Distributed micro-city traffic simulation device and method based on supercomputing platform

By designing a distributed micro-urban traffic simulation device on the supercomputing platform, using configuration nodes and computing nodes to divide and transmit urban traffic networks, the problem of insufficient computing power in large-scale urban traffic simulation is solved, and efficient simulation and Extensibility.

CN118761244BActive Publication Date: 2024-11-19TSINGHUA UNIVERSITY +1
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411245258.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-06
Publication Date
2024-11-19
Estimated Expiration
2044-09-06

AI Technical Summary

Technical Problem

The prior art is difficult to fully utilize the computing power of parallel computing platforms in large-scale urban traffic simulation, resulting in inefficient simulations and inability to scale up on a large scale.

Method used

Using a distributed micro-urban traffic simulation device based on supercomputing platform, the division and data transmission of urban traffic topology networks are realized by configuring nodes, computing nodes, urban traffic network division modules, data channel construction modules and dynamic task scheduling modules, and improving simulation Efficiency and scalability.

Benefits of technology

It improves the execution efficiency of large-scale micro-urban traffic simulation, ensures the scalability of the simulation scale, can effectively process large-scale urban traffic data and improves the analysis ability of simulation results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118761244B_ABST
    Figure CN118761244B_ABST
Patent Text Reader

Abstract

The present disclosure relates to the technical field of urban traffic simulation, and in particular to a distributed micro-city traffic simulation device and method based on a supercomputer platform. The device includes a configuration node, multiple computing nodes and a dynamic task scheduling module, the multiple computing nodes include a master computing node and multiple slave computing nodes, the configuration node is used to respond to the user to perform information configuration and task request reception, the master computing node is used to divide the computing tasks of each computing node for the task request, and each computing node respectively executes the corresponding computing tasks in the traffic simulation. By forming a distributed micro-city traffic simulation device based on a supercomputer platform, based on the large-scale cluster parallel computing capability provided by the supercomputer platform, by designing an extensible task division and an efficient data transmission scheme, while improving the execution efficiency of large-scale micro-city traffic simulation, the scalability of the simulation scale is guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the technical field of urban traffic simulation, and in particular to a distributed micro-urban traffic simulation device and method based on a supercomputing platform. Background Art

[0002] Micro urban traffic simulation is a method that uses computer technology and related models to simulate and analyze the operation of urban traffic systems and restore fine-grained urban traffic operation conditions, starting from urban traffic participants. However, for large-scale urban traffic simulation, how to ensure the efficiency and scalability of large-scale urban traffic simulation tasks based on the actual corresponding software framework and the full use of the computing power of the parallel computing platform has become an important issue that needs to be solved. Summary of the invention

[0003] In view of this, the present disclosure proposes a distributed micro-city traffic simulation device and method based on a supercomputing platform.

[0004] According to one aspect of the present disclosure, a distributed micro-city traffic simulation device based on a supercomputing platform is provided, the device comprising: a configuration node, a plurality of computing nodes and a dynamic task scheduling module, the plurality of computing nodes comprising a master computing node and a plurality of slave computing nodes, each of the computing nodes comprising a data loading module and a simulation computing engine, the master computing node further comprising an urban traffic network division module and a data channel construction module,

[0005] The configuration node is used to generate a task request in response to a user operation and send the task request to the main computing node, wherein the task request carries configuration information, and the configuration information includes a target number of computing nodes participating in the simulation and identification information indicating a target city to be simulated;

[0006] The urban traffic network partitioning module is used to obtain the urban data set of the target city from the shared database based on the identification information; construct an urban traffic topology network according to the urban data set, divide the urban traffic topology network according to the target number to form multiple topology subgraphs, and synchronize the multiple topology subgraphs to the corresponding computing nodes, and the number of the topology subgraphs is the same as the target number;

[0007] The data channel construction module is used to determine the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and control each computing node to construct a corresponding channel, wherein the channel is used for data transmission between the computing nodes;

[0008] The dynamic task scheduling module is used to determine that when a traffic unit moves to a boundary point of a current topological subgraph during a traffic simulation, the traffic unit that can enter the next topological subgraph from the current topological subgraph is stored in the traffic unit set as a traffic unit to be scheduled; form traffic unit information of each traffic unit to be scheduled and store it in the call-out data cache of the corresponding computing node; for the call-out task of each traffic unit to be scheduled, control the current computing node where each traffic unit to be scheduled is located to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel; for the call-in task of each traffic unit to be scheduled, control the current computing node to which each traffic unit to be scheduled is called to store the corresponding data in the call-out data cache sent from the associated node into the call-in data cache through the corresponding channel;

[0009] The dynamic task scheduling module is also used to generate a computing task corresponding to the traffic unit to be scheduled according to the transferred-in task of each traffic unit to be scheduled, and send the computing task to the computing task pool of the computing node that needs to be transferred into the traffic unit to be scheduled;

[0010] The simulation computing engine is used to obtain a computing task from a computing task pool of the computing node, obtain required basic information based on the computing task, and execute the computing task based on the basic information to obtain a computing result;

[0011] Each of the data loading modules is used to load data from a shared database according to the data requirements of the simulation computing engine for executing the computing task;

[0012] The basic information includes data from the task scheduling data cache of the computing node and data loaded from the shared database, and the data loaded from the shared database includes at least one of map data, traffic unit data, and environmental data.

[0013] In a possible implementation, the urban traffic network division module includes:

[0014] The urban topology network construction submodule is used to obtain the urban data set of the target city from the shared database based on the identification information, and parse the urban data set to obtain the road set and intersection set required for constructing the urban traffic topology network; taking the roads as edges and the intersections as nodes, constructing the urban traffic topology network corresponding to the target city according to the road set and the intersection set;

[0015] The traffic network area division submodule is used to determine the area information of the target area to be simulated for the target city based on the urban traffic topology network; divide the part of the urban traffic topology network corresponding to the target area based on the area information to obtain a plurality of topology subgraphs of the target number, and synchronize the plurality of topology subgraphs to the corresponding computing nodes;

[0016] The road set includes a plurality of road entries, and each of the road entries corresponds to a different road in the target city. Each of the road entries includes a road sign, road geographic information, and associated road information of the corresponding road. The road geographic information is represented by points and lines. The road geographic information is used to indicate at least one of the starting position information, the ending position information, and the route track of the corresponding road. The associated road information is used to indicate a road that is connected to the road corresponding to the road entry.

[0017] The intersection set includes a plurality of intersection entries, and each of the intersection entries corresponds to a different intersection in the target city, and each of the intersection entries includes an intersection identifier, intersection geographic information, and associated intersection information of the corresponding intersection; the intersection geographic information is used to indicate at least one of the geographic location of the intersection and the lane conditions in the intersection; the associated intersection information is used to indicate the roads connected to the intersection corresponding to the intersection entry, and the in-and-out relationship between each connected road and the intersection;

[0018] The target area is a part or all of the target city. The target area is determined according to the target simulation range of the target city. The area information includes at least one of the length, width and boundary information of the target area.

[0019] In a possible implementation, the urban topology network construction submodule is further used to delete isolated edges and isolated nodes in the urban traffic topology network to form a fully connected urban traffic topology network;

[0020] The isolated edges include edges that are not connected to nodes and other edges, and the isolated nodes include nodes that are not connected to other nodes.

[0021] In a possible implementation, a portion of the urban traffic topology network corresponding to the target area is divided based on the area information to obtain a plurality of topology subgraphs having the target number, including:

[0022] Determine a plurality of nodes in the target area in the urban traffic topology network based on the area information, and create a node longitude sequence and a node latitude sequence based on the plurality of nodes, wherein the node longitude sequence includes the plurality of nodes and the nodes are arranged in ascending order according to the longitude of the geographical location, and the node latitude sequence includes the plurality of nodes and the nodes are arranged in ascending order according to the latitude of the geographical location;

[0023] Taking the first node in the node longitude sequence as the starting point and according to the order of the nodes in the node longitude sequence, a node longitude chain is formed; dividing the node longitude chain equally according to the first quantity to form a plurality of node longitude sub-chains; according to the order of the nodes in the node latitude sequence, forming corresponding node latitude sub-chains for all nodes in each of the node longitude sub-chains, and dividing each of the node latitude sub-chains equally according to the second quantity to form a plurality of node latitude sub-chains, thereby completing node division;

[0024] For all node latitude sub-chains, traverse the roads, and according to the topological association of the roads, add the roads whose road-associated intersections are in the same node latitude sub-chain to the corresponding node latitude sub-chain;

[0025] Constructing an associated road allocation quantity matrix corresponding to each node latitude slave chain, traversing the remaining undivided roads based on a greedy strategy, and adding each of the remaining roads to the node latitude slave chain with the smallest matrix value of the current associated road allocation quantity matrix;

[0026] Based on the latitude sub-links of the nodes to which the roads are added, a corresponding topological sub-graph is formed to obtain a plurality of topological sub-graphs having the target number;

[0027] The product of the first number and the second number is equal to the target number.

