entrance road from intersection i to intersection j.

[0050] Considering that the traffic flow between intersections is bidirectional, the traffic flow correlation degree I from intersection i to j is q(i→j) The correlation degree I with the traffic flow from intersection j to iq(j→i) Therefore, the traffic flow correlation between intersections j and i is expressed as I q(i→j) with I q(j→i) Average of:

[0051]

[0052] The calculation formula of signal period correlation is:

[0053]

[0054] Where R is the maximum possible value of the ratio of signal control cycles of adjacent intersections, C i is the signal period value of intersection i.

[0055] The calculation formula of the road section traffic density correlation is:

[0056]

[0057] Among them, ρ s(i→j) represents the vehicle density from the upstream intersection i to the downstream intersection j under saturation conditions, q i→j is the unit vehicle flow of the entrance road from intersection i to intersection j, N i→j Indicates the number of lanes, L i→j is the length of the connecting section from intersection i to j.

[0058] In summary, the intersection correlation model is given as:

[0059]

[0060] In the Louvain community detection algorithm, a community is represented by a group consisting of one or several nodes, which are continuously merged in the algorithm process. The modularity is used to measure the quality of the current community division. The above association model is used as the initial edge weight of the road section to give the modularity formula as follows:

[0061]

[0062] Among them, k i represents the sum of the associations of the road sections connected to intersection i, δ(c i ,c j ) indicates whether intersection i and intersection j belong to the same community.

[0063] Step 2.2: Use the association degree as the road segment weight to divide the road network into sub-areas

[0064] In the actual division process, Figure 2 The road network model shown in the figure is divided into sub-areas, and the following is obtained: Figure 3 The area division results shown are as follows:

[0065] Step 2.2.1: Modularity optimization phase

[0066] (1) Collect and count the traffic information of intersections and connecting sections in the sub-area (initially the entire road network), substitute it into the established correlation model for calculation, and obtain the correlation between adjacent intersections.

[0067] (2) Initially, each intersection is regarded as a node, and each node is regarded as a community. At this time, the number of communities divided by the algorithm is the number of nodes.

[0068] (3) Merge nodes with adjacent communities and calculate the modularity increment:

[0069]

[0070] Among them, ∑ in represents the sum of the correlations of all road sections within community C during the process of merging into community C at the intersection, ∑ out represents the sum of the association degrees of all road sections pointing to the intersections in community C, k i,in Represents the sum of the associations of all road sections connecting intersection i and community C.

[0071] (4) If all modularity increments obtained during the merging of all communities adjacent to the intersection with the node are less than 0, the intersection is returned to the original community. Otherwise, the intersection is merged into the community with the largest modularity increment and the community and corresponding data are updated until the community where any intersection is located no longer changes, and the modularity optimization stage ends.

[0072] Step 2.2.2: Divide the network into cohesive phases

[0073] (1) Each community divided in the modularity optimization phase is merged and compressed into a new node, and the weight is updated according to the relevance of the road segments connected to the new node. The sum of the relevance of all road segments connecting two different communities is used as the weight between the new nodes compressed and merged from the two communities.

[0074] (2) Go back to step 2.2.1 and repeatedly optimize the modularity until the modularity of the entire map no longer changes. The community finally merged is a controlled traffic sub-area. The division result is as follows: Figure 3 shown.

[0075] Step 3: In Vissim, gradually increase the traffic volume of each sub-area divided in step 2 and calibrate them separately, draw the macro basic diagram of each sub-area, and establish the relationship between the cumulative number of vehicles and the trip completion flow.

[0076] Specifically, step 3.1: using Vissim to input the traffic flow of each sub-area according to the traffic flow data.

[0077] Step 3.2: Data collection points are set up on each road section. In order to obtain the free flow, critical flow and congested flow of the macro basic diagram, the vehicle flow input needs to be gradually increased during the input process.

