CN117975736B - Unmanned vehicle road cooperative application scene test method and system - Google Patents

Unmanned vehicle road cooperative application scene test method and system Download PDF

Info

Publication number
CN117975736B
CN117975736B CN202410371229.XA CN202410371229A CN117975736B CN 117975736 B CN117975736 B CN 117975736B CN 202410371229 A CN202410371229 A CN 202410371229A CN 117975736 B CN117975736 B CN 117975736B
Authority
CN
China
Prior art keywords
scene
traffic
data
flow
road
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202410371229.XA
Other languages
Chinese (zh)
Other versions
CN117975736A (en
Inventor
戴金洲
张琳
刘嘉靖
韩超
吕庆斌
邬洋
王雪
沈上圯
郭鸾
沙硕
刘元晟
姚瑶
郭子君
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
BEIJING INSTITUTE OF METROLOGY
Original Assignee
BEIJING INSTITUTE OF METROLOGY
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by BEIJING INSTITUTE OF METROLOGY filed Critical BEIJING INSTITUTE OF METROLOGY
Priority to CN202410371229.XA priority Critical patent/CN117975736B/en
Publication of CN117975736A publication Critical patent/CN117975736A/en
Application granted granted Critical
Publication of CN117975736B publication Critical patent/CN117975736B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/25—Integrating or interfacing systems involving database management systems
    • G06F16/254—Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/2431—Multiple classes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/042—Knowledge-based neural networks; Logical representations of neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G06N3/0455—Auto-encoder networks; Encoder-decoder networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0475—Generative networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/094—Adversarial learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0125—Traffic data processing
    • G08G1/0133—Traffic data processing for classifying traffic situation
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G08G1/096766—Systems involving transmission of highway information, e.g. weather, speed limits where the system is characterised by the origin of the information transmission
    • G08G1/096783—Systems involving transmission of highway information, e.g. weather, speed limits where the system is characterised by the origin of the information transmission where the origin of the information is a roadside individual element
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2123/00—Data types
    • G06F2123/02—Data types in the time domain, e.g. time-series data
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Human Resources & Organizations (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Databases & Information Systems (AREA)
  • Tourism & Hospitality (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Chemical & Material Sciences (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Development Economics (AREA)
  • Atmospheric Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Remote Sensing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Quality & Reliability (AREA)

Abstract

The invention discloses a method and a system for testing a vehicle-road cooperative application scene of an unmanned vehicle, which relate to the technical field of vehicle-road cooperative testing and comprise the following steps: arranging sensors to collect traffic control data, constructing a traffic control scene library, classifying scenes, and determining scene parameters based on scene classification results; constructing a traffic scene simulation algorithm based on scene parameters in a traffic control scene library, and simulating traffic flow and vehicle behaviors in a specific scene; deploying and verifying a communication protocol model in a simulation environment, training through a machine learning algorithm, and identifying scene classification by using traffic data and scene parameters; and according to the identification result, formulating and executing an adaptive strategy, and adjusting the vehicle behavior and the traffic control strategy to adapt to the current traffic scene. The unmanned vehicle road cooperative application scene test method provided by the invention obviously improves the efficiency of traffic flow and enhances the safety of a traffic system. Reduces environmental pollution and improves the quality of life of urban residents.

Description

Unmanned vehicle road cooperative application scene test method and system
Technical Field
The invention relates to the technical field of vehicle-road cooperative testing, in particular to a method and a system for testing a vehicle-road cooperative application scene of an unmanned vehicle.
Background
The recent developments in traffic monitoring technology and unmanned technology have provided traffic control systems with a large amount of real-time data, but efficient integration, analysis and utilization of the data remains a challenge. Conventional traffic control systems often fail to make full use of these data for accurate traffic flow prediction and management, and lack sufficient flexibility and intelligence to cope with dynamic changes in urban traffic networks.
The traditional method mainly relies on a fixed signal plan and simple sensor feedback for traffic control. These methods have obvious effects in dealing with simple traffic conditions, but tend to be frustrating in the face of complex urban traffic environments, such as variable traffic flows, emergencies, and extreme weather conditions. The traditional method lacks the capability of carrying out the depth analysis and prediction on the real-time traffic data, and cannot realize the dynamic adaptation and optimization on the traffic conditions, so that the problems of traffic jam, accident frequency, environmental pollution and the like are caused.
Therefore, a method for testing the cooperative application scene of the unmanned vehicle and the vehicle road is needed to combine traffic data fusion, scene recognition, machine learning and dynamic traffic control strategies, so as to realize more intelligent and dynamic traffic control. The traffic jam is effectively relieved, and the road safety and the traffic efficiency are improved.
Disclosure of Invention
The present invention has been made in view of the above-described problems.
Therefore, the technical problems solved by the invention are as follows: the existing static traffic management and control method has the problems that the flexibility is insufficient, the real-time data cannot be fully utilized for dynamic adjustment, and the comprehensive analysis and the application optimization are performed according to the real-time traffic condition and the multi-source data.
