Large-scale road network parallel microscopic simulation method, device, equipment and medium

Abstract

The invention provides a large-scale road network parallel microcosmic simulation method, device, equipment and medium, and relates to the technical field of traffic simulation, and the method comprises the steps: carrying out the path flow distribution processing according to road network data and user heterogeneous data, carrying out the iterative calculation through combining with the sensitivity of different users to the average travel time, the travel time standard deviation and the pass cost, and obtaining a path flow distribution result; obtaining a path flow distribution scheme; matrix construction processing is carried out according to the path flow distribution scheme, and a Laplacian matrix is constructed by analyzing connectivity and structural characteristics of a traffic network; road network segmentation processing is carried out according to the eigenvalue and the eigenvector of the Laplacian matrix, and a road network segmentation result is obtained; and performing spatial decomposition processing on the road network segmentation result to obtain a spatial decomposition result, and defining vehicle driving information in a boundary region in the spatial decomposition result to obtain a simulation model. According to the invention, the problem that traffic decision makers cannot be assisted to make scientific decisions in time due to the fact that accurate prediction results cannot be fed back in time in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of traffic simulation technology, and in particular to a large-scale road network parallel microscopic simulation method, device, equipment and medium. Background Art

[0002] In the existing field of traffic technology, macro traffic simulation and micro traffic simulation are generally used. Macro traffic simulation uses relatively small amounts of data and simplified traffic models, and can process and analyze large-scale traffic networks, providing traffic planners and decision makers with a global overview of traffic conditions. However, while pursuing computational efficiency, macro traffic simulation has to sacrifice a certain degree of accuracy. Because it uses simplified traffic models and less data input, it is unable to simulate the specific behaviors and interactions of individual vehicles in detail, and is insufficient in capturing details in traffic flow; micro traffic simulation focuses on smaller granularity of traffic information, and can simulate the detailed behaviors of individual vehicles, but because the model is more complex, micro simulation requires a lot of computing resources and time to run the simulation. When faced with large-scale highway network simulation, the micro simulation method is difficult to cope with the computing load brought by the large-scale road network. Both the macro simulation method and the micro simulation method cannot timely feedback accurate prediction results, resulting in the problem of being unable to assist traffic decision makers in making scientific decisions in a timely manner.

[0003] Therefore, there is an urgent need for a large-scale road network parallel micro-simulation method, device, equipment and medium to solve the problem that accurate prediction results cannot be fed back in time, resulting in the inability to assist traffic decision makers to make scientific decisions in a timely manner. Summary of the invention

[0004] The purpose of the present invention is to provide a large-scale road network parallel microscopic simulation method, device, equipment and medium to improve the above problems. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows: In a first aspect, the present application provides a large-scale road network parallel microscopic simulation method, comprising: Acquire road network data and user heterogeneous data, wherein the road network data includes road network node data and road network link data, and the user heterogeneous data includes traffic demand data of each category of users; Performing path flow allocation processing according to the road network data and the user heterogeneous data, and performing iterative calculation in the traffic network by combining different users' sensitivity to average travel time, travel time standard deviation and toll, to obtain a path flow allocation scheme; Performing matrix construction processing according to the path flow distribution scheme, and constructing a Laplace matrix by analyzing the connectivity and structural characteristics of the transportation network; Performing road network segmentation processing according to the eigenvalues ​​and eigenvectors of the Laplace matrix, treating the highway road network as an undirected graph and performing image segmentation on the undirected graph to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks; The road network segmentation result is subjected to spatial decomposition processing to obtain a spatial decomposition result, and vehicle driving information is defined in a boundary area of ​​the spatial decomposition result to obtain a simulation model.