[0028] In a possible implementation, based on the area information, a portion of the urban traffic topology network corresponding to the target area is divided to obtain a plurality of topology subgraphs having the target number, further comprising:

[0029] Taking the first node in the node latitude sequence as the starting point and according to the order of the nodes in the node latitude sequence, a node latitude chain is formed; dividing the node latitude chain equally according to the second quantity to form a plurality of node latitude sub-chains; according to the order of the nodes in the node longitude sequence, forming corresponding node longitude sub-chains for all nodes in each of the node latitude sub-chains, and dividing each of the node longitude sub-chains equally according to the first quantity to form a plurality of node longitude sub-chains, thereby completing node division;

[0030] For all node longitude sub-chains, traverse the roads, and according to the topological association of the roads, add the roads whose road-associated intersections are in the same sub-chain to the corresponding node longitude sub-chain;

[0031] Constructing an associated road allocation quantity matrix corresponding to each node longitude sub-chain, traversing the remaining undivided roads based on a greedy strategy, and adding each of the remaining roads to the node longitude sub-chain with the smallest matrix value of the current associated road allocation quantity matrix;

[0032] Based on the longitude chains of the nodes to which the roads are added, corresponding topological subgraphs are formed, and a plurality of topological subgraphs having the target number are obtained.

[0033] In a possible implementation, based on the multiple topology subgraphs, determining channels that need to be constructed between different computing nodes, and controlling each computing node to perform corresponding channel construction includes:

[0034] Based on each of the topological subgraphs, a plurality of boundary points between different topological subgraphs are determined, each of the boundary points being a connection point between a corresponding road and an intersection at a boundary of the topological subgraph;

[0035] Based on the associated intersection information of the intersection corresponding to each of the boundary points, determining the transmission direction of the data at each of the boundary points, the transmission direction including input and / or output;

[0036] Based on the road association relationship between different topological subgraphs and the transmission direction of each boundary point, a data channel matrix for all topological subgraphs is constructed, and the data channel matrix is ​​segmented to obtain a data channel construction set corresponding to each topological subgraph;

[0037] Based on each of the data channel construction sets, generate corresponding channel construction information, and send each channel construction information to the corresponding computing node;

[0038] Each of the computing nodes is further configured to construct a channel with other computing nodes based on the received channel construction information.

[0039] In a possible implementation, the data channel construction module is further used to set a transmission structure for data transmission between the computing nodes;

[0040] The transmission structure setting includes at least one of the following: setting a basic data structure of each type of data to be transmitted in each channel according to the type of data to be transmitted between the computing nodes; setting a composite data structure based on multiple basic data structures; setting the transmission type of the data transmitted in the channel according to the transmission content generated during the traffic simulation process;

[0041] Among them, the transmission type includes traffic unit transmission type, intersection transfer-in type, and transfer-out type, and the traffic unit transmission type includes traffic unit basic information type and traffic unit motion status information type.

[0042] In a possible implementation, each computing node is provided with a task scheduling data cache, and the task scheduling data cache includes: a scheduling basis cache for storing scheduling-related information of boundary points required for implementing task scheduling, an input data cache for storing input data, and an output data cache for storing output data, wherein the scheduling-related information of the boundary points includes at least one of traffic light information and road information;

[0043] Wherein, the dynamic task scheduling module includes:

[0044] A scheduling determination submodule is used to mark the traffic unit when it is determined that the traffic unit moves to the boundary point of the current topological subgraph, and store the marked traffic unit in the traffic unit set as a traffic unit to be scheduled if it is determined that the traffic unit can enter the next topological subgraph from the current topological subgraph according to the scheduling association information of the boundary point;

[0045] A dispatching task data organization submodule, used to form the traffic unit information of each traffic unit to be dispatched and store it in the corresponding call-out data cache, wherein the traffic unit information includes dispatching task information, calculation status information, transmission structure information and channel information, wherein the dispatching task information is used to represent the basic information of the traffic unit, the calculation status information is used to represent the calculation parameter information of the traffic unit, the transmission structure information is determined according to the target channel for transmitting the traffic unit information, and the channel information is used to indicate the required target channel;

[0046] The task scheduling submodule is used to control the current computing node where each of the traffic units to be scheduled is located to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel for the call-out task of each of the traffic units to be scheduled; and to control the current computing node to which each of the traffic units to be scheduled is called to store the corresponding data in the call-out data cache sent by the associated node into the call-in data cache through the corresponding channel for the call-in task of each of the traffic units to be scheduled;

[0047] A computing task setting submodule is used to generate computing tasks corresponding to the traffic units to be dispatched according to the incoming tasks of each traffic unit to be dispatched, and send the computing tasks to the computing task pool of the computing node that the traffic unit to be dispatched needs to be called into;

[0048] The associated nodes of a computing node are other computing nodes with which a channel exists.

[0049] According to another aspect of the present disclosure, a distributed micro-city traffic simulation method based on a supercomputing platform is provided, which is applied to the above-mentioned distributed micro-city traffic simulation device based on a supercomputing platform, and the method comprises:

[0050] The control configuration node generates a task request in response to a user operation, and sends the task request to the main computing node, wherein the task request carries configuration information, including a target number of computing nodes participating in the simulation and identification information indicating a target city to be simulated;

[0051] Control the city traffic network module in the main computing node to obtain the city data set of the target city from the shared database based on the identification information; construct an urban traffic topology network according to the city data set, divide the urban traffic topology network according to the target number to form a plurality of topology subgraphs, and synchronize the plurality of topology subgraphs to the corresponding computing nodes, wherein the number of the topology subgraphs is the same as the target number;

[0052] Controlling the data channel construction module in the main computing node to determine the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and controlling each computing node to construct a corresponding channel, wherein the channel is used for data transmission between the computing nodes;

[0053] When the dynamic task scheduling module is controlled to determine that the traffic unit moves to the current topological subgraph and moves to the boundary point during the traffic simulation process, the traffic unit that can enter the next topological subgraph from the current topological subgraph is stored in the traffic unit set as the traffic unit to be scheduled; the traffic unit information of each of the traffic units to be scheduled is formed and stored in the call-out data cache of the corresponding computing node; for the call-out task of each of the traffic units to be scheduled, the current computing node where each of the traffic units to be scheduled is located is controlled to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel; for the call-in task of each of the traffic units to be scheduled, the current computing node to which each of the traffic units to be scheduled is called is controlled to store the corresponding data in the call-out data cache sent from the associated node into the call-in data cache through the corresponding channel;

[0054] Control the dynamic task scheduling module to generate a computing task corresponding to the traffic unit to be scheduled according to the transferred-in task of each traffic unit to be scheduled, and send the computing task to the computing task pool of the computing node that the traffic unit to be scheduled needs to be transferred into;

[0055] Controlling the simulation computing engine in each computing node to obtain a computing task from a computing task pool of the computing node, obtain required basic information based on the computing task, and execute the computing task based on the basic information to obtain a computing result;

[0056] Controlling the data loading module in each computing node to load data from the shared database according to the data requirement of the simulation computing engine to execute the computing task;

[0057] The basic information includes data from the task scheduling data cache of the computing node and data loaded from the shared database, and the data loaded from the shared database includes at least one of map data, traffic unit data, and environmental data.

[0058] According to another aspect of the present disclosure, a distributed micro-city traffic simulation device based on a supercomputing platform is provided, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0059] According to another aspect of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions implement the above method when executed by a processor.

[0060] According to another aspect of the present disclosure, a computer program product is provided, including a computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in a processor of an electronic device, the processor in the electronic device executes the above method.