In order to solve the technical problems, the invention provides the following technical scheme: a method for testing a vehicle-road cooperative application scene of an unmanned vehicle comprises the following steps: arranging sensors to collect traffic control data, constructing a traffic control scene library, classifying scenes, and determining scene parameters based on scene classification results; constructing a traffic scene simulation algorithm based on scene parameters in a traffic control scene library, and simulating traffic flow and vehicle behaviors in a specific scene; deploying and verifying a communication protocol model in a simulation environment, training through a machine learning algorithm, and identifying scene classification by using traffic data and scene parameters; according to the identification result, an adaptive strategy is formulated and executed, and the vehicle behavior and the traffic control strategy are adjusted to adapt to the current traffic scene; the traffic control data comprises traffic flow data, intersection type and signal lamp data, vehicle behavior data, pedestrian flow data, environmental factor data, special event data and real-time traffic state data; the construction of the traffic control scene library comprises the steps of processing data from different sources through an ETL data fusion technology, extracting vehicle flow and pedestrian flow information based on a convolutional neural network, and extracting key features from sensor data by utilizing time sequence analysis, wherein the key features comprise vehicle and pedestrian flow, vehicle running speed and waiting time; using a K-means algorithm to automatically classify scenes into an early peak flow scene, a weekend park people stream scene, a night residence scene, a construction recombination traffic scene and an emergency traffic scene; and carrying out scene definition on the judged scene based on the key features, and constructing a parameter model under the scene definition by applying a machine learning method according to the time characteristics, the space characteristics and the event types of the scene.
As a preferable scheme of the unmanned vehicle road cooperative application scene test method, the invention comprises the following steps: the construction of the traffic scene simulation algorithm comprises the steps of performing feature compression and coding on a parameter model under scene definition by adopting a self-coder network, and converting the parameter model into a scene representation vector; generating a traffic flow pattern in a specific scene based on the scene representation vector by using the generation countermeasure network GAN; introducing a graph neural network GNN to simulate the behavior mode of a vehicle and a pedestrian in a specific traffic scene; adjusting scene parameters in the simulation process by combining a reinforcement learning algorithm; applying the simulation result to a digital twin environment, synchronizing with real-time traffic data, and providing real-time simulation and prediction feedback; the scene classification identification by using traffic data and scene parameters comprises the steps of adopting a 5G network-based communication protocol model, introducing a graph-based deep learning technology, extracting network structure features from traffic flow and behavior modes, and encoding the structure information of the traffic network into feature vectors by a graph embedding technology; training an integrated deep neural network model by utilizing the extracted network structural features and traffic features, classifying scenes into an early peak flow scene, a weekend park people flow scene, a night residence scene, a construction recombination traffic scene and an emergency traffic scene by combining with a GAT graph attention network, feeding back the identification result to a communication protocol model, and optimizing a data transmission strategy; the step of formulating and executing the adaptive strategy according to the identification result comprises executing a first adjustment strategy if the adaptive strategy is classified as an early peak traffic scene; if the scene is classified as a weekend park people stream scene, executing a second adjustment strategy; a night house scene, executing a third adjustment strategy; constructing a recombined traffic scene and executing a fourth adjustment strategy; and executing a fifth adjustment strategy in the emergency traffic scene.
As a preferable scheme of the unmanned vehicle road cooperative application scene test method, the invention comprises the following steps: the first adjustment strategy is executed, wherein the first adjustment strategy comprises the steps of monitoring traffic flow data and intersection type data in real time when an early peak flow scene is judged, identifying regional congestion conditions through a congestion index identification model, executing a scheduling strategy based on the congestion conditions, and sending alternative route information to vehicles through V2X communication; monitoring the strategy effect and adjusting the scheduling strategy according to feedback;
The congestion index identification model is expressed as,
Wherein,Expressed in position/>And time/>Congestion index,/>Representing the spatial position of the observation point,/>Representing the point in time of the observation,/>The starting spatial position of the integral is shown, the starting point of the observation section is shown, 0 is the starting time point of the integral, and the starting time point is the beginning of the analysis period,/>Expressed in position/>And time/>Traffic density of/>Representing location infinitesimal,/>Representing time infinitesimal,/>Representation considering traffic density/>And environmental parameters/>Speed of traffic at time,/>Representing a set of environmental parameters,/>Representing maximum traffic flow speed in absence of congestion,/>Traffic density representing current time position and time,/>Representing the maximum density that causes traffic flow stagnation,/>Representing a non-linear parameter describing the effect of density on speed,/>An adjustment coefficient indicating the influence of environmental factors on the vehicle flow velocity.
As a preferable scheme of the unmanned vehicle road cooperative application scene test method, the invention comprises the following steps: the executing of the second adjustment strategy comprises the steps of constructing a people stream density identification model by utilizing pedestrian flow data and environmental factor data if the people stream scene of the weekend park is judged, and identifying a people stream high density area;
The people stream density recognition model is expressed as,
Wherein,Represents the/>Actual number of people in each zone,/>Represents the/>Area of the monitoring area,/>Representing environmental condition impact,/>Representing the people flow density per unit area,/>Respectively represent the comprehensive influence of environmental conditions, special events and time periods on people stream density,/>Representing the temperature effect, based on the difference between the current temperature and the optimal temperature,/>Indicating the effect of humidity, based on the difference between the current humidity and the optimal humidity,/>Representing weather effects, assigning values according to weather conditions,/>Representing special event effects, assigning values according to the special event effects,/>Representing event type, based on event expected people stream appeal, influencing assignment according to people stream appeal,/>Representing the number of predicted participants,/>Representing time period influence,/>Values representing the influence of different time periods of the day,/>Representing the difference impact value between weekends and weekdays,/>、/>Respectively represent the minimum value and the maximum value of the people flow density in the observation period and are used for normalization processing,/>Representing adjustment coefficient,/>Representing the normalized people stream density;
If it is According to the method, a traffic flow high-density area is judged, when traffic signals are dynamically adjusted according to traffic flow density identification results at park entrances and exits and traffic intersections, the green light time of pedestrian crossing streets is prolonged, the green light time of road sections near park exits is prolonged in a traffic flow peak period, temporary pedestrian crossing street areas are arranged in the traffic flow high-density area to disperse traffic flows, pedestrians are guided to use the designated crossing street areas by using movable railings and temporary marks to avoid randomly crossing the roads, traffic conditions, recommended routes and parking lot vacancy information of park periphery are issued in real time through road side units RSU, electronic information boards and mobile applications, and access suggestions and time selection guides are issued in advance through social media and park official network channels for tourists reserved to visit parks on weekends.