[0005] In a second aspect, the present application also provides a large-scale road network parallel microscopic simulation device, comprising: An acquisition model is used to acquire road network data and user heterogeneous data, wherein the road network data includes road network node data and road network link data, and the user heterogeneous data includes traffic demand data of each category of users; A first processing model is used to perform path flow allocation processing according to the road network data and the user heterogeneous data, and obtain a path flow allocation solution by iteratively calculating in the traffic network in combination with different users' sensitivity to average travel time, travel time standard deviation and toll; A second processing model is used to perform matrix construction processing according to the path flow allocation scheme, and to construct a Laplace matrix by analyzing the connectivity and structural characteristics of the transportation network; A third processing model is used to perform road network segmentation processing according to the eigenvalues ​​and eigenvectors of the Laplace matrix, by treating the highway road network as an undirected graph and performing image segmentation on the undirected graph to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks; The segmentation model is used to perform spatial decomposition processing on the road network segmentation result to obtain a spatial decomposition result, and define vehicle driving information in a boundary area of ​​the spatial decomposition result to obtain a simulation model.

[0006] In a third aspect, the present application also provides a large-scale road network parallel microscopic simulation device, comprising: Memory for storing computer programs; A processor is used to implement the steps of the large-scale road network parallel microscopic simulation method when executing the computer program.

[0007] In a fourth aspect, the present application further provides a medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned method based on parallel microscopic simulation of a large-scale road network are implemented.

[0008] The beneficial effects of the present invention are: The present invention utilizes user heterogeneous data and the eigenvalues ​​and eigenvectors of the Laplace matrix to simulate the user's tendency to choose different travel paths in the face of a series of uncertain factors such as weather, congestion, and accidents in the highway network, and realizes the reasonable allocation of the network path flow by comprehensively considering the sensitivity of different users to travel time. Furthermore, the present invention constructs the road network segmentation result based on the eigenvalues ​​and eigenvectors of the Laplace matrix, and considers the weights of the cut edges in the highway road network image and the weights of the edges in the segmented area to ensure the balance of each segmented area; in addition, the present invention also combines vehicle driving information for simulation in the road network space, thereby effectively ensuring the efficiency and accuracy of the simulation.

[0009] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 It is a schematic diagram of the process flow of the large-scale road network parallel microscopic simulation method described in an embodiment of the present invention; Figure 2 Schematic diagram of the structure of the large-scale road network parallel microscopic simulation equipment described in an embodiment of the present invention.

[0012] Markings in the figure: 800, large-scale road network parallel micro-simulation equipment; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. DETAILED DESCRIPTION

[0013] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0014] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0015] Embodiment 1: This embodiment provides a large-scale road network parallel microscopic simulation method.

[0016] See also Figure 1 , the figure shows that the method includes steps S1 to S5, including: S1: Acquire road network data and user heterogeneous data, wherein the road network data includes road network node data and road network link data, and the user heterogeneous data includes traffic demand data of each category of users; In this step, the traffic demand data of each category of users is the traffic demand between OD pairs under the user category; The OD pair is a combination of a specific travel start point and end point.

[0017] Since travel is affected by a series of factors such as weather, congestion, accidents, etc., there is uncertainty in travel time. Different users have different transportation needs. Therefore, the heterogeneity of users is considered in combination with actual traffic conditions.

[0018] S2: performing path flow allocation processing according to the road network data and the user heterogeneous data, and performing iterative calculation in the traffic network by combining different users' sensitivity to average travel time, travel time standard deviation and toll, to obtain a path flow allocation scheme; In order to clarify the specific calculation method of the path flow distribution scheme, step S2 includes S21 to S2, which are specifically: S21: performing initial path selection processing according to the road network data and the user heterogeneous data, and obtaining an initial path flow distribution scheme by distributing flow to the path with the shortest average travel time; In this step, the path with the shortest average travel time is selected based on the road network data and the user heterogeneous data. , allocating all the traffic of the OD pair to the path with the smallest average travel time On, that is, for , have And set OD pair User Class Use path set ; in, is the path with the shortest average travel time, For the user class, is a collection of user classes, For a single OD pair, is the set of all OD pairs, Indicates the highway user class In the road network OD Middle Path The traffic volume under For a specific user class OD The transportation demand between For a specific user class OD The set of paths used, A collection containing only one specific path.