[0078] Step 3.3: Based on the collected data of input and output traffic of each sub-area, calculate the cumulative number of vehicles in the sub-area:

[0079]

[0080] Where n(k) represents the cumulative number of vehicles in the sub-area during the kth sampling period, n in (k) represents the number of vehicles entering the sub-area during the kth sampling period, n out (k) represents the number of vehicles passing through the separation zone in the kth sampling period, Δt represents the time interval of the sampling period, q(k) represents the traffic flow rate passing through the separation zone, and s represents the total number of lanes passing through the separation zone.

[0081] Step 3.4: Draw a scatter plot with the cumulative number of vehicles as the horizontal axis and the trip completion flow as the vertical axis. Use the least squares method to fit the relationship between the cumulative number of vehicles and the trip completion flow into a cubic function, and draw the characteristic curve of the macro basic graph. The cubic function is expressed by the formula:

[0082] G(n(k))=an(k) 3 +bn(k) 2 +cn(k)+d

[0083] Where n(k) is the cumulative number of vehicles, G(n(k)) is the trip completion flow, and a, b, c, and d are the coefficients of the fitted cubic function.

[0084] Step 4: Decompose the road network optimization problem into multiple sub-area collaborative optimization problems, determine the control objectives of the upper-level multi-sub-area collaborative optimization model, and use the distributed optimization algorithm CBADMM to solve the global optimal solution, that is, the optimal cumulative number of vehicles under the collaborative control of each sub-area.

[0085] The collaborative optimization goal needs to consider the interaction and influence of information between sub-areas, which also means that all sub-areas cannot achieve the theoretical optimal cumulative number of vehicles at the same time. We can only find the optimal cumulative number of vehicles in each sub-area under the coordinated game. Therefore, the collaborative control aims to maximize the overall traffic efficiency of the road network. At the same time, the asymmetric balance of traffic flow between sub-areas is caused by the differences in the traffic cost between each traffic sub-area and the size and capacity of each traffic sub-area. The objective function is a non-convex function, so the CBADMM algorithm considering non-convex optimization is used to solve the optimal cumulative number of vehicles under collaborative control. CBADMM is an algorithm based on the decomposition and coordination process. It finds a solution to a large global problem by coordinating the solutions of small local sub-problems. Specifically:

[0086] Step 4.1: Construction of multi-sub-area collaborative optimization model

[0087] (1) Setting collaborative optimization goals

[0088] When designing the upper-level collaborative optimization objectives, it is necessary to consider the overall traffic status of the road network, the traffic operation efficiency between sub-areas, and the asymmetric balance of vehicle distribution in sub-areas.

[0089] According to the properties of the macro basic graph, when the cumulative number of vehicles in a sub-area reaches the optimal cumulative number of its basic graph, its trip completion flow is the largest, that is, the traffic efficiency is the largest. Therefore, in the coordinated optimization process, it should be ensured that the solution of each sub-area is near the left side of the optimal cumulative number of vehicles. Considering the differences in the macro basic graphs of different traffic sub-areas, that is, the trip completion flow varies with the cumulative number of vehicles to different degrees, the square of the difference between the trip completion flow and the maximum trip completion flow is used to reflect the quality of the traffic status:

[0090]

[0091] After normalization, the optimization target is given:

[0092]

[0093] Where R represents the sub-area set, n d represents the optimal cumulative number of vehicles, G i (n d ) represents the trip completion flow of sub-area i under the optimal cumulative number of vehicles, G i (n) represents the trip completion flow when the cumulative number of vehicles in sub-area i is n at the current moment.

[0094] When performing sub-area collaborative optimization, a very important focus is the efficiency of traffic transfer between each mutually coupled sub-area. The efficiency of traffic transfer between coupled sub-areas can effectively reflect the quality of sub-area collaborative optimization. The relative road resistance on the road section connecting the sub-areas is taken as the optimization target. By minimizing the relative road resistance, reasonable traffic distribution is achieved to avoid excessive congestion in some sub-areas. Road resistance is often expressed by the travel cost on the road section, so the optimization target is given:

[0095]

[0096] Where l represents the connecting section from upstream sub-area i to downstream sub-area j, t l,0 is the travel time through section l in the free flow phase, t l,s is the travel time at the saturation moment, q l,ij is the traffic volume on road section l, C l is the traffic capacity of section l, α and β are empirical parameters that affect road resistance, usually taken as α=0.15 and β=4.