[0061] Through the distributed micro-city traffic simulation device and method based on a supercomputer platform provided by the embodiments of the present invention, a distributed micro-city traffic simulation device is formed based on a supercomputer platform. Based on the large-scale cluster parallel computing capability provided by the supercomputer platform, by designing scalable task division and efficient data transmission solutions, the execution efficiency of large-scale micro-city traffic simulation is improved while ensuring the scalability of the simulation scale.

[0062] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0064] Figure 1A block diagram of a distributed micro-city traffic simulation device based on a supercomputing platform according to an embodiment of the present disclosure is shown.

[0065] Figure 2 A schematic diagram showing the working process of an urban traffic network division module in a distributed micro-urban traffic simulation device based on a supercomputing platform according to an embodiment of the present disclosure is shown.

[0066] Figure 3 A schematic diagram showing the working process of a data channel construction module in a distributed micro-city traffic simulation device based on a supercomputing platform according to an embodiment of the present disclosure is shown.

[0067] Figure 4 A schematic diagram showing the working process of a dynamic task scheduling module in a distributed micro-city traffic simulation device based on a supercomputing platform according to an embodiment of the present disclosure is shown.

[0068] Figure 5 A schematic diagram of the simulation process of an example of a distributed micro-city traffic simulation device based on a supercomputing platform according to an embodiment of the present disclosure is shown.

[0069] Figure 6 It is a block diagram of a device 1900 for distributed micro-city traffic simulation based on a supercomputing platform according to an exemplary embodiment. DETAILED DESCRIPTION

[0070] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0071] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0072] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0073] Micro-urban traffic simulation is a method that uses computer technology and related models to simulate and analyze the operation of urban traffic systems and restore fine-grained urban traffic operation conditions, starting from urban traffic participants. It can help researchers, planners, and decision makers understand the dynamic behavior of traffic flow, evaluate the effectiveness of traffic management strategies, and predict future traffic development trends. Large-scale urban micro-traffic simulation not only poses challenges to the data processing and computational efficiency of simulation solutions, but also places demands on the scalability of simulation solutions due to the variability of the urban environment. To meet the above requirements, the software framework supporting urban micro-traffic simulation should be able to make full use of computing resources, while ensuring the correctness of the calculations, and improving the overall simulation efficiency and scalability.

[0074] However, the city-level traffic simulation system in the related technology mainly adopts a single-node-multi-threaded parallel computing solution, such as CityFlow, QarSUMO, etc. Compared with the city traffic simulation system that adopts a single-node-single-thread computing solution, such as SUMO, although the single-node-multi-threaded simulation solution can more effectively utilize the computing resources of a single computer, it still cannot meet the needs of large-scale urban traffic simulation. This is because the available resources of a single node are limited. Not only can the computing resources not meet the efficiency requirements of the simulation task, but some devices cannot even load large-scale urban traffic data, which greatly limits the downstream work such as analysis, iteration, model training, etc. based on the results of the simulation task execution. With the rapid development of cloud computing, heterogeneous computing power computing, and cluster computing and other technical services, large-scale parallel computing platforms represented by supercomputing platforms have provided a basic platform for various computing-intensive and resource-intensive tasks, and promoted the development of application fields including numerical computing, bioinformatics simulation, simulation acceleration, and deep learning.

[0075] Therefore, for large-scale urban traffic simulation, how to ensure the efficiency and scalability of large-scale urban traffic simulation tasks based on the actual corresponding software framework and the full use of the computing power of the parallel computing platform has become an important issue that needs to be solved. The following describes the technical solutions existing in the relevant technology:

[0076] Related technical solution 1: describes a distributed traffic simulation platform based on workflow technology and its simulation method. The platform architecture includes a user traffic platform layer: receiving user input, displaying simulation status and results; a regional simulation server group layer: consisting of multiple simulation servers, responsible for executing traffic simulation; a workflow server layer: managing the simulation process, including workflow engine, regional coordination, simulation model loading and decomposition, etc.; a database server layer: storing simulation workflow models, configuration data and urban traffic network data.

[0077] Related technical solution 2: A distributed traffic simulation system based on variable regional division and its simulation method are proposed, aiming to solve the load balancing problem in large city traffic simulation, improve the simulation speed, and make simulation-based traffic optimization scheduling and control possible. The system architecture includes a regional division server: responsible for providing road network files and regional division schemes for corresponding time periods; regional coordination server: generates a full road network model based on the road network file, and decomposes the model according to the regional division scheme in different time periods and distributes it to the simulation server; regional simulation server group: composed of multiple simulation servers, each server is responsible for traffic simulation in a specific area.

[0078] Related technical solution 3: describes a distributed traffic simulation platform based on service-oriented architecture (SOA) and its simulation method. The platform architecture includes a user interface layer: including simulation client and data maintenance client; coordinated simulation multi-agent layer: including interactive agent, road network generation agent, coordination agent, simulation result processing agent and regional simulation module; data service center layer: including data service center, covering Hibernate ORM submodule, data generation submodule and data dispatch submodule; simulation resource layer: using Oracle database to store road network data information and signal configuration information required for simulation.

[0079] Related technical solution 4: describes a distributed simulation system and simulation method for urban rail transit traction power supply system, which belongs to the field of urban rail transit technology. The system aims to simplify the simulation analysis process and improve the accuracy of simulation results to meet the technical challenges brought about by the rapid development of urban rail transit systems. The main steps of the simulation method include: determining the parameters required for simulation analysis, such as the running path, speed, output voltage and current of the power supply system of the rail vehicle; establishing an equivalent model, selecting the line, vehicle, power supply and traction station model equivalent to the actual operation; setting the initial value of the variable to be simulated, and setting the simulation time step; obtaining the power-on current of the vehicle model, and obtaining the calculated value of the variable to be simulated based on this and the initial value; verifying the steps to ensure the accuracy of the calculated value, and recalculating if there is an error.

[0080] Related technical solution 5: describes a distributed accelerated computing method, which aims to effectively accelerate the speed of traffic simulation at the city scale. The main steps of the method include: establishing a mapping from a road network to a graph model, simplifying the road network into intersections and road sections, and forming a graph model G(V,E); using a multi-layer k-way partitioning method to partition the graph so that the vertex weights of each subgraph are basically the same, while minimizing the edge intercept weights generated by the partitioning; mapping the partitioned graph to a loop network through mapping, and assigning it to different computing cores for parallel simulation; dividing a period of time into multiple time slices, and setting a vehicle generator and a traffic buffer for traffic settlement for the sections between different cores; the traffic buffer collects vehicle conditions to form a vehicle set, and uses the maximum likelihood method to fit the arrival time and destination distribution; the vehicle generator synchronizes the distribution in the traffic buffer, and uses the Monte Carlo method to generate vehicles, and puts them into the road section to which the traffic buffer belongs according to the time distribution and destination distribution.

[0081] Related technical solution 6: describes a GPU-based micro-traffic simulation method and device, which adaptively uses GPU in micro-traffic simulation to undertake the massive computing power required for fine-grained micro-traffic simulation at the city level, thereby greatly improving the efficiency of micro-traffic simulation. The main contents include: constructing a main chain-branch chain list for each lane of the urban road network; based on the main chain-branch chain list, generating a vehicle index for each vehicle in the urban road network; simulating the traffic operation of each vehicle according to the vehicle index and path planning. In addition, the patent also involves specific methods for simulating vehicle traffic operation, including the generation process of preset driving strategies and path planning.

[0082] However, the solutions to this problem in related technologies have the following limitations: single-node simulation computing solutions are limited by limited computing and storage resources. In the absence of a distributed guidance model and software framework, the simulation efficiency is low and cannot be expanded on a large scale. Existing distributed traffic simulation methods are designed for macro traffic simulation and cannot migrate to city-level micro traffic simulation.

[0083] In order to solve the above technical problems, the embodiments of the present invention provide a distributed micro-city traffic simulation device and method based on a supercomputing platform. The distributed micro-city traffic simulation device is formed based on a supercomputing cluster (also called a supercomputing platform). Based on the large-scale cluster parallel computing capability provided by the supercomputing platform, by designing scalable task division and efficient data transmission solutions, the execution efficiency of large-scale micro-city traffic simulation is improved while ensuring the scalability of the simulation scale.