[0019] S22: performing effective path generation processing according to the initial path flow allocation scheme, and generating an effective path set under three indicators of average travel time, travel time standard deviation and toll by using a three-target label algorithm; In order to clarify the specific method of obtaining the valid path set, step S22 includes S221 to S224, which are: S221: performing label initialization processing according to the path nodes in the initial path traffic distribution scheme, setting the initial source node label to the source node label set and performing access construction processing to obtain a label scan set; In this step, the node label set is , the source node is , the initial source node label is , the source node label set is , the label scan set is ; Initialize the labels of the path nodes to obtain the initial source node labels , the initial source node label Set to the source node label set, then the source node label set is For all other nodes, set their label sets to empty, indicating that they have not been visited yet, and initialize a label scan set, which is .

[0020] The label scan set is used to store label information of all nodes for subsequent calls.

[0021] S222: performing a lexicographical judgment on the three indicators of average travel time, travel time standard deviation and toll on the unprocessed node labels in the label scanning set to obtain nodes that meet the lexicographical requirements; In this step, the unprocessed nodes are , the unprocessed node label is ; Preferably, the dictionary order is to prioritize the minimum average travel time, then the minimum standard deviation of travel time, and finally the minimum toll. When the unprocessed node label satisfies the dictionary order, the unprocessed node label is Scan set from the tag to delete it.

[0022] S223: Expand the dictionary-compliant nodes according to the path with the shortest average travel time to obtain a next node label; In this step, we ensure that we continue to expand from the path with the shortest average travel time each time.

[0023] S224: Perform label dominance judgment based on the next node label until the next node label reaches a termination condition, perform set processing on the initial source node label and the next node label, and generate a valid path set.

[0024] In this step, the label dominance judgment is: if the next node label Unlabeled collection If any tag in is dominant, it will be added to the tag set and label scan set ; If the tag set There is a label in the next node Dominate Tags , then from the tag set And the label scan set Delete the check set ; The termination condition is: if the label scan set If it is empty, the algorithm is terminated, and the initial source node label and the next node label are processed as a set to generate a valid path set; otherwise, it returns to step S222.

[0025] The next node label is , Indicates link The starting node of Indicates link The cumulative average travel time on For Link The cumulative travel time variance on For Link The accumulated travel expenses on is a tag collection, To check the collection Are any labels in the next node label dominate; dominate; In step S22, unused paths between OD pairs are deleted during the iteration process, which can significantly reduce the memory space required by the algorithm when processing large amounts of data, while improving the calculation speed.

[0026] S23: performing iterative optimization processing of path flow according to the valid path set, and adjusting the path flow distribution by using a gradient projection method to obtain an optimized path flow distribution scheme; In order to clarify the specific method of obtaining the optimized path flow distribution solution, step S23 includes S231 to S234, which are specifically: S231: updating the average travel time and the standard deviation of the travel time according to the valid path set, and performing an extreme value update on the generalized cost of the current path flow distribution scheme based on the updated time result to obtain a cost update result; In this step, the generalized cost is: (1); In the above formula (1), Represents the user class In the road network OD Middle Path Under generalized cost, Indicates the path In OD The average travel time under Indicates the path In OD The standard deviation of travel time under Indicates the path In OD Tolls under Represents the user class The generalized time value of Represents the user class Reliability value; Among them, the path In OD The average travel time under is the average travel time, the path In OD The standard deviation of travel time under is the standard deviation of travel time, and the path In OD The tolls below are toll charges; The generalized cost difference is: (2); In the above formula (2), Indicates that in OD belongs to the user class Path and The generalized cost difference of the path is represents the highest generalized cost, represents the minimum generalized cost, Indicates In the first iteration, the highway user class In the road network OD Middle Path The traffic volume under Represents the user class The generalized time value of Represents the user class Reliability value; The valid path set is , at this time the valid path set is ; In the OD belongs to the user class Path and The path generalized cost difference is the extreme value of the generalized cost of the current path traffic allocation solution; Preferably, the generalized time value is: the opportunity income or loss of relevant entities that may be caused by time saving or delay during the travel process, and the reliability value is: the economic value related to the reliability of travel time and the degree of attention that travelers attach to the uncertainty of travel time.