[0097] In order to avoid the problem of unbalanced traffic distribution caused by excessively high traffic loads in some sub-areas and low traffic loads in other sub-areas during the collaborative optimization process, an equilibrium term is introduced; considering that the balance of sub-intervals is asymmetric, the imbalance of sub-intervals is reduced by minimizing the relative traffic pressure of coupled sub-intervals.

[0098] Given the sub-area balance optimization objectives:

[0099]

[0100] Among them, n i is the current number of vehicles in sub-area i, n i,max Indicates the maximum number of vehicles that sub-area i can carry.

[0101] (2) Given constraints

[0102]

[0103] Among them, n i is the current cumulative number of vehicles in sub-area i, n ii represents the portion of the current cumulative number of vehicles in sub-area i whose destinations are inside the sub-area, n ij represents the portion of the current cumulative number of vehicles in sub-area i heading to sub-area j, C net represents the road network capacity, n i,max represents the maximum number of vehicles that can be accommodated in sub-area i, l represents a road section connecting the upstream sub-area i and the downstream sub-area j, C l Represents the traffic capacity of road section l.

[0104] Step 4.2: Solve the collaborative optimization objective using the CBADMM algorithm

[0105] (1) Initialization parameters

[0106] Initialize the timing plan for each intersection through the Webster timing method, and initialize the cumulative number of vehicles in each sub-area based on the current road network information Lagrange multiplier λ 0 , penalty parameter ρ, and in order to achieve information interaction and coordinated control between sub-areas in the distribution optimization process and solve the consistency optimization problem, the co-identification variable z is introduced 0 As an augmentation term of the objective function:

[0107]

[0108] Given the consistency constraints:

[0109] z i =n i

[0110] Where ρ is the penalty coefficient used to adjust the impact of the equilibrium term to ensure that the traffic efficiency of individual sub-areas will not be sacrificed in the process of balancing the vehicle distribution in the sub-areas, n d The optimal cumulative number of vehicles representing the macro basic graph of the sub-area, z i is the part of the global consensus variable z that is related to sub-area i. In short, z i When is a linear function of z; the local decisions of multiple sub-areas are coordinated through the global consensus variable, the traffic distribution of each sub-area is adjusted, and the ideal cumulative number of vehicles in each sub-area under coordination is obtained.

[0111] The augmented Lagrangian function form is further given:

[0112]

[0113] (2) Update the cumulative number of vehicles n:

[0114]

[0115] Where k represents the number of iterations, Δ G represents the Bregman divergence, represents the gradient; represents the inner product, since G i (n) is a univariate function of the sub-area with respect to the cumulative number of vehicles, and its value is and The product of .

[0116] (3) Update the consensus variables:

[0117]

[0118] (4) Update the Lagrange multiplier λ:

[0119]

[0120] (5) Determine whether the current solution converges or whether the number of iterations meets the maximum number of iterations. If so, output the upper-level collaborative objective and connect it as a constraint of the lower-level optimization objective. Otherwise, proceed to step (6).

[0121] Raw residual:

[0122]

[0123] Dual residual:

[0124]

[0125] where ε pri represents the original error, ε dual represents the dual error.

[0126] (6) According to the upper-layer collaborative target at the current iteration number, the upstream and downstream intersections of the sub-area connecting section are adjusted, the section flow is obtained through the simulation software, and the step (2) is returned to update the parameters.

[0127] Step 5: The optimal solution of the upper-level multi-sub-area collaborative control model in each iteration is used as the constraint of the lower-level single-sub-area optimization control model, and the feedback controller is used to achieve the upper-level collaborative control goal.