As a preferable scheme of the unmanned vehicle road cooperative application scene test method, the invention comprises the following steps: the third regulation strategy is executed, wherein the third regulation strategy comprises the steps of if a night residence scene is judged, making a preliminary regulation rule for limiting the speed of the vehicle and timing the signal lamps, if the traffic flow is lower than 20% of the average level in the evening, limiting the speed of the vehicle to 30 km/h, reducing the conversion frequency of the signal lamps, and prolonging the duration of the green lamps and the red lamps to 1.5 times of the normal period; installing noise monitoring equipment in a residential area, monitoring the night noise level in real time, setting a noise safety threshold as a feedback index of an adjustment strategy, executing a speed and signal lamp timing adjustment rule, monitoring the influence of the speed and the signal lamp timing adjustment rule on the noise level in real time, judging that the preliminary adjustment rule fails to reduce noise pollution if the monitored noise level exceeds the safety threshold, and dynamically adjusting the speed limit and the signal lamp timing according to the noise monitoring result until the noise level falls below the safety threshold; the fourth adjustment strategy is executed, wherein the fourth adjustment strategy comprises the steps of collecting special event data and construction area information if construction recombination traffic scenes are judged, setting temporary traffic signs and signals, issuing bypass information, informing a driver of the construction information and the bypass route through V2X communication, and setting according to the traffic flow change adjustment signs and signals.
As a preferable scheme of the unmanned vehicle road cooperative application scene test method, the invention comprises the following steps: executing the fifth adjustment strategy comprises processing special event data and traffic flow information in real time if the emergency traffic scene is judged, identifying an accident influence range, implementing emergency traffic control, temporarily closing an accident road section, starting emergency route navigation, providing real-time traffic information for a driver through V2X and recommending an optimal avoidance route.
As a preferable scheme of the unmanned vehicle road cooperative application scene test method, the invention comprises the following steps: identifying the accident influence range comprises calculating a congestion index according to traffic flow and speed data through a congestion index identification model; evaluating the people stream density of the accident area and the periphery thereof through a people stream density identification model; the method comprises the steps of taking a result of congestion index and people flow density evaluation as input of a decision tree model to analyze comprehensive influences of accidents on traffic and people flow, and identifying an accident influence range according to the comprehensive analysis result, wherein the accident influence range comprises road sections influenced by the accidents, accident influence degrees and people flow areas influenced by the accidents; the recommended optimal avoidance route comprises road sections affected by accidents, accident influence degrees and people flow areas affected by the accidents, and the road sections, the accident influence degrees and the people flow areas are integrated into input parameters of a path planning algorithm; updating the states of the affected road sections in the road network diagram model according to the analysis result of the accident influence range, wherein the states comprise increasing the passing cost of the affected road or setting the affected road to be non-passable; by means ofThe algorithm combines GANs drawing attention network to carry out dynamic path planning; according to the road network state and traffic information updated in real time, calculating an optimal avoidance route from the current position to the target position; said/>The algorithm is combined with GANs the attention network of the map, and comprises the steps of converting information data of road sections influenced by accidents, accident influence degrees and people flow areas influenced by the accidents into attributes of nodes and edges in a map model, wherein the road sections correspond to the edges in the map, and the accident influence degrees and the people flow areas correspond to the weights of the edges and the nodes; in the road network graph model, dynamically updating the state of the affected road section according to the analysis result of the accident influence range; and analyzing the updated road network graph model by using GANs graph attention network, and automatically learning the importance of each road section and each intersection by using a graph attention mechanism.
Another object of the present invention is to provide a system for testing a vehicle-road cooperative application scenario of an unmanned vehicle, which can solve the problem that the conventional static traffic control method cannot adapt to complex and variable traffic environments through a dynamic analysis and decision mechanism driven by real-time data.
In order to solve the technical problems, the invention provides the following technical scheme: an unmanned vehicle road cooperative application scene test system, comprising: the system comprises a data acquisition module, a scene simulation module, a scene classification module and a scene test module; the data acquisition module is used for arranging sensors to collect traffic control data, constructing a traffic control scene library, classifying scenes and determining scene parameters based on scene classification results; the scene simulation module is used for constructing a traffic scene simulation algorithm based on scene parameters in the traffic control scene library and simulating traffic flow and vehicle behaviors in a specific scene; the scene classification module is used for deploying and verifying a communication protocol model in a simulation environment, training through a machine learning algorithm, and identifying scene classification by utilizing traffic data and scene parameters; the scene test module is used for making and executing an adaptive strategy according to the identification result, and adjusting the vehicle behavior and the traffic control strategy to adapt to the current traffic scene.
A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the unmanned vehicle road cooperative application scenario test method as described above.
A computer readable storage medium having stored thereon a computer program, characterized in that the computer program when executed by a processor implements the steps of the unmanned vehicle road cooperative application scenario test method as described above.