[0027] S232: Summarizing the traffic flows of multiple paths according to the initial path flow allocation scheme to obtain the total traffic flow of paired paths; In this step, the total traffic flow on the paired paths is: (3); In the above formula (3), represents the total traffic flow on the paired paths, Indicates the highway user class in the initial state In the road network OD Middle Path The minimum traffic volume under Indicates the highway user class in the initial state In the road network OD Middle Path The highest traffic volume under.

[0028] S233: performing iterative optimization processing on the path flow according to the cost update result to obtain a vehicle flow update result; In this step, the vehicle flow update result is: (4); In the above formula (4), Indicates In the first iteration, the highway user category In the road network OD Middle Path The traffic volume under Indicates In the first iteration, the highway user class In the road network OD Middle Path The traffic volume under represents the step size, Indicates that in OD belongs to the user class Path and The generalized cost difference of the path is represents the total traffic flow on the paired paths; Among them, In the first iteration, the highway user category In the road network OD Middle Path The following traffic flow is the updated result of traffic flow.

[0029] S234: adjusting the path flow distribution according to the vehicle flow update result and the total vehicle flow of the paired paths to obtain an optimized path flow distribution solution.

[0030] In this step, after adjusting the path flow distribution according to the vehicle flow update result and the total vehicle flow of the paired paths, the average travel time and the travel time standard deviation are updated again to obtain an optimized path flow distribution solution.

[0031] S24: Perform convergence judgment processing according to the optimized path flow allocation scheme, and judge whether the current path flow allocation scheme has reached an equilibrium state by calculating the relative error between the generalized cost of the current path flow allocation scheme and the theoretically optimal generalized cost. If the relative error is less than a preset threshold, the final path flow allocation scheme is obtained.

[0032] In this step, the relative error is: (5); In the above formula (5), Relative error, Indicates the inner loop iteration execution The minimum cumulative generalized cost is For the user class, is a collection of user classes, For a specific user class OD The transportation demand between For a specific user class OD The set of paths used, Indicates the highway user class In the road network OD Middle Path The traffic volume under represents the generalized cost, Indicates the inner loop iteration execution Second, the path The flow rate on Represents the user class The generalized time value of Represents the user class Reliability value; Among them, the is a set of valid paths A path in .

[0033] S3: performing matrix construction processing according to the path flow allocation scheme, and constructing a Laplace matrix by analyzing the connectivity and structural characteristics of the transportation network; In order to clarify the specific method of obtaining the Laplace matrix, step S3 includes S31 to S33, which are specifically: S31: constructing the number of road network nodes according to the path flow distribution scheme, performing extreme value normalization processing on the path flow by setting the number of road network nodes, and generating link weights between nodes; In this step, the link weight between the nodes is: (6); In the above formula (6), Represents the normalized road node and nodes The flow value between Represents a road node and The distribution flow between represents the minimum value of the flow distribution between all road node pairs, It represents the maximum value of the flow distribution between all road node pairs; The normalized road node and nodes The flow value between nodes is used to represent and nodes The weight of the link between the nodes and nodes The weights of the links between them are the diagonal elements in the matrix, which are used to construct the adjacency matrix; The node represents a connection point of different sections of a highway or an end point of a section, and the link represents a road section connecting two nodes.

[0034] S32: constructing weights according to the number of nodes in the road network and the weights of links between the nodes to obtain a total weight of links connecting the nodes; In this step, the total weight of the links connecting the nodes is: (7); In the above formula (7), represents the total weight of the links connecting the nodes, Represents the normalized road node and nodes The flow value between Indicates the number of network nodes, The index variable representing the sum; The total weight of the links connecting the nodes is the elements on the diagonal of the matrix, which is used to construct the degree matrix.

[0035] S33: According to the analysis of the connectivity and structural characteristics of the transportation network, a matrix is ​​constructed for the link weights between the nodes and the total weights of the links connecting the nodes to obtain a Laplace matrix.

[0036] In this step, the adjacency matrix and the degree matrix are constructed according to the analysis of the connectivity and structural characteristics of the transportation network to obtain a Laplace matrix.

[0037] By analyzing the connectivity and structural characteristics of the transportation network, we comprehensively examine the various nodes, road sections and the connection relationships between them in the network to reveal the complex structure within the network and provide a basis for the road network segmentation process.