[0128] Solve the lower framework, and use the optimal solution of the upper multi-sub-area coordinated control model as the connection constraint of the signal timing optimization model in the lower sub-area. Use the feedback controller to perform boundary control to maintain the optimal number of vehicles under the coordinated control of each sub-area, and use the genetic-simulated annealing algorithm to optimize the timing of each intersection in the sub-area. Consider the dynamics of traffic flow and set the maximum control time of the road network. Once the maximum control time of the road network is exceeded, it is necessary to re-enter the sub-area optimization stage. Specifically:

[0129] Step 5.1 Use feedback controller to implement boundary control on sub-areas

[0130] The boundary control target is the optimal cumulative number of vehicles under the sub-area cooperative control required by the upper model, and the control object is the traffic lights at the boundary intersections of each sub-area.

[0131] The optimal cumulative number of vehicles and the deviation of the actual cumulative number of vehicles in the next sampling period are calculated using the traffic flow entering and leaving the sub-area boundary intersection and the multi-sub-area boundary traffic flow model, and used as the input of the feedback controller. Given the multi-sub-area boundary traffic flow model:

[0132]

[0133] Among them, n i (k) is the cumulative number of vehicles in sub-area i during the kth sampling period, q ji is the traffic flow from sub-area j to sub-area i, and Δt represents the time interval of the sampling period.

[0134] The green-signal ratio of the boundary intersections is allocated according to the optimal cumulative number of vehicles, the deviation of the actual cumulative number of vehicles in the sub-area in the next sampling period, and the entry and exit flow of each intersection at the boundary of the sub-area.

[0135] Step 5.2 Optimization of intersections within sub-areas

[0136] Under the constraint of maintaining the optimal sub-area cumulative number of vehicles under boundary control, the signal timing of each intersection is optimized under the framework of a distributed optimization algorithm, taking delay time and queue length as optimization objectives.

[0137] Output the timing plan to determine whether the maximum control time of the road network has been reached. If so, return the data to the upper-level framework for sub-area coordination and optimization. Otherwise, execute the current plan until the maximum control time of the road network has been reached.

[0138] Example 2

[0139] In one embodiment of the present disclosure, a multi-traffic sub-area distributed collaborative optimization system based on a double-layer architecture is provided, comprising:

[0140] The model building module is configured to: build a road network model according to the topological structure of the controlled road network and the collected traffic flow data, and improve the traffic flow data in the road network model through simulation;

[0141] The area division module is configured to: construct a weighted directed graph using a road network model, obtain a number of traffic sub-areas by dividing the weighted directed graph into regions, and draw a macro basic graph of each traffic sub-area;

[0142] The collaborative optimization module is configured to: decompose the road network optimization problem into multiple sub-area collaborative optimization problems based on the macro basic graph of each traffic sub-area, construct an upper-level multi-sub-area collaborative optimization model, and use a distributed optimization algorithm to solve the global optimal solution to obtain the optimal cumulative number of vehicles in each traffic sub-area under collaborative control;

[0143] The timing optimization module is configured to use the optimal cumulative number of vehicles in each traffic sub-area as a constraint of the lower-level single-sub-area optimization control model, optimize the timing of each intersection within the traffic sub-area, and obtain the optimal timing plan for the controlled road network.

[0144] Example 3

[0145] In one embodiment of the present disclosure, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the distributed collaborative optimization method of multiple traffic sub-areas based on a double-layer architecture is implemented.

[0146] Example 4

[0147] In one embodiment of the present disclosure, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the distributed collaborative optimization method of multiple traffic sub-areas based on a double-layer architecture is implemented.

[0148] Example 5

[0149] In one embodiment of the present disclosure, an electronic device is provided, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture.