[0084] like Figure 1As shown, a distributed micro-city traffic simulation device based on a supercomputing platform includes: a configuration node, multiple computing nodes and a dynamic task scheduling module, wherein the multiple computing nodes include a master computing node and multiple slave computing nodes, each of the computing nodes includes a data loading module and a simulation computing engine, and the master computing node also includes an urban traffic network division module and a data channel construction module. Among them, the configuration node and the multiple computing nodes can be implemented by computing nodes in the supercomputing platform.

[0085] The configuration node is used to generate a task request in response to a user operation and send the task request to the main computing node. The task request carries configuration information, which includes the target number of computing nodes participating in the simulation and identification information indicating the target city that needs to be simulated.

[0086] In this embodiment, the target number can be set according to the number of computing nodes in the supercomputing platform to ensure that each computing node of the supercomputing platform can participate in the entire simulation process, thereby improving the efficiency and speed of the simulation. The user can issue a task request through the operation configuration to realize the preliminary configuration of the micro-city traffic simulation.

[0087] The urban traffic network division module is used to obtain the urban data set of the target city from the shared database based on the identification information; construct an urban traffic topology network according to the urban data set, divide the urban traffic topology network according to the target number to form multiple topology subgraphs, and synchronize the multiple topology subgraphs to the corresponding computing nodes, and the number of the topology subgraphs is the same as the target number. While maintaining the integrity of the road structure, a concrete data communication model is provided for urban traffic simulation in a distributed environment to provide support for data transmission. In some embodiments, the shared database can be a database of a supercomputing platform.

[0088] like Figure 1 As shown, the urban traffic network division module includes an urban topology network construction submodule and a traffic network area division submodule. Figure 2 The working process of each submodule in the urban traffic network division module is schematically explained.

[0089] The city topology network construction submodule is used to obtain the city data set of the target city from the shared database based on the identification information, and parse the city data set to obtain the road set and intersection set required to construct the city traffic topology network (such as Figure 2 ); taking roads as edges and intersections as nodes, constructing an urban traffic topology network corresponding to the target city according to the road set and intersection set (such as Figure 2 Topological network construction in ).

[0090] In this embodiment, city data sets of different cities may be stored in the shared database.

[0091] In this embodiment, the road set includes a plurality of road entries and each of the road entries corresponds to a different road in the target city. Each of the road entries includes a road identification (such as an ID) of the corresponding road, road geographic information, and associated road information. The road geographic information is represented by points and lines, and the road geographic information is used to indicate at least one of the starting position information, the ending position information, and the route trajectory of the corresponding road, wherein the road geographic information may be a set of points and lines. The associated road information is used to indicate a road that is connected to the road corresponding to the road entry, that is, to indicate a road that has a connection attribute with the road corresponding to the road entry.

[0092] In this embodiment, the intersection set includes a plurality of intersection entries and each of the intersection entries corresponds to a different intersection in the target city, and each of the intersection entries includes an intersection identifier (such as an ID) of the corresponding intersection, intersection geographic information, and associated intersection information. The intersection geographic information is used to indicate at least one of the geographical location of the intersection and the lane conditions in the intersection. The associated intersection information is used to indicate the roads connected to the intersection corresponding to the intersection entry, and the in-and-out relationship between each connected road and the intersection.

[0093] In this embodiment, the associated network of urban roads can be constructed based on the associated road information contained in the road entries, on the basis of taking the roads in the target city as edges and the intersections as nodes (e.g. Figure 2 The intersection is embedded into the road association network with the associated intersection information contained in the intersection entry (such as Figure 2 The intersection associations shown in the figure constitute the urban traffic topology network.

[0094] In one possible implementation, Figure 2 As shown, the urban topology network construction submodule is also used for information screening, that is, it is also used to delete isolated edges and isolated nodes in the urban traffic topology network to form a fully connected urban traffic topology network; wherein, the isolated edges include edges that are not connected to nodes and other edges, and the isolated nodes include nodes that are not connected to other nodes.

[0095] The traffic network area division submodule is used to determine the area information of the target area to be simulated for the target city based on the urban traffic topology network (such as Figure 2The target area is a part or all of the target city, and the target area is determined according to the target simulation range of the target city. The area information includes at least one of the length, width and boundary information of the target area.

[0096] In this embodiment, the part corresponding to the target area in the urban traffic topology network is divided based on the area information to obtain a plurality of topology subgraphs having the target number. There are two implementation methods as follows:

[0097] Method 1:

[0098] Based on the regional information, multiple nodes in the urban traffic topology network that are in the target area are determined, and a node longitude sequence and a node latitude sequence are created based on the multiple nodes. The node longitude sequence includes the multiple nodes and the nodes are arranged from small to large according to the longitude of the geographical location. The node latitude sequence includes the multiple nodes and the nodes are arranged from small to large according to the latitude of the geographical location. Taking the first node in the node longitude sequence as the starting point and according to the order of the nodes in the node longitude sequence, a node longitude chain is formed; the node longitude chain is divided equally according to a first quantity to form a plurality of node longitude sub-chains; according to the order of the nodes in the node latitude sequence, corresponding node latitude sub-chains are formed for all nodes in each of the node longitude sub-chains, and each of the node latitude sub-chains is divided equally according to a second quantity to form a plurality of node latitude sub-chains, thereby completing the node division (the above process is implemented) Figure 2 For all node latitude sub-chains, traverse each of the roads, and according to the topological association of the roads, add the roads whose road-associated intersections are in the same node latitude sub-chain to the corresponding node latitude sub-chain, so as to achieve Figure 2 The road division based on topological association is shown. Construct the associated road allocation quantity matrix corresponding to each node latitude chain, traverse the remaining undivided roads based on the greedy strategy, and add the remaining roads to the node latitude chain with the smallest matrix value of the current associated road allocation quantity matrix (if the matrix values ​​are the same, the road can be randomly assigned to any node latitude chain), to achieve Figure 2 The road division based on the greedy strategy is shown. Based on the latitude chain of each node to which the road is added, a corresponding topological subgraph is formed, and a plurality of topological subgraphs having the target number are obtained.

[0099] The product of the first number and the second number is equal to the target number.

[0100] Taking the target number of 16 as an example, the node longitude chain is first divided equally into 4 node longitude sub-chains, and each node latitude sub-chain is divided equally into 4 node latitude sub-chains, and finally 16 node latitude sub-chains are obtained. Each node latitude sub-chain includes the same number of nodes.

[0101] Method 2:

[0102] Based on the regional information, multiple nodes in the urban traffic topology network that are in the target area are determined, and a node longitude sequence and a node latitude sequence are created based on the multiple nodes, wherein the node longitude sequence includes the multiple nodes and the nodes are arranged from small to large according to the longitude of the geographical location, and the node latitude sequence includes the multiple nodes and the nodes are arranged from small to large according to the latitude of the geographical location. Taking the first node in the node latitude sequence as the starting point and according to the order of the nodes in the node latitude sequence, a node latitude chain is formed; the node latitude chain is divided equally according to the second quantity to form multiple node latitude sub-chains; according to the order of the nodes in the node longitude sequence, corresponding node longitude sub-chains are formed for all nodes in each of the node latitude sub-chains, and each of the node longitude sub-chains is divided equally according to the first quantity to form multiple node longitude sub-chains, thereby completing the node division (the above process is implemented) Figure 2 For all node longitude sub-links, traverse each of the roads, and according to the topological association of the roads, add the roads whose road-associated intersections are in the same sub-link to the corresponding node longitude sub-link, to achieve Figure 2 The road division based on topological association is shown. Construct the associated road allocation quantity matrix corresponding to each node longitude chain, traverse the remaining undivided roads based on the greedy strategy, and add the remaining roads to the node longitude chain with the smallest matrix value in the current associated road allocation quantity matrix (if the matrix values ​​are the same, the roads can be randomly assigned to any node longitude chain), to achieve Figure 2 The road division based on the greedy strategy is shown. Based on the longitude chain of each node to which the road is added, a corresponding topological subgraph is formed, and a plurality of topological subgraphs having the target number are obtained.

[0103] The product of the first number and the second number is equal to the target number.

[0104] The data channel construction module is used to determine the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and control each computing node to construct a corresponding channel, and the channel is used for data transmission between the computing nodes.