[0038] S4: performing road network segmentation processing according to the eigenvalues ​​and eigenvectors of the Laplace matrix, treating the highway road network as an undirected graph and performing image segmentation on the undirected graph to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks; In order to clarify the specific method of obtaining the road network segmentation result, step S4 includes S41 to S45, which are specifically: S41: constructing a set according to the eigenvalues ​​of the Laplace matrix and the number of links connecting the nodes of the eigenvector to obtain the sum of all node degrees of the road network node subset; In this step, the eigenvalues ​​of the Laplace matrix are the connectivity and structural characteristics of the traffic network graph, and the eigenvectors of the Laplace matrix are the coordinates of the nodes under the structural characteristics; According to the connectivity, structural characteristics and coordinates of the traffic network graph, the number of links connected to all nodes is aggregated to obtain the sum of all node degrees of the road network node subset, and the sum of all node degrees of the road network node subset is used for the sub-road network simulation computing load.

[0039] S42: constructing an optimal path solution model based on the sum of all node degrees of the road network node subset and the regional connection weights in the undirected graph; In this step, the path solution model is: (8); In the above formula (8), represents the normalized cutoff value, Represents an undirected graph A partitioned subset, represents the sub-network simulation computing load, represents the total frequency of vehicle information exchange between each sub-network, represents the partition subset, represents the sum index; The simulation computing load of the sub-road network is the sum of all node degrees of the road network node subset, and the total frequency of vehicle information exchange between the sub-road networks is the area in the undirected graph. With area Connection weights.

[0040] In this step, the optimal path solution model is: (9); In the above formula (9), Represents the total frequency of vehicle information exchange between each sub-network; represents the partition subset, represents the sum index; The simulation speed depends on the sub-network with the longest simulation time in each simulation step. The simulation operation efficiency of the entire network is improved by balancing the simulation computing load of each sub-network.

[0041] S43: calculating the eigenvalues ​​and eigenvectors of the Laplace matrix according to the optimal path solution model to obtain a clustering structure of the highway network image; In this step, the Lanczos iteration method is used to calculate the Laplace matrix. The smallest non-zero eigenvalue and eigenvector, the eigenvector corresponding to the eigenvalue, the eigenvectors are arranged together to form a Matrix, each row of which A vector in the dimensional space is clustered using the K-means algorithm to obtain the clustering structure of the highway road network image.

[0042] S44: performing spectral clustering optimization processing on the highway network image according to the segmented boundaries in the cluster structure, by setting the weights of the cut edges and the weights of the edges in the segmented regions, and obtaining an optimal segmentation scheme.

[0043] In this step, the minimized segmented boundary is selected, and the balance of each segmented area is ensured by considering the weight of the cut edge and the weight of the edge in the segmented area. The highway network image is spectrally clustered according to the balance. The category to which each row in the clustering result belongs is the original The category of the sub-network area to which the network node belongs. So far, the optimal segmentation solution is obtained.

[0044] S45: Segmenting the highway road network image based on the optimal segmentation scheme to obtain a road network segmentation result.

[0045] In this step, the road network segmentation result includes at least two sub-road networks.

[0046] S5: performing spatial decomposition processing on the road network segmentation result to obtain a spatial decomposition result, and defining vehicle driving information in a boundary area of ​​the spatial decomposition result to obtain a simulation model.

[0047] In this step, the spatial decomposition result is to divide the highway traffic network into multiple sub-areas. The vehicle number, type, instantaneous speed, section number and lane number are transmitted through the boundary area in the vehicle driving information to optimize the information transmission between the vehicles in each sub-road network area to ensure the maximum simulation efficiency and simulation accuracy. Furthermore, the boundary communication between adjacent sub-road networks is carried out through the parallel computing framework, and the simulation computing tasks of the sub-road network are then assigned to different processes of the computer. The MPI message passing protocol is used to realize communication and synchronization between different processes, thereby accelerating the execution of complex simulation computing tasks.