The invention has the beneficial effects that: according to the unmanned vehicle and vehicle road cooperative application scene test method, the traffic scenes are accurately identified and classified, and the traffic control strategy is dynamically adjusted based on scene information, so that traffic jam is effectively reduced, traffic time is shortened, and road traffic capacity is improved. The emergency is monitored and responded in real time, the occurrence of traffic accidents is reduced, and the safety of unmanned vehicles is improved. In addition, the optimized traffic flow reduces vehicle emission, is beneficial to reducing urban air pollution and energy consumption, and promotes the development of green traffic. The personalized navigation service improves the travel experience of the user. The invention provides important support for the development of intelligent traffic systems, and has important significance for improving the efficiency and effect of urban traffic management.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is an overall flowchart of a method for testing a vehicle-road cooperative application scenario of an unmanned vehicle according to an embodiment of the present invention.
Detailed Description
So that the manner in which the above recited objects, features and advantages of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments, some of which are illustrated in the appended drawings. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.
Example 1
Referring to fig. 1, for one embodiment of the present invention, a method for testing a vehicle-road cooperative application scenario of an unmanned vehicle is provided, including:
the arrangement sensor collects traffic control data, builds a traffic control scene library, classifies scenes, and determines scene parameters based on scene classification results.
Based on scene parameters in the traffic control scene library, a traffic scene simulation algorithm is constructed to simulate traffic flow and vehicle behavior in a specific scene.
The communication protocol model is deployed and verified in the simulation environment, and the scene classification is identified by using traffic data and scene parameters through training of a machine learning algorithm.
And according to the identification result, formulating and executing an adaptive strategy, and adjusting the vehicle behavior and the traffic control strategy to adapt to the current traffic scene.
Traffic control data includes traffic flow data, intersection type and signal light data, vehicle behavior data, pedestrian flow data, environmental factor data, special event data, and real-time traffic state data.
Traffic flow data includes vehicle counts: the number of vehicles passing through a particular road segment in each time period; vehicle model distribution: the proportion and distribution of different types of vehicles (e.g. cars, vans, buses); intersection type and signal lamp data includes intersection layout: the type (such as T type, X type, ring) of the intersection and the dimension information thereof; signal lamp mode: the timing scheme of each intersection signal lamp comprises the duration time, the phase difference and the like of the traffic lights.
The vehicle behavior data includes a speed profile: vehicle running speeds at different times and different road sections; behavior pattern: behavior patterns of the vehicle at the intersection, such as straight, left turn, right turn, etc.; the pedestrian flow data includes pedestrian counts: the number of pedestrians passing through a specific area within a specific time period; pedestrian behavior: behavior patterns of pedestrians at intersections and the like, such as crossing roads, waiting for signal lights and the like.
The environmental factor data includes weather conditions: the influence of weather conditions such as temperature, rainfall, snow, fog and the like on traffic; illumination conditions: lighting conditions during the day, night or under specific weather conditions; special event data includes traffic accidents: time, place, type of accident and impact on traffic flow; and (3) road construction: time, place, type of construction and traffic control measures during the construction.
The real-time traffic state data includes traffic congestion index: real-time traffic congestion conditions for different road segments; real-time vehicle speed: real-time average speed of main road section.
Constructing a traffic control scene library, namely processing data from different sources through an ETL data fusion technology, extracting vehicle flow and pedestrian flow information based on a convolutional neural network, and extracting key features from sensor data by utilizing time sequence analysis, wherein the key features comprise vehicle and pedestrian flow, vehicle running speed and waiting time; using a K-means algorithm to automatically classify scenes into an early peak flow scene, a weekend park people stream scene, a night residence scene, a construction recombination traffic scene and an emergency traffic scene; and (3) carrying out scene definition on the judged scene based on the key characteristics, and constructing a parameter model under scene definition by applying a machine learning method according to the time characteristics, the space characteristics and the event types of the scene.
Automatically collecting data from traffic cameras, sensors, social media and weather applications, cleaning the data to remove errors and inconsistencies, removing non-traffic elements in images, standardizing social media message formats, converting all data into a unified format, unifying all timestamps into UTC time, unifying all position information into GPS coordinates, merging the data by using a weighted average algorithm, and improving the accuracy and reliability of the data. The method comprises the steps of extracting key information which is helpful for understanding traffic scenes from the fused data, identifying the number, the type and the moving direction of vehicles and pedestrians in images by using a convolutional neural network, analyzing traffic flow and speed change trend in sensor data by using a time sequence analysis method, and extracting traffic related events and emotion from social media data by using a natural language processing technology. Converting the extracted features into a numerical vector form, preparing for the input of a clustering algorithm, clustering the feature vectors by using a K-means algorithm, automatically grouping the feature vectors according to the similarity of the features to form different scenes, and distributing a clear scene tag data set for each clustering result to divide: according to the scene classification results, training and testing data sets are divided for each scene, a machine learning model is selected for each scene, and the data set training model of the corresponding scene is used. And (3) evaluating the accuracy and generalization capability of the model by using the test data set, adjusting model parameters according to the evaluation result, and optimizing the performance of the model.
Constructing a traffic scene simulation algorithm, wherein the traffic scene simulation algorithm comprises the steps of performing feature compression and coding on a parameter model under scene definition by adopting a self-coder network, and converting the parameter model into a scene representation vector; generating a traffic flow pattern in a specific scene based on the scene representation vector by using the generation countermeasure network GAN; introducing a graph neural network GNN to simulate the behavior mode of a vehicle and a pedestrian in a specific traffic scene; adjusting scene parameters in the simulation process by combining a reinforcement learning algorithm; the simulation result is applied to a digital twin environment and is synchronized with real-time traffic data, so that real-time simulation and prediction feedback is provided.