[0105] In this embodiment, if Figure 3As shown, based on the multiple topological subgraphs, the channels that need to be constructed between different computing nodes are determined, and each computing node is controlled to perform corresponding channel construction, including the following steps: area boundary point identification, bidirectional data transmission channel identification, point-to-point data channel matching, construction of a data channel construction set, and each computing node performing data channel construction.

[0106] like Figure 3 As shown, the regional boundary point identification is first performed, that is, based on each of the topological subgraphs, a plurality of boundary points between different topological subgraphs are determined, each of the boundary points is the connection between a corresponding road and an intersection at the boundary of the topological subgraph; then the bidirectional data transmission channel identification is performed, that is, based on the associated intersection information of the intersection corresponding to each of the boundary points, the data transmission direction at each of the boundary points is determined, and the transmission direction includes calling in and / or calling out; then point-to-point data channel matching is performed, that is, based on the road association relationship between different topological subgraphs and the transmission direction of each of the boundary points, a data channel matrix for all topological subgraphs is constructed, and the data channel matrix is ​​segmented to obtain a data channel construction set corresponding to each of the topological subgraphs; based on each of the data channel construction sets, corresponding channel construction information is generated, and each of the channel construction information is sent to a corresponding computing node; each of the computing nodes is also used to construct a channel with other computing nodes based on the received channel construction information.

[0107] In this embodiment, in the point-to-point data channel matching, since there is a strong correlation between the topological subgraphs, that is, if one of the two topological subgraphs contains an incoming transmission channel, the other subgraph must contain an outgoing transmission channel, then by constructing a data channel matrix, the data channel information is symmetrically managed, and a complete data channel matrix is ​​obtained after all boundary points are processed. Through data segmentation, a data channel construction set suitable for each subgraph is obtained. The data channel construction set can include associated node information, channel identification information (i.e., target road or target intersection identification), channel type information, and channel quantity information. The associated nodes of a computing node are other computing nodes that have channels with the computing node.

[0108] In this embodiment, based on the high-speed data transmission channel provided by the supercomputing platform, the channel for boundary data transmission can be identified and constructed to provide support for dynamic data transmission and task scheduling, thereby ensuring the efficiency and reliability of data transmission while maintaining resource availability. Based on MPI or other available parallel computing frameworks, the dynamic task scheduling module provides a unified interface for data interaction to ensure the correctness and consistency of simulation calculations.

[0109] In this embodiment, after completing the synchronization of the data channel construction set, each computing node completes the data channel construction according to the received channel construction information. The channels constructed by each computing node include: an outgoing channel, an incoming channel, and a scheduling basis channel. Among them, for the outgoing channel, the content transmitted includes scheduling-related information and traffic unit information. Scheduling-related information refers to the data content that a traffic unit (such as a vehicle) relies on when it wants to drive out of the boundary of a topological subgraph and enter another topological subgraph (based on a simulation calculation model). For the incoming channel, the content transmitted is the traffic unit information. Due to the corresponding relationship between driving in and driving out, the entering party in the associated subgraph should maintain the scheduling-related information required by the other subgraph. In general, each computing node constructs a traffic unit transmission channel (for vehicle transmission) and an entry / exit basis transmission channel (for access judgment of vehicle migration) based on the information provided by the channel construction information.

[0110] In this embodiment, the traffic unit may be a participant in urban traffic, such as a vehicle, a pedestrian, and the like.

[0111] In one possible implementation, Figure 3 As shown, the data channel construction module is also used to set the transmission structure for data transmission between the computing nodes.

[0112] The transmission structure setting includes at least one of the following:

[0113] The basic data structure of each type of data to be transmitted in each channel is set according to the type of data to be transmitted between the computing nodes; a composite data structure is set based on multiple basic data structures; and the transmission type of the data transmitted in the channel is set according to the transmission content generated during the traffic simulation process.

[0114] Among them, the transmission type includes traffic unit transmission type, intersection transfer-in type, and transfer-out type, and the traffic unit transmission type includes traffic unit basic information type and traffic unit motion status information type.

[0115] In this implementation, different distributed transmission devices provide different types of support for basic data type selection. Taking MPI as an example, it provides a variety of basic data types including MPI_INT (integer type) and MPI_FLOAT (floating point type). When making a selection, you can use a basic data type selection scheme based on data intervals, that is, optimize the basic data type selection by identifying the dynamic range of the data to be transmitted. For example, for urban driving speed, the dynamic range is (-80,80) km / h, then you can choose MPI_CHAR (fixed-length data type) type instead of MPI_INT type (where MPI_CHAR type is 1 byte and MPI_INT type is 4 bytes) to optimize data transmission efficiency.

[0116] In this implementation, the composite data structure is defined by aggregating multiple selected basic data types to complete the definition of the composite data structure in a minimized and reusable solution.

[0117] In this implementation, for the transmission type, that is, the specific data transmission unit, the present disclosure defines the transmission type based on the actual transmission content generated during the traffic simulation process.

[0118] like Figure 1 As shown, the dynamic task scheduling module is used to store the traffic units that can enter the next topological subgraph from the current topological subgraph as traffic units to be scheduled in the traffic unit set when it is determined that the traffic unit moves to the current topological subgraph and moves to the boundary point; form the traffic unit information of each traffic unit to be scheduled and store it in the call-out data cache of the corresponding computing node; for the call-out task of each traffic unit to be scheduled, control the current computing node where each traffic unit to be scheduled is located to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel; for the call-in task of each traffic unit to be scheduled, control the current computing node where each traffic unit to be scheduled is called in, and store the corresponding data in the call-out data cache sent from the associated node into the call-in data cache through the corresponding channel.

[0119] The dynamic task scheduling module is also used to generate computing tasks corresponding to the traffic units to be scheduled according to the incoming tasks of each traffic unit to be scheduled, and send the computing tasks to the computing task pool of the computing node that the traffic unit to be scheduled needs to be called into.

[0120] In one possible implementation, Figure 4 As shown, each computing node is provided with a task scheduling data cache, and the task scheduling data cache includes: a scheduling basis cache for storing scheduling-related information of boundary points required for implementing task scheduling, an input data cache for storing input data, and an output data cache for storing output data, and the scheduling-related information of the boundary points includes at least one of traffic light information and road information. Traffic light information may refer to the red, yellow, and green change information of the relevant traffic light, so as to determine whether the traffic unit can be scheduled. Road information may include movement information (speed, acceleration, position, whether it is driving slowly, whether it is congested, etc.) of traffic units traveling on each road.

[0121] In some embodiments, the scheduling basis cache includes a call-out basis cache for storing the call-out basis in the scheduling association information and a call-in basis cache for storing the call-in basis in the scheduling association information. At the beginning of each simulation step, the computing node first needs to send the information in the call-in basis cache to the associated node, and at the same time receive the information from the associated node and put it into the call-out basis cache. The call-in data channel set and the call-out data channel set respectively include multiple groups of call-in data caches and call-out data caches, each group of call-in data caches corresponds to a data transmission channel (the call-in data cache corresponds to the incoming data channel, and the call-out data cache corresponds to the outgoing data channel), which is used to maintain the basic information of task scheduling between computing nodes.

[0122] Among them, Figure 4 As shown, the dynamic task scheduling module includes: a scheduling determination submodule, a scheduling task data organization submodule, a task scheduling submodule and a computing task setting submodule.

[0123] The scheduling determination submodule is used to mark the traffic unit when it is determined that the traffic unit moves to the boundary point of the current topological subgraph. If it is determined based on the scheduling association information of the boundary point that the traffic unit can enter the next topological subgraph from the current topological subgraph, the marked traffic unit is stored in the traffic unit set as a traffic unit to be scheduled.

[0124] In this embodiment, during the simulation process, when the scheduling determination submodule determines that the traffic unit has traveled to a boundary point, it obtains the call-out basis of the current boundary point from the call-out basis cache based on the information of the current boundary point, and determines whether the topological subgraph corresponding to the associated node can be called in at the current simulation moment. If the judgment result is that it can be called in, the identification information of the current traffic unit is recorded and the traffic unit is marked as needing to be scheduled as a traffic unit to be scheduled; otherwise, it is determined that the current traffic unit does not need to be scheduled.

[0125] In this embodiment, the traffic unit may refer to an object that can travel on the road in urban traffic, such as a vehicle, a pedestrian, etc., and the present disclosure does not impose any limitation on this.