[0048] Embodiment 2: This embodiment provides a large-scale road network parallel microscopic simulation device, the device comprising: an acquisition model, used to acquire road network data and user heterogeneous data, the road network data comprising road network node data and road network link data, the user heterogeneous data comprising traffic demand data of each category of users; A first processing model is used to perform path flow allocation processing according to the road network data and the user heterogeneous data, and obtain a path flow allocation solution by iteratively calculating in the traffic network in combination with different users' sensitivity to average travel time, travel time standard deviation and toll; To clarify the specific method of obtaining the first processing model, the specific methods are: A first iteration unit is used to perform initial path selection processing according to the road network data and the user heterogeneous data, and obtain an initial path flow distribution plan by distributing the flow to the path with the shortest average travel time; The second iteration unit is used to perform effective path generation processing according to the initial path flow allocation scheme, and generate an effective path set under three indicators of average travel time, travel time standard deviation and toll by using a three-target label algorithm; A third iterative unit is used to perform iterative optimization processing of path flow according to the effective path set, and use a gradient projection method to adjust the path flow distribution to obtain an optimized path flow distribution plan; The fourth iterative unit is used to perform convergence judgment processing according to the optimized path flow allocation plan, and judge whether the current path flow allocation plan has reached an equilibrium state by calculating the relative error between the generalized cost of the current path flow allocation plan and the theoretically optimal generalized cost. If the relative error is less than a preset threshold, the final path flow allocation plan is obtained.

[0049] A second processing model is used to perform matrix construction processing according to the path flow allocation scheme, and to construct a Laplace matrix by analyzing the connectivity and structural characteristics of the transportation network; To clarify the specific method of obtaining the second processing model, the specific methods are: A first construction unit is used to construct the number of road network nodes according to the path flow distribution scheme, perform extreme value normalization processing on the path flow by setting the number of road network nodes, and generate link weights between nodes; A fourth construction unit is used to construct weights according to the number of nodes in the road network and the weights of links between the nodes to obtain a total weight of links connecting the nodes; The third construction unit is used to construct a matrix of the link weights between the nodes and the total weights of the links connected to the nodes according to the analysis of the connectivity and structural characteristics of the transportation network, so as to obtain a Laplace matrix.

[0050] A third processing model is used to perform road network segmentation processing according to the eigenvalues ​​and eigenvectors of the Laplace matrix, by treating the highway road network as an undirected graph and performing image segmentation on the undirected graph to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks; The segmentation model is used to perform spatial decomposition processing on the road network segmentation result to obtain a spatial decomposition result, and define vehicle driving information in a boundary area of ​​the spatial decomposition result to obtain a simulation model.

[0051] It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0052] Embodiment 3: Corresponding to the above method embodiment, a large-scale road network parallel microscopic simulation device is also provided in this embodiment. The large-scale road network parallel microscopic simulation device described below and the large-scale road network parallel microscopic simulation method described above can refer to each other.

[0053] Figure 2 FIG. 8 is a block diagram of a large-scale road network parallel microscopic simulation device 800 according to an exemplary embodiment. Figure 2 As shown, the large-scale road network parallel micro-simulation device 800 may include: a processor 801 and a memory 802. The large-scale road network parallel micro-simulation device 800 may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0054] The processor 801 is used to control the overall operation of the large-scale road network parallel micro-simulation device 800 to complete all or part of the steps in the large-scale road network parallel micro-simulation method described above. The memory 802 is used to store various types of data to support the operation of the large-scale road network parallel micro-simulation device 800, and these data may include, for example, instructions for any application or method operating on the large-scale road network parallel micro-simulation device 800, and application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone, which is used to receive external audio signals. The received audio signal may be further stored in the memory 802 or sent via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules can be keyboards, mice, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the large-scale road network parallel micro-simulation device 800 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 can include: Wi-Fi module, Bluetooth module, NFC module.

[0055] In an exemplary embodiment, the large-scale road network parallel microscopic simulation device 800 can be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, referred to as ASIC), digital signal processors (Digital Signal Processor, referred to as DSP), digital signal processing devices (Digital Signal Processing Device, referred to as DSPD), programmable logic devices (Programmable Logic Device, referred to as PLD), field programmable gate arrays (Field Programmable Gate Array, referred to as FPGA), controllers, microcontrollers, microprocessors or other electronic components, and are used to execute the above-mentioned large-scale road network parallel microscopic simulation method.