And directly verifying the accuracy of traffic flow and behavior patterns generated by the simulation algorithm by synchronizing with real-time traffic data. The key step of ensuring the reliability of the simulation results provides a real-time correction mechanism. The feedback of the real-time data enables the simulation system to adjust and optimize according to the latest traffic conditions. The adaptability of the simulation system is improved, and the simulation system can reflect and predict complex and variable actual traffic scenes. The real-time feedback provides a direct basis for evaluating the effects of different traffic control strategies, supports iterative optimization of traffic management decisions, and helps find the control strategy most suitable for the current traffic conditions. And synchronizing the simulation result with the real-time data to form a closed-loop testing, evaluating, optimizing and feedback system. This not only promotes the practicality and accuracy of the simulation, but also provides a mechanism for continuously improving traffic control strategies.
Identifying scene classification by using traffic data and scene parameters comprises adopting a 5G network-based communication protocol model, introducing a graph-based deep learning technology, extracting network structural features from traffic flows and behavior modes, and encoding structural information of a traffic network into feature vectors by a graph embedding technology;
and training an integrated deep neural network model by utilizing the extracted network structural features and traffic features, classifying the scene into an early peak flow scene, a weekend park people flow scene, a night residence scene, a construction recombination traffic scene and an emergency traffic scene by combining with a GAT graph attention network, feeding back the identification result to a communication protocol model, and optimizing a data transmission strategy.
The optimized data transmission strategy is based on the scene recognition effect and the implemented traffic control strategy collecting feedback, and the scene recognition model and the traffic control strategy are continuously and iteratively optimized. By using the online learning method, the system can adapt to the traffic environment with complex changes.
According to the identification result, formulating and executing an adaptive strategy comprises executing a first adjustment strategy if the adaptive strategy is classified as an early peak traffic scene; if the scene is classified as a weekend park people stream scene, executing a second adjustment strategy; a night house scene, executing a third adjustment strategy; constructing a recombined traffic scene and executing a fourth adjustment strategy; and executing a fifth adjustment strategy in the emergency traffic scene.
The self-encoder network application comprises a self-encoder network for feature compression and encoding of parametric models under complex scene definitions, generating scene representation vectors. The high-dimensional scene parameter data is converted into a low-dimensional scene representation vector by an encoder portion of the self-encoder. The compression process captures the most important features of scene parameters, providing a compact and accurate input for the subsequent generation of traffic flow patterns in a particular scene.
Based on the scene representation vector obtained in the previous step, a countermeasure network (GAN) is generated to generate a traffic flow pattern in a specific scene. The GAN is composed of a generator that is responsible for generating data and a arbiter that is responsible for determining whether the data is authentic or generated by the generator. Here, the generator attempts to generate a traffic flow pattern conforming to the real scene, and the arbiter directs the generator to improve the traffic flow pattern it generates until the generated data is indistinguishable from the real scene data.
A Graph Neural Network (GNN) was introduced to simulate the behavior patterns of vehicles and pedestrians in a specific traffic scenario. GNNs are able to process graph structure data making them particularly useful for analysis of traffic networks. Each node may represent an intersection or road segment and the edges represent roads. GNNs learn complex interactions in traffic flows through graph structures to simulate the behavior patterns of vehicles and pedestrians.
And adjusting scene parameters in the simulation process by combining a reinforcement learning algorithm. The reinforcement learning algorithm guides the model to self-optimize through a reward mechanism, and a strategy for optimizing traffic flow in a given scene is found through a continuous learning process. The algorithm can try different scene parameter configurations, and parameters are adjusted through simulation result feedback so as to achieve the purpose of optimizing the simulation result.
The simulation result is applied to a digital twin environment and is synchronized with real-time traffic data to provide real-time simulation and prediction feedback. The digital twin environment is a virtual mirror image, and can reflect the state of the physical world in real time. By applying the simulation result to digital twin, the real-time mapping and prediction of the real traffic environment are realized, and a traffic manager is helped to make more accurate and timely decisions.
And extracting network structure characteristics from traffic flow and behavior modes based on a communication protocol model of the 5G network and a deep learning technology, further training an integrated deep neural network model, and classifying scenes. By utilizing the graph embedding technology and the GAT graph annotation network, accurate classification is provided for each scene, and the data transmission strategy is optimized through the 5G network according to the identification result, so that real-time and accurate transmission of information is ensured.
Executing a first adjustment strategy, namely monitoring traffic flow data and intersection type data in real time when the traffic flow is judged to be an early peak flow scene, identifying regional congestion conditions through a congestion index identification model, executing a scheduling strategy based on the congestion conditions, and sending alternative route information to vehicles through V2X communication; monitoring the strategy effect and adjusting the scheduling strategy according to feedback;
the congestion index identification model is expressed as,
Wherein,Expressed in position/>And time/>Congestion index,/>Representing the spatial position of the observation point,/>Representing the point in time of the observation,/>The starting spatial position of the integral is shown, the starting point of the observation section is shown, 0 is the starting time point of the integral, and the starting time point is the beginning of the analysis period,/>Expressed in position/>And time/>Traffic density per unit length,/>Representing location infinitesimal,/>Representing time infinitesimal,/>Representation considering traffic density/>And environmental parameters/>Speed of traffic at time,/>Representing a set of environmental parameters, including various factors affecting traffic flow speed, such as weather conditions, time periods, special events, etc.,Representing maximum traffic flow speed in absence of congestion,/>Traffic density representing current time position and time,/>Representing the maximum density that causes traffic flow stagnation,/>Representing a non-linear parameter describing the effect of density on speed,/>An adjustment coefficient indicating the influence of environmental factors on the vehicle flow velocity.
The traffic density is obtained through real-time monitoring and reflects the number of vehicles in unit length on the road at a specific position x and time t.