[0150] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0152] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Technical personnel in the relevant field should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. A multi-traffic sub-area distributed collaborative optimization method based on a two-layer architecture, characterized in that: include: According to the topological structure of the controlled road network and the collected traffic flow data, a road network model is established, and the traffic flow data in the road network model is improved through simulation; The road network model is used to construct a weighted directed graph. By dividing the region of the weighted directed graph, several traffic sub-areas are obtained, and the macro basic graph of each traffic sub-area is drawn; Based on the macro basic graph of each traffic sub-area, the road network optimization problem is decomposed into multiple sub-area collaborative optimization problems, and the upper-level multi-sub-area collaborative optimization model is constructed. The distributed optimization algorithm is used to solve the global optimal solution, and the optimal cumulative number of vehicles in each traffic sub-area under collaborative control is obtained. The optimal cumulative number of vehicles in each traffic sub-area is used as the constraint of the lower-level single-sub-area optimization control model, and the timing of each intersection within the traffic sub-area is optimized to obtain the optimal timing plan for the controlled road network.

2. The multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture as claimed in claim 1, characterized in that: The vehicle flow data is the unit vehicle flow collected through pre-set data collection points and collection cycles; Based on the topological structure of the controlled road network and the collected traffic flow data, a road network model is established using traffic simulation modeling tools.

3. The multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture as claimed in claim 1, characterized in that: The construction of the weighted directed graph is specifically as follows: With intersections as nodes, there are bidirectional edges between two adjacent intersections, and the correlation between the intersections is used as the edge weight; The intersection correlation is calculated through a correlation model constructed based on the traffic density correlation, traffic volume correlation and signal cycle correlation.

4. The multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture as claimed in claim 1, characterized in that: The above method of obtaining several traffic sub-areas by dividing the area of ​​the weighted directed graph adopts the Louvain community detection algorithm, takes the maximum modularity increment as the goal, and merges nodes into communities until the communities where each node is located no longer change; The modularity is calculated based on the intersection correlation.

5. The multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture as claimed in claim 1, characterized in that: The multi-sub-area collaborative optimization model takes the maximization of the overall traffic efficiency of the road network as the optimization goal, and uses the Bregman alternating direction multiplier algorithm considering consensus variables to solve the optimal cumulative number of vehicles in each traffic sub-area.

6. The multi-traffic sub-area distributed collaborative optimization method based on a double-layer architecture as claimed in claim 1, characterized in that: The single sub-area optimization control model uses a feedback controller to perform boundary control to maintain the optimal cumulative number of vehicles under the coordinated control of each traffic sub-area, and uses a genetic-simulated annealing algorithm to optimize the timing of each intersection within the traffic sub-area.

7. A multi-traffic sub-area distributed collaborative optimization system based on a two-layer architecture, characterized in that: include: The model building module is configured to: build a road network model according to the topological structure of the controlled road network and the collected traffic flow data, and improve the traffic flow data in the road network model through simulation; The area division module is configured to: construct a weighted directed graph using a road network model, obtain a number of traffic sub-areas by dividing the weighted directed graph into regions, and draw a macro basic graph of each traffic sub-area; The collaborative optimization module is configured to: decompose the road network optimization problem into multiple sub-area collaborative optimization problems based on the macro basic graph of each traffic sub-area, construct an upper-level multi-sub-area collaborative optimization model, and use a distributed optimization algorithm to solve the global optimal solution to obtain the optimal cumulative number of vehicles in each traffic sub-area under collaborative control; The timing optimization module is configured to use the optimal cumulative number of vehicles in each traffic sub-area as a constraint of the lower-level single-sub-area optimization control model, optimize the timing of each intersection within the traffic sub-area, and obtain the optimal timing plan for the controlled road network.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the distributed collaborative optimization method for multiple traffic sub-areas based on a double-layer architecture as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the distributed collaborative optimization method of multiple traffic sub-areas based on a double-layer architecture as described in any one of claims 1-6 is implemented.

10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes the distributed collaborative optimization method for multiple traffic sub-areas based on a double-layer architecture as described in any one of claims 1-6.

Images