[0126] The scheduling task data organization submodule is used to form the traffic unit information of each traffic unit to be scheduled and store it in the corresponding call-out data cache. The traffic unit information includes scheduling task information, calculation status information, transmission structure information and channel identification. The scheduling task information is used to represent the basic information of the traffic unit, the calculation status information is used to represent the calculation parameter information of the traffic unit, the transmission structure information is determined according to the target channel for transmitting the traffic unit information, and the channel information is used to indicate the required target channel.

[0127] In this embodiment, based on the set of traffic units determined by the scheduling, the scheduling task data organization submodule organizes the relevant information of each traffic unit to be scheduled into a standard data format and stores it in the corresponding called-out data cache according to the format requirements of the data called-out in the task scheduling data cache. Among them, the scheduling task information may refer to the basic information of the traffic unit to be scheduled, such as the ID information of the traffic unit to be scheduled, the physical information of the traffic unit to be scheduled, etc. The calculation parameter information described by the calculation state information may be the speed information, acceleration information and other information related to the motion state of the traffic unit to be scheduled. The transmission structure information is directly associated with the transmission structure definition defined by the data channel, that is, the format adjustment and organization of the traffic unit information to be transmitted. The channel information may be the channel identification ID of the called-out data cache, so as to transmit the traffic unit information to the called-out data cache through the channel corresponding to the channel identification.

[0128] The task scheduling submodule is used to control the current computing node where each of the traffic units to be scheduled is located to send the corresponding data in the call-out data cache of the current computing node (such as traffic unit information in a standard format, the number of traffic units to be called out, etc.) to the call-in data cache of the associated node through the channel for the call-out task of each of the traffic units to be scheduled; and control the current computing node where each of the traffic units to be scheduled is called in to store the corresponding data in the call-out data cache sent by the associated node from the associated node into the call-in data cache through the corresponding channel for the call-in task of each of the traffic units to be scheduled.

[0129] The computing task setting submodule is used to generate computing tasks corresponding to the traffic units to be dispatched according to the incoming tasks of each traffic unit to be dispatched, and send the computing tasks to the computing task pool of the computing node that the traffic unit to be dispatched needs to be called into.

[0130] In this embodiment, after receiving the incoming task of the traffic unit to be dispatched, the computing task setting submodule needs to generate the corresponding computing task and initialize and restore the computing state of the task. The computing task setting submodule can obtain or receive information from the incoming data cache of the computing node corresponding to the topological subgraph to be called into by the traffic unit to be dispatched, and then parse the information into a standard task format and load it into the computing task pool of the computing node. In addition, the computing task setting submodule completes the reloading of the computing state and inserts the computing target into the target intersection or road corresponding to the traffic unit to be dispatched based on the state information provided by the dispatching task.

[0131] The simulation computing engine is used to obtain computing tasks from a computing task pool of the computing node, obtain required basic information based on the computing tasks, and execute the computing tasks based on the basic information to obtain computing results.

[0132] In this embodiment, the simulation computing engine is responsible for the task calculation within the computing node. For the micro-city traffic simulation, the simulation computing engine may be provided with driving and following models, signal control models, environmental models, etc. The simulation computing engine may use these models to perform task calculations.

[0133] Each of the data loading modules is used to load data from a shared database according to the data requirements of the simulation computing engine for executing the computing task.

[0134] The basic information includes data from the task scheduling data cache of the computing node and data loaded from the shared database, and the data loaded from the shared database includes at least one of map data, traffic unit data, and environmental data.

[0135] In this embodiment, the channel created by the master computing node also includes a start / end channel between the master computing node and each slave computing node. After determining that each computing node has completed the channel creation, the master computing node is also used to send a start simulation signal to each slave computing node through the start / end channel, and each computing node starts the corresponding simulation calculation work. In addition, when the master computing node determines that the simulation end condition is met (such as the simulation duration specified by the user configuration has been reached), the end information is sent to each slave computing node through the start / end channel; each slave computing node ends the simulation. Then the master computing node generates a task result for the task request based on the calculation results of each computing node and returns it to the configuration node or stores it in a shared database. The task result may include the total simulation duration, traffic flow distribution information, urban traffic congestion status, vehicle speed distribution information, etc.

[0136] For example, a user wants to use the above-mentioned distributed micro-city traffic simulation device based on the supercomputing platform to conduct a micro-city traffic simulation within the Fifth Ring Road of xx City involving 1 million traffic participants on a scale of 9 nodes on the supercomputing platform.

[0137] like Figure 5 As shown, first, the user will obtain the urban road network data within the Fifth Ring Road of xx City, urban intersection data, urban AOI (Area of ​​Interest) data, and traffic unit data in JSON format in the shared database of the supercomputing platform. The main contents of these urban data are shown in Table 1 below:

[0138] Table 1 Example of data stored in a shared database

[0139]

[0140] The user configures information on the configuration node of the supercomputing platform. The configuration information includes the following contents as shown in Table 2:

[0141] Table 2 Configuration information content example

[0142]

[0143] The supercomputing platform allocates 9 nodes to the current simulation task based on the user configuration. The main computing node receives the task request for traffic simulation from the user, reads the configuration information and loads the data; the slave computing node waits for the information from the main computing node.

[0144] The urban traffic network partitioning module in the master computing node uses the minimum cut algorithm to partition the constructed urban traffic topology network and obtains the topology subgraphs of 9 sub-areas. The master computing node synchronizes the partitioning results to each slave computing node so that each slave computing node and the master computing node can determine the topology subgraph they need to calculate. Each slave computing node loads the corresponding data from the shared database (these data include the urban road network, urban AOI, urban intersection, and traffic unit in the area corresponding to the corresponding topology subgraph).

[0145] Then, the data channel construction module in the main computing node determines the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and controls each of the computing nodes to construct the corresponding channels.

[0146] The channel created by the master computing node also includes the MPI_COMM_GROUP association group (used to activate data transmission and dynamic scheduling). After the master computing node determines that each computing node has completed channel creation, it sends a start simulation signal through MPI_COMM_GROUP, and each computing node starts the corresponding simulation calculation work.

[0147] During the operation of the urban traffic simulation, when cross-node traffic unit flows occur, each computing node sends the data to be transmitted to the corresponding computing node through the previously created channel. The dynamic task scheduling module implements data transmission control and task scheduling in the manner described above. Each computing node obtains computing tasks from its own computing task pool and executes computing tasks to obtain computing results.

[0148] When the master computing node determines that the simulation end conditions are met (such as the simulation duration specified by the user configuration has been reached), it sends the end information to each slave computing node based on MPI_COMM_GROUP; each slave computing node ends the simulation. The master computing node then generates task results based on the calculation results of each computing node and transmits them back to the configuration node or stores them in the shared database. The task results include the total simulation duration, traffic flow distribution information, urban traffic congestion status, vehicle speed distribution information, etc.

[0149] The embodiment of the present disclosure also provides a distributed micro-city traffic simulation method based on a supercomputing platform, which is applied to a distributed micro-city traffic simulation device based on a supercomputing platform. The method includes:

[0150] The control configuration node generates a task request in response to a user operation, and sends the task request to the main computing node, wherein the task request carries configuration information, including a target number of computing nodes participating in the simulation and identification information indicating a target city to be simulated;

[0151] Control the city traffic network module in the main computing node to obtain the city data set of the target city from the shared database based on the identification information; construct an urban traffic topology network according to the city data set, divide the urban traffic topology network according to the target number to form a plurality of topology subgraphs, and synchronize the plurality of topology subgraphs to the corresponding computing nodes, wherein the number of the topology subgraphs is the same as the target number;

[0152] Controlling the data channel construction module in the main computing node to determine the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and controlling each computing node to construct a corresponding channel, wherein the channel is used for data transmission between the computing nodes;

[0153] When the dynamic task scheduling module is controlled to determine that the traffic unit moves to the current topological subgraph and moves to the boundary point during the traffic simulation process, the traffic unit that can enter the next topological subgraph from the current topological subgraph is stored in the traffic unit set as the traffic unit to be scheduled; the traffic unit information of each of the traffic units to be scheduled is formed and stored in the call-out data cache of the corresponding computing node; for the call-out task of each of the traffic units to be scheduled, the current computing node where each of the traffic units to be scheduled is located is controlled to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel; for the call-in task of each of the traffic units to be scheduled, the current computing node to which each of the traffic units to be scheduled is called is controlled to store the corresponding data in the call-out data cache sent from the associated node into the call-in data cache through the corresponding channel;

[0154] Control the dynamic task scheduling module to generate a computing task corresponding to the traffic unit to be scheduled according to the transferred-in task of each traffic unit to be scheduled, and send the computing task to the computing task pool of the computing node that the traffic unit to be scheduled needs to be transferred into;

[0155] Controlling the simulation computing engine in each computing node to obtain a computing task from a computing task pool of the computing node, obtain required basic information based on the computing task, and execute the computing task based on the basic information to obtain a computing result;

[0156] The data loading module in each computing node is controlled to load data from the shared database according to the data requirement of the simulation computing engine to execute the computing task.