The nonlinear parameter α is a parameter describing a nonlinear relationship between traffic density and vehicle flow speed, and has a value ranging between 0 and 1, which helps to ensure that the model behavior, in which the speed decreases with increasing density, coincides with the actual observation.
When α is close to 0, this means that the traffic flow speed is very sensitive to changes in traffic density; i.e. a slight increase in traffic density will result in a substantial decrease in traffic speed.
When α is close to 1, the traffic flow speed is less sensitive to changes in traffic density; this means that the traffic flow speed drops significantly only when the traffic density approaches its maximum value.
The optimal value of alpha should be determined according to specific traffic characteristics and actual observation data so as to ensure that the model can accurately reflect the traffic flow dynamics in the real world. In practical applications, experiments and data analysis are required to determine alpha values suitable for a particular road or traffic situation.
If it isIf the congestion condition of the current road section is larger than the first congestion threshold value, judging that the congestion condition of the current road section occurs, and immediately acquiring sensor data of the road section in the opposite direction to acquire the congestion condition classification of the road section in the opposite direction; the current road section is a road section A, and the opposite direction road section is a road section B;
When (when) When the traffic signal is greater than a first congestion threshold value and smaller than a second congestion threshold value, judging that the traffic signal is slightly congested, comparing the traffic signal with the congestion condition of a road section B, if the congestion condition of the road section B also occurs, maintaining the current signal, if the congestion condition of the road section B does not occur, adjusting the signal, increasing the green light time of the road section A for 10 seconds, performing real-time traffic data analysis on the green light time of the current road section after the traffic light time is adjusted, not adversely affecting other directions of an intersection, issuing real-time traffic information of the smooth road section B to a road section A driver through V2X communication, recommending the road section B to be used as an alternative route, performing adjustment effect evaluation on traffic condition changes of the road section A and the road section B after the signal is adjusted when the signal is continuously monitored, and dynamically optimizing an adjustment strategy;
When (when) When the traffic information issuing system is larger than or equal to the second congestion threshold value and smaller than or equal to the third congestion threshold value, judging that the traffic information issuing system is moderately congested, comparing the congestion situation of the road section B, if the road section B is not congested, increasing the green light time of the road section A to be at most 10 seconds, recommending a driver to bypass the road section B by using the traffic information issuing system, and providing clear route switching guidance through V2X communication; if the road section B is slightly congested, the green light time of the road section A is increased by 5 seconds, the road section A and the road section B are distributed uniformly through a traffic system, and the real-time traffic states of the road section A and the road section B are issued at the same time, so that a driver is guided to make an optimal decision according to personal conditions and real-time traffic information; if the road section B is moderate congestion, keeping the signal lamp timing of the road section A unchanged, clearly informing that the two road sections have moderate congestion through strengthening traffic information release, searching other alternative routes, and considering to adjust travel time to avoid peak time; if the road section B is severely congested, emergency traffic control measures are implemented on the road section A, temporary lane adjustment or entrance restriction are implemented, meanwhile, traffic warning and emergency blocking avoidance information are issued, the road section A and the road section B are recommended to be avoided, and detailed detour navigation advice is provided;
When (when) When the traffic congestion is larger than a third congestion threshold value, judging heavy congestion, not judging congestion condition classification of a section B, and starting a preset emergency response plan, wherein the emergency response plan comprises temporarily changing a route usage rule, starting an emergency lane, temporarily prohibiting a vehicle from entering a specific area and setting a unidirectional traffic section; and issuing emergency blocking avoidance information and alternative route suggestions through the V2X communication and social media platform, and guiding a driver to select other smooth routes according to real-time traffic conditions.
Executing a second adjustment strategy, namely if the scene is judged to be a weekend park people stream scene, constructing a people stream density identification model by utilizing pedestrian flow data and environmental factor data, and identifying a people stream high density area;
The people stream density recognition model is expressed as,
Wherein,Represents the/>Actual number of people in each zone,/>Represents the/>Area of the monitoring area,/>Representing environmental condition impact,/>Representing the people flow density per unit area,/>Respectively represent the comprehensive influence of environmental conditions, special events and time periods on people stream density,/>Representing the temperature effect, based on the difference between the current temperature and the optimal temperature,/>Indicating the effect of humidity, based on the difference between the current humidity and the optimal humidity,/>Representing weather effects, assigning values according to weather conditions,/>Representing special event effects, assigning values according to the special event effects,/>Representing event type, based on event expected people stream appeal, influencing assignment according to people stream appeal,/>Representing the number of predicted participants,/>Representing time period influence,/>Values representing the influence of different time periods of the day,/>Representing the difference impact value between weekends and weekdays,/>、/>Respectively represent the minimum value and the maximum value of the people flow density in the observation period and are used for normalization processing,/>Representing adjustment coefficient,/>Representing normalized people stream density.
If it isAccording to the method, a traffic flow high-density area is judged, when traffic signals are dynamically adjusted according to traffic flow density identification results at park entrances and exits and traffic intersections, the green light time of pedestrian crossing streets is prolonged, the green light time of road sections near park exits is prolonged in a traffic flow peak period, temporary pedestrian crossing street areas are arranged in the traffic flow high-density area to disperse traffic flows, pedestrians are guided to use the designated crossing street areas by using movable railings and temporary marks to avoid randomly crossing the roads, traffic conditions, recommended routes and parking lot vacancy information of park periphery are issued in real time through road side units RSU, electronic information boards and mobile applications, and access suggestions and time selection guides are issued in advance through social media and park official network channels for tourists reserved to visit parks on weekends.