[0056] Embodiment 4: Corresponding to the above method embodiment, a readable storage medium is also provided in this embodiment. The readable storage medium described below and the large-scale road network parallel microscopic simulation method described above can refer to each other.

[0057] A readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the large-scale road network parallel microscopic simulation method of the above method embodiment.

[0058] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0059] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0060] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A large-scale road network parallel microscopic simulation method, characterized in that: include: Acquire road network data and user heterogeneous data, wherein the road network data includes road network node data and road network link data, and the user heterogeneous data includes traffic demand data of each category of users; Performing path flow allocation processing according to the road network data and the user heterogeneous data, and performing iterative calculation in the traffic network by combining different users' sensitivity to average travel time, travel time standard deviation and toll, to obtain a path flow allocation scheme; Performing matrix construction processing according to the path flow distribution scheme, and constructing a Laplace matrix by analyzing the connectivity and structural characteristics of the transportation network; Performing road network segmentation processing according to the eigenvalues ​​and eigenvectors of the Laplace matrix, treating the highway road network as an undirected graph and performing image segmentation on the undirected graph to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks; The road network segmentation result is subjected to spatial decomposition processing to obtain a spatial decomposition result, and vehicle driving information is defined in a boundary area of ​​the spatial decomposition result to obtain a simulation model.

2. The large-scale road network parallel microscopic simulation method according to claim 1, characterized in that: Path flow allocation processing is performed according to the road network data and the user heterogeneous data, and a path flow allocation scheme is obtained by iteratively calculating in the traffic network in combination with different users' sensitivity to average travel time, travel time standard deviation and toll, including: Performing initial path selection processing according to the road network data and the user heterogeneous data, and obtaining an initial path flow distribution scheme by distributing flow to the path with the shortest average travel time; Performing effective path generation processing according to the initial path flow allocation scheme, generating an effective path set under the three indicators of average travel time, travel time standard deviation and toll by using a three-target label algorithm; Performing iterative optimization processing of path flow according to the effective path set, and adjusting path flow distribution by using a gradient projection method to obtain an optimized path flow distribution scheme; A convergence judgment process is performed according to the optimized path flow allocation scheme. By calculating the relative error between the generalized cost of the current path flow allocation scheme and the theoretically optimal generalized cost, it is judged whether the current path flow allocation scheme has reached an equilibrium state. If the relative error is less than a preset threshold, the final path flow allocation scheme is obtained.

3. The large-scale road network parallel microscopic simulation method according to claim 2, characterized in that: According to the initial path flow allocation scheme, effective path generation processing is performed, and by using a three-target label algorithm, an effective path set is generated under the three indicators of average travel time, travel time standard deviation and toll, including: Perform label initialization processing according to the path nodes in the initial path traffic distribution scheme, obtain a label scan set by setting the initial source node label to the source node label set and performing access construction processing; Performing a lexicographical judgment on the three indicators of average travel time, travel time standard deviation and toll on the unprocessed node labels in the label scanning set, and obtaining nodes that meet the dictionary requirements; Expand the dictionary nodes according to the path with the shortest average travel time to obtain the next node label; A label dominance judgment is performed based on the next node label until the next node label reaches a termination condition, and the initial source node label and the next node label are processed in a set to generate a valid path set.

4. The large-scale road network parallel microscopic simulation method according to claim 1, characterized in that: According to the path flow allocation scheme, a matrix construction process is performed, and a Laplace matrix is ​​constructed by analyzing the connectivity and structural characteristics of the traffic network, including: According to the path flow distribution scheme, the number of road network nodes is constructed and processed, and the path flow is normalized by setting the number of road network nodes to generate link weights between nodes; Weights are constructed according to the number of nodes in the road network and the weights of links between the nodes to obtain the total weight of links connecting the nodes; According to the analysis of the connectivity and structural characteristics of the transportation network, a matrix is ​​constructed for the link weights between the nodes and the total weights of the links connected to the nodes to obtain a Laplace matrix.