[0157] The basic information includes data from the task scheduling data cache of the computing node and data loaded from the shared database, and the data loaded from the shared database includes at least one of map data, traffic unit data, and environmental data.

[0158] It should be noted that, although the above embodiments are used as examples to introduce the distributed micro-city traffic simulation device and method based on the supercomputing platform, those skilled in the art will appreciate that the present disclosure should not be limited thereto. In fact, the user can flexibly set each step and module according to personal preferences and / or actual application scenarios, as long as it complies with the technical solution of the present disclosure.

[0159] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the methods described in the above method embodiments, and their specific implementations can be referenced to each other. For the sake of brevity, they will not be repeated here.

[0160] The embodiment of the present disclosure also provides a computer-readable storage medium on which computer program instructions are stored, and the computer program instructions implement the above method when executed by a processor. The computer-readable storage medium can be a volatile or non-volatile computer-readable storage medium.

[0161] An embodiment of the present disclosure further proposes an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0162] The embodiments of the present disclosure also provide a computer program product, including a computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in a processor of an electronic device, the processor in the electronic device executes the above method.

[0163] Figure 6 1 is a block diagram of a device 1900 for distributed microscopic urban traffic simulation based on a supercomputing platform according to an exemplary embodiment. For example, the device 1900 can be provided as a server or a terminal device. Figure 6, the apparatus 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932 for storing instructions, such as an application, that can be executed by the processing component 1922. The application stored in the memory 1932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0164] The device 1900 may also include a power supply component 1926 configured to perform power management of the device 1900, a wired or wireless network interface 1950 configured to connect the device 1900 to a network, and an input / output interface 1958 (I / O interface). The device 1900 may operate based on an operating system stored in the memory 1932, such as Windows Server 2000. TM , MacOS X TM , Unix TM , Linux TM , FreeBSD TM or similar.

[0165] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions, which can be executed by the processing component 1922 of the device 1900 to perform the above method.

[0166] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0167] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the above. The computer-readable storage medium used herein is not to be interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0168] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0169] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect through the Internet). In some embodiments, by using the state information of the computer-readable program instructions to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit may execute the computer-readable program instructions, thereby implementing various aspects of the present disclosure.

[0170] Various aspects of the present disclosure are described herein with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0171] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0172] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0173] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0174] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A distributed micro-city traffic simulation device based on a supercomputing platform, characterized in that: The device comprises: a configuration node, a plurality of computing nodes and a dynamic task scheduling module, wherein the plurality of computing nodes comprises a master computing node and a plurality of slave computing nodes, each of the computing nodes comprises a data loading module and a simulation computing engine, and the master computing node further comprises a city traffic network division module and a data channel construction module. The configuration node is used to generate a task request in response to a user operation and send the task request to the main computing node, wherein the task request carries configuration information, and the configuration information includes a target number of computing nodes participating in the simulation and identification information indicating a target city to be simulated; The urban traffic network partitioning module is used to obtain the urban data set of the target city from the shared database based on the identification information; construct an urban traffic topology network according to the urban data set, divide the urban traffic topology network according to the target number to form multiple topology subgraphs, and synchronize the multiple topology subgraphs to the corresponding computing nodes, and the number of the topology subgraphs is the same as the target number; The data channel construction module is used to determine the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and control each computing node to construct a corresponding channel, wherein the channel is used for data transmission between the computing nodes; The dynamic task scheduling module is used to determine that when a traffic unit moves to a boundary point of a current topological subgraph during a traffic simulation, the traffic unit that can enter the next topological subgraph from the current topological subgraph is stored in the traffic unit set as a traffic unit to be scheduled; form traffic unit information of each traffic unit to be scheduled and store it in the call-out data cache of the corresponding computing node; for the call-out task of each traffic unit to be scheduled, control the current computing node where each traffic unit to be scheduled is located to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel; for the call-in task of each traffic unit to be scheduled, control the current computing node to which each traffic unit to be scheduled is called to store the corresponding data in the call-out data cache sent from the associated node into the call-in data cache through the corresponding channel; The dynamic task scheduling module is also used to generate a computing task corresponding to the traffic unit to be scheduled according to the transferred-in task of each traffic unit to be scheduled, and send the computing task to the computing task pool of the computing node that needs to be transferred into the traffic unit to be scheduled; The simulation computing engine is used to obtain a computing task from a computing task pool of the computing node, obtain required basic information based on the computing task, and execute the computing task based on the basic information to obtain a computing result; Each of the data loading modules is used to load data from a shared database according to the data requirements of the simulation computing engine for executing the computing task; The basic information includes data from the task scheduling data cache of the computing node and data loaded from the shared database, and the data loaded from the shared database includes at least one of map data, traffic unit data, and environmental data.

2. The device according to claim 1, characterized in that The urban traffic network division module includes: The urban topology network construction submodule is used to obtain the urban data set of the target city from the shared database based on the identification information, and parse the urban data set to obtain the road set and intersection set required for constructing the urban traffic topology network; taking the roads as edges and the intersections as nodes, constructing the urban traffic topology network corresponding to the target city according to the road set and the intersection set; The traffic network area division submodule is used to determine the area information of the target area to be simulated for the target city based on the urban traffic topology network; divide the part of the urban traffic topology network corresponding to the target area based on the area information to obtain a plurality of topology subgraphs of the target number, and synchronize the plurality of topology subgraphs to the corresponding computing nodes; The road set includes a plurality of road entries, and each of the road entries corresponds to a different road in the target city. Each of the road entries includes a road sign, road geographic information, and associated road information of the corresponding road. The road geographic information is represented by points and lines. The road geographic information is used to indicate at least one of the starting position information, the ending position information, and the route track of the corresponding road. The associated road information is used to indicate a road that is connected to the road corresponding to the road entry. The intersection set includes a plurality of intersection entries, and each of the intersection entries corresponds to a different intersection in the target city, and each of the intersection entries includes an intersection identifier, intersection geographic information, and associated intersection information of the corresponding intersection; the intersection geographic information is used to indicate at least one of the geographic location of the intersection and the lane conditions in the intersection; the associated intersection information is used to indicate the roads connected to the intersection corresponding to the intersection entry, and the in-and-out relationship between each connected road and the intersection; The target area is a part or all of the target city. The target area is determined according to the target simulation range of the target city. The area information includes at least one of the length, width and boundary information of the target area.

3. The device according to claim 2, characterized in that The urban topology network construction submodule is further used to delete isolated edges and isolated nodes in the urban traffic topology network to form a fully connected urban traffic topology network; The isolated edges include edges that are not connected to nodes and other edges, and the isolated nodes include nodes that are not connected to other nodes.

4. The device according to claim 2, characterized in that Based on the area information, a portion of the urban traffic topology network corresponding to the target area is divided to obtain a plurality of topology subgraphs having the target number, including: Determine a plurality of nodes in the target area in the urban traffic topology network based on the area information, and create a node longitude sequence and a node latitude sequence based on the plurality of nodes, wherein the node longitude sequence includes the plurality of nodes and the nodes are arranged in ascending order according to the longitude of the geographical location, and the node latitude sequence includes the plurality of nodes and the nodes are arranged in ascending order according to the latitude of the geographical location; Taking the first node in the node longitude sequence as the starting point and according to the order of the nodes in the node longitude sequence, a node longitude chain is formed; dividing the node longitude chain equally according to the first quantity to form a plurality of node longitude sub-chains; according to the order of the nodes in the node latitude sequence, forming corresponding node latitude sub-chains for all nodes in each of the node longitude sub-chains, and dividing each of the node latitude sub-chains equally according to the second quantity to form a plurality of node latitude sub-chains, thereby completing node division; For all node latitude sub-chains, traverse the roads, and according to the topological association of the roads, add the roads whose road-associated intersections are in the same node latitude sub-chain to the corresponding node latitude sub-chain; Constructing an associated road allocation quantity matrix corresponding to each node latitude slave chain, traversing the remaining undivided roads based on a greedy strategy, and adding each of the remaining roads to the node latitude slave chain with the smallest matrix value of the current associated road allocation quantity matrix; Based on the latitude sub-links of the nodes to which the roads are added, a corresponding topological sub-graph is formed to obtain a plurality of topological sub-graphs having the target number; The product of the first number and the second number is equal to the target number.