Effect of temperatureAnd carrying out correlation analysis by using the historical weather data and the people flow data to determine an optimal temperature interval. And then comparing the real-time temperature data with the optimal temperature interval to calculate a temperature effect value. If the optimum temperature interval is/>To/>The current temperature is/>The calculated temperature effect value reflects the extent to which the temperature exceeds the optimum interval.
Effect of humidityA method similar to the temperature effect determines a comfort humidity interval by analyzing historical humidity data and people stream data. The difference between the current humidity and the comfort humidity interval is used to calculate a humidity effect value, indicating the potential impact of humidity on the flow of people.
Weather effectDifferent weather conditions (sunny, rainy, snowy, etc.) are given different weights based on historical weather conditions and people flow data. These weight values are learned by a machine learning model that uses historical weather conditions and corresponding people stream data as inputs during model training.
Event typeAnd analyzing the influence of various events in the history on the flow of people, and distributing weights for the events of different types. These weights may be derived based on historical attractions of the event and engagement statistics analysis.
Number of people expected to participateFor upcoming events, the predictive model is used to estimate the number of persons expected to participate, based on factors such as their type, scale, and promotional strength. By analyzing historical data of similar events, predictions are made in connection with social media trend analysis, and information provided by event organizers.
Time period influenceBy analyzing the historical people stream data, the mode of people stream change in one day is identified, and different influence values are given to different time periods.
Weekend and day of work differenceAnd analyzing the difference of the people stream data of the weekdays and the weekends to determine the quantized value of the weekend effect. And obtaining the water flow difference between the weekends and the weekdays at the specific places through statistical analysis.
Executing a third regulation strategy, namely if the night residence scene is judged, formulating a vehicle speed limit and a preliminary regulation rule when the traffic light is timed, if the traffic flow is lower than 20% of the average level in the evening, reducing the vehicle speed limit to 30 km/h, reducing the conversion frequency of the traffic light, and prolonging the duration of the green light and the red light to 1.5 times of the normal period;
Installing noise monitoring equipment in a residential area, monitoring the night noise level in real time, setting a noise safety threshold as a feedback index of an adjustment strategy, executing a speed and signal lamp timing adjustment rule, monitoring the influence of the speed and the signal lamp timing adjustment rule on the noise level in real time, judging that the preliminary adjustment rule fails to reduce noise pollution if the monitored noise level exceeds the safety threshold, dynamically adjusting the speed limit and the signal lamp timing according to the noise monitoring result, reducing the speed limit, prolonging the red light duration of the signal lamp, reducing the frequency of vehicles passing through the residential area until the noise level falls below the safety threshold;
Executing the fourth adjustment strategy comprises collecting special event data and construction area information if the construction recombination traffic scene is judged, setting temporary traffic signs and signals, issuing detour information, informing a driver of the construction information and detour route through V2X communication, and setting the adjustment signs and signals according to the traffic flow.
Executing the fifth adjustment strategy comprises processing special event data and traffic flow information in real time if the emergency traffic scene is judged, identifying an accident influence range, implementing emergency traffic control, temporarily closing an accident road section, starting emergency route navigation, providing real-time traffic information for a driver through V2X and recommending an optimal avoidance route.
Identifying the accident influence range comprises calculating a congestion index according to traffic flow and speed data through a congestion index identification model; evaluating the people stream density of the accident area and the periphery thereof through a people stream density identification model; the result of the congestion index and the people flow density evaluation is used as the input of a decision tree model to analyze the comprehensive influence of accidents on traffic and people flow, and the accident influence range is identified according to the comprehensive analysis result, wherein the accident influence range comprises road sections influenced by the accidents, the accident influence degree and people flow areas influenced by the accidents;
The recommended optimal avoidance route comprises integrating road sections affected by accidents, the accident influence degree and people flow areas affected by the accidents into input parameters of a path planning algorithm; updating the state of the affected road section in the road network diagram model according to the analysis result of the accident influence range, wherein the state comprises the passing cost of the affected road or the condition that the affected road is not passed; by means of The algorithm combines GANs drawing attention network to carry out dynamic path planning; according to the road network state and traffic information updated in real time, calculating an optimal avoidance route from the current position to the target position; utilization/>The algorithm combines GANs drawing attention network to carry out dynamic path planning; and calculating an optimal avoidance route from the current position to the target position according to the road network state and the traffic information updated in real time.
The algorithm is combined with GANs the attention network of the map, and comprises the steps of converting information data of road sections influenced by accidents, accident influence degrees and people flow areas influenced by the accidents into attributes of nodes and edges in a map model, wherein the road sections correspond to the edges in the map, and the accident influence degrees and the people flow areas correspond to the weights of the edges and the nodes; in the road network graph model, dynamically updating the states of the affected road segments according to the analysis result of the accident influence range, which involves adjusting the weights of the corresponding road segments (edges), and simulating the non-passable state by increasing the weights (passing cost) of the road segments with serious accident or setting extremely high weight values for the road segments with serious accident; the method comprises the steps of analyzing an updated road network graph model by using GANs graph attention network, automatically learning the importance of each road section and each intersection by using the graph attention mechanism, enabling the model to pay more attention to the road section and the intersection with the greatest influence on avoiding route selection during route planning in traffic flow recombination under the influence of accidents, and applying/> based on the road network graph model weighted by the graph attention networkThe algorithm adopts a heuristic function when planning a path, the heuristic function considers the traditional distance cost, and combines the accident influence weight provided by the graph annotation force network and the real-time traffic information; /(I)The algorithm searches an optimal path from a starting point to an end point based on the road network graph model, and judges the optimal avoidance route.