5. The large-scale road network parallel microscopic simulation method according to claim 1, characterized in that: The road network segmentation processing is performed according to the eigenvalues ​​and eigenvectors of the Laplace matrix, and the highway road network is regarded as an undirected graph and the undirected graph is subjected to image segmentation to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks, including: A set is constructed based on the number of node-connected links of the eigenvalues ​​and eigenvectors of the Laplace matrix to obtain the sum of all node degrees of the road network node subset; The optimal path solution model is constructed based on the sum of all node degrees of the road network node subset and the regional connection weights in the undirected graph; Calculating the eigenvalues ​​and eigenvectors of the Laplace matrix according to the optimal path solution model to obtain a clustering structure of the highway network image; According to the segmented boundaries in the cluster structure, the spectral clustering optimization processing is performed on the highway network image by setting the weights of the cut edges and the weights of the edges in the segmented area to obtain the optimal segmentation scheme; The highway road network image is segmented based on the optimal segmentation scheme to obtain a road network segmentation result.

6. A large-scale road network parallel microscopic simulation device, characterized in that: include: An acquisition model is used to acquire road network data and user heterogeneous data, wherein the road network data includes road network node data and road network link data, and the user heterogeneous data includes traffic demand data of each category of users; A first processing model is used to perform path flow allocation processing according to the road network data and the user heterogeneous data, and obtain a path flow allocation solution by iteratively calculating in the traffic network in combination with different users' sensitivity to average travel time, travel time standard deviation and toll; A second processing model is used to perform matrix construction processing according to the path flow allocation scheme, and to construct a Laplace matrix by analyzing the connectivity and structural characteristics of the transportation network; A third processing model is used to perform road network segmentation processing according to the eigenvalues ​​and eigenvectors of the Laplace matrix, by treating the highway road network as an undirected graph and performing image segmentation on the undirected graph to obtain a road network segmentation result, wherein the road network segmentation result includes at least two sub-road networks; The segmentation model is used to perform spatial decomposition processing on the road network segmentation result to obtain a spatial decomposition result, and define vehicle driving information in a boundary area of ​​the spatial decomposition result to obtain a simulation model.

7. The large-scale road network parallel microscopic simulation device according to claim 6, characterized in that: The first processing model comprises: A first iteration unit is used to perform initial path selection processing according to the road network data and the user heterogeneous data, and obtain an initial path flow distribution plan by distributing the flow to the path with the shortest average travel time; The second iteration unit is used to perform effective path generation processing according to the initial path flow allocation scheme, and generate an effective path set under three indicators of average travel time, travel time standard deviation and toll by using a three-target label algorithm; A third iterative unit is used to perform iterative optimization processing of path flow according to the effective path set, and use a gradient projection method to adjust the path flow distribution to obtain an optimized path flow distribution plan; The fourth iterative unit is used to perform convergence judgment processing according to the optimized path flow allocation plan, and judge whether the current path flow allocation plan has reached an equilibrium state by calculating the relative error between the generalized cost of the current path flow allocation plan and the theoretically optimal generalized cost. If the relative error is less than a preset threshold, the final path flow allocation plan is obtained.

8. The large-scale road network parallel microscopic simulation device according to claim 6, characterized in that: The second processing model comprises: A first construction unit is used to construct the number of road network nodes according to the path flow distribution scheme, perform extreme value normalization processing on the path flow by setting the number of road network nodes, and generate link weights between nodes; A fourth construction unit is used to construct weights according to the number of nodes in the road network and the weights of links between the nodes to obtain a total weight of links connecting the nodes; The third construction unit is used to construct a matrix of the link weights between the nodes and the total weights of the links connected to the nodes according to the analysis of the connectivity and structural characteristics of the transportation network, so as to obtain a Laplace matrix.

9. A large-scale road network parallel microscopic simulation device, characterized in that: include: Memory for storing computer programs; A processor, used to implement the steps of the large-scale road network parallel microscopic simulation method as claimed in any one of claims 1 to 5 when executing the computer program.

10. A medium, characterized in that The medium stores a computer program, which, when executed by a processor, implements the steps of the large-scale road network parallel microscopic simulation method as claimed in any one of claims 1 to 5.

Images