5. The device according to claim 4, characterized in that The method further includes dividing the part of the urban traffic topology network corresponding to the target area based on the area information to obtain a plurality of topology subgraphs having the target number. Taking the first node in the node latitude sequence as the starting point and according to the order of the nodes in the node latitude sequence, a node latitude chain is formed; dividing the node latitude chain equally according to the second quantity to form a plurality of node latitude sub-chains; according to the order of the nodes in the node longitude sequence, forming corresponding node longitude sub-chains for all nodes in each of the node latitude sub-chains, and dividing each of the node longitude sub-chains equally according to the first quantity to form a plurality of node longitude sub-chains, thereby completing node division; For all node longitude sub-chains, traverse the roads, and according to the topological association of the roads, add the roads whose road-associated intersections are in the same sub-chain to the corresponding node longitude sub-chain; Constructing an associated road allocation quantity matrix corresponding to each node longitude sub-chain, traversing the remaining undivided roads based on a greedy strategy, and adding each of the remaining roads to the node longitude sub-chain with the smallest matrix value of the current associated road allocation quantity matrix; Based on the longitude chains of the nodes to which the roads are added, corresponding topological subgraphs are formed, and a plurality of topological subgraphs having the target number are obtained.

6. The device according to any one of claims 1 to 5, characterized in that: Based on the multiple topological subgraphs, channels that need to be constructed between different computing nodes are determined, and each computing node is controlled to perform corresponding channel construction, including: Based on each of the topological subgraphs, a plurality of boundary points between different topological subgraphs are determined, each of the boundary points being a connection point between a corresponding road and an intersection at a boundary of the topological subgraph; Based on the associated intersection information of the intersection corresponding to each of the boundary points, determining the transmission direction of the data at each of the boundary points, the transmission direction including input and / or output; Based on the road association relationship between different topological subgraphs and the transmission direction of each boundary point, a data channel matrix for all topological subgraphs is constructed, and the data channel matrix is ​​segmented to obtain a data channel construction set corresponding to each topological subgraph; Based on each of the data channel construction sets, generate corresponding channel construction information, and send each channel construction information to the corresponding computing node; Each of the computing nodes is further configured to construct a channel with other computing nodes based on the received channel construction information.

7. The device according to claim 1 or 5, characterized in that: The data channel construction module is also used to set the transmission structure for data transmission between the computing nodes; The transmission structure setting includes at least one of the following: setting a basic data structure of each type of data to be transmitted in each channel according to the type of data to be transmitted between the computing nodes; setting a composite data structure based on multiple basic data structures; setting the transmission type of the data transmitted in the channel according to the transmission content generated during the traffic simulation process; Among them, the transmission type includes traffic unit transmission type, intersection transfer-in type, and transfer-out type, and the traffic unit transmission type includes traffic unit basic information type and traffic unit motion status information type.

8. The device according to claim 1 or 5, characterized in that: Each computing node is provided with a task scheduling data cache, and the task scheduling data cache includes: a scheduling basis cache for storing scheduling-related information of boundary points required for implementing task scheduling, an input data cache for storing input data, and an output data cache for storing output data, wherein the scheduling-related information of the boundary points includes at least one of traffic light information and road information; Wherein, the dynamic task scheduling module includes: A scheduling determination submodule is used to mark the traffic unit when it is determined that the traffic unit moves to the boundary point of the current topological subgraph, and store the marked traffic unit in the traffic unit set as a traffic unit to be scheduled if it is determined that the traffic unit can enter the next topological subgraph from the current topological subgraph according to the scheduling association information of the boundary point; A dispatching task data organization submodule, used to form the traffic unit information of each traffic unit to be dispatched and store it in the corresponding call-out data cache, wherein the traffic unit information includes dispatching task information, calculation status information, transmission structure information and channel information, wherein the dispatching task information is used to represent the basic information of the traffic unit, the calculation status information is used to represent the calculation parameter information of the traffic unit, the transmission structure information is determined according to the target channel for transmitting the traffic unit information, and the channel information is used to indicate the required target channel; The task scheduling submodule is used to control the current computing node where each of the traffic units to be scheduled is located to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel for the call-out task of each of the traffic units to be scheduled; and to control the current computing node to which each of the traffic units to be scheduled is called to store the corresponding data in the call-out data cache sent by the associated node into the call-in data cache through the corresponding channel for the call-in task of each of the traffic units to be scheduled; A computing task setting submodule is used to generate computing tasks corresponding to the traffic units to be dispatched according to the incoming tasks of each traffic unit to be dispatched, and send the computing tasks to the computing task pool of the computing node that the traffic unit to be dispatched needs to be called into; The associated nodes of a computing node are other computing nodes with which a channel exists.

9. A distributed micro-city traffic simulation method based on a supercomputing platform, characterized in that: Applied to a distributed microscopic urban traffic simulation device based on a supercomputing platform, the method comprises: The control configuration node generates a task request in response to a user operation, and sends the task request to the main computing node, wherein the task request carries configuration information, including a target number of computing nodes participating in the simulation and identification information indicating a target city to be simulated; Control the city traffic network module in the main computing node to obtain the city data set of the target city from the shared database based on the identification information; construct an urban traffic topology network according to the city data set, divide the urban traffic topology network according to the target number to form a plurality of topology subgraphs, and synchronize the plurality of topology subgraphs to the corresponding computing nodes, wherein the number of the topology subgraphs is the same as the target number; Controlling the data channel construction module in the main computing node to determine the channels that need to be constructed between different computing nodes based on the multiple topological subgraphs, and controlling each computing node to construct a corresponding channel, wherein the channel is used for data transmission between the computing nodes; When the dynamic task scheduling module is controlled to determine that the traffic unit moves to the current topological subgraph and moves to the boundary point during the traffic simulation process, the traffic unit that can enter the next topological subgraph from the current topological subgraph is stored in the traffic unit set as the traffic unit to be scheduled; the traffic unit information of each of the traffic units to be scheduled is formed and stored in the call-out data cache of the corresponding computing node; for the call-out task of each of the traffic units to be scheduled, the current computing node where each of the traffic units to be scheduled is located is controlled to send the corresponding data in the call-out data cache of the current computing node to the call-in data cache of the associated node through the channel; for the call-in task of each of the traffic units to be scheduled, the current computing node to which each of the traffic units to be scheduled is called is controlled to store the corresponding data in the call-out data cache sent from the associated node into the call-in data cache through the corresponding channel; Control the dynamic task scheduling module to generate a computing task corresponding to the traffic unit to be scheduled according to the transferred-in task of each traffic unit to be scheduled, and send the computing task to the computing task pool of the computing node that the traffic unit to be scheduled needs to be transferred into; Controlling the simulation computing engine in each computing node to obtain a computing task from a computing task pool of the computing node, obtain required basic information based on the computing task, and execute the computing task based on the basic information to obtain a computing result; Controlling the data loading module in each computing node to load data from the shared database according to the data requirement of the simulation computing engine to execute the computing task; The basic information includes data from the task scheduling data cache of the computing node and data loaded from the shared database, and the data loaded from the shared database includes at least one of map data, traffic unit data, and environmental data.

10. A distributed micro-city traffic simulation device based on a supercomputing platform, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the method of claim 9 when executing the instructions stored in the memory.

11. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method of claim 9 is implemented.

Citation Information

Patent Citations

  • System and method for testing urban rail train security detection sensing network

    CN103647684A

  • Microcosmic traffic simulation system based on GPU

    CN117350017A

  • Road traffic information detection method

    DE19858477A1

  • Collaborative simulation calculation method for urban rail transit network and station

    WO2023065595A1