Example 2
The second embodiment of the invention provides a system for testing a vehicle-road cooperative application scene of an unmanned vehicle, which comprises the following components:
the system comprises a data acquisition module, a scene simulation module, a scene classification module and a scene test module.
The data acquisition module is used for arranging sensors to collect traffic control data, constructing a traffic control scene library, classifying scenes and determining scene parameters based on scene classification results.
The scene simulation module is used for constructing a traffic scene simulation algorithm based on scene parameters in the traffic control scene library and simulating traffic flow and vehicle behaviors in a specific scene.
The scene classification module is used for deploying and verifying a communication protocol model in a simulation environment, is trained through a machine learning algorithm, and is used for identifying scene classification by using traffic data and scene parameters.
The scene test module is used for making and executing an adaptive strategy according to the identification result, and adjusting the vehicle behavior and the traffic control strategy to adapt to the current traffic scene.
Example 3
One embodiment of the present invention, which is different from the first two embodiments, is:
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
Example 4
For verifying the beneficial effects of the invention, the effects of the technical scheme of the invention in the aspects of improving traffic efficiency, reducing accident rate, reducing vehicle emission and improving user satisfaction are evaluated, compared with the traditional traffic management method, and scientific demonstration is carried out through simulation experiments.
A virtual city comprising a plurality of traffic scenes is designed, and the virtual city has a complex road network, various traffic flows and environmental conditions.
And simulating traffic flow through SUMO simulation software, and collecting traffic flow, speed, accident record, emission data and other information.
The experiment was repeated 3 times over different time periods including peak morning and evening, peak evening, weekend and holiday, each time period was simulated for at least 24 hours to obtain stable experimental results.
The traditional method adopts a fixed signal lamp timing scheme and standard traffic flow management measures, and does not adjust traffic control strategies according to real-time traffic data.
The method implements dynamic signal lamp timing adjustment, traffic flow prediction and congestion identification based on real-time data, and an adaptive traffic control strategy.
The experimental process comprises the preparation stage, setting a virtual city environment, and defining a road network, a traffic flow model and environment variables; simulation operation, namely respectively operating the traditional method and the method of the invention, and collecting experimental index data; data analysis: carrying out statistical analysis on the collected data, and evaluating the performance difference of the two schemes; results comparison the difference in performance of the two schemes on different indicators was demonstrated using a graph and statistical method and the experimental results are shown in table 1.
Table 1 comparison of experimental results
The method adopts the congestion index identification model and the deep neural network based on the real-time traffic data, and optimizes traffic flow and signal lamp scheduling. Compared with the traditional fixed time period signal timing method, the method reduces the average passing time and improves the road use efficiency.
By utilizing the people flow density recognition model and the environmental factor data analysis, the dynamic adjustment of the traffic signal and the pedestrian crossing area is realized. Compared with the traditional method which is adjusted by experience, the method is more scientific and accurate, and the pedestrian safety and the traffic smoothness are obviously improved.
By monitoring the night traffic flow and the noise level, the invention dynamically adjusts the speed limit and the signal lamp timing, and compared with the traditional non-dynamic adjusting method, the invention effectively reduces the noise pollution and improves the life quality of residents.
The traffic flow reorganization strategy based on real-time traffic and construction information is adopted, so that the traffic flow reorganization strategy is more flexible and effective than the traditional preset construction route adjustment strategy, traffic delay is reduced, and the satisfaction degree of a driver is improved.
By combining the real-time monitoring and analysis of traffic flow and people flow density and by an improved path planning algorithm, the emergency is responded quickly, and compared with the traditional method, the traffic management efficiency and safety under the emergency are improved obviously.
The method of the invention realizes the optimization of traffic flow in a plurality of key traffic scenes by integrating multisource data processing, deep learning and vehicle-road cooperative technology, improves the safety and environmental friendliness, and is obviously superior to the traditional traffic management method. Not only improves the overall efficiency of the traffic system, but also enhances the adaptability of the system to various conditions.
It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.

Claims (7)

1. The method for testing the cooperative application scene of the unmanned vehicle and the vehicle road is characterized by comprising the following steps of:
Arranging sensors to collect traffic control data, constructing a traffic control scene library, classifying scenes, and determining scene parameters based on scene classification results;
constructing a traffic scene simulation algorithm based on scene parameters in a traffic control scene library, and simulating traffic flow and vehicle behaviors in a specific scene;
deploying and verifying a communication protocol model in a simulation environment, training through a machine learning algorithm, and identifying scene classification by using traffic data and scene parameters;
According to the identification result, an adaptive strategy is formulated and executed, and the vehicle behavior and the traffic control strategy are adjusted to adapt to the current traffic scene;
The traffic control data comprises traffic flow data, intersection type and signal lamp data, vehicle behavior data, pedestrian flow data, environmental factor data, special event data and real-time traffic state data;
The construction of the traffic control scene library comprises the steps of processing data from different sources through an ETL data fusion technology, extracting vehicle flow and pedestrian flow information based on a convolutional neural network, and extracting key features from sensor data by utilizing time sequence analysis, wherein the key features comprise vehicle and pedestrian flow, vehicle running speed and waiting time; using a K-means algorithm to automatically classify scenes into an early peak flow scene, a weekend park people stream scene, a night residence scene, a construction recombination traffic scene and an emergency traffic scene; performing scene definition on the judged scene based on the key features, and constructing a parameter model under scene definition by applying a machine learning method according to the time characteristics, the space characteristics and the event types of the scene;
The construction of the traffic scene simulation algorithm comprises the steps of performing feature compression and coding on a parameter model under scene definition by adopting a self-encoder network, and converting the parameter model into a scene representation vector; generating a traffic flow pattern in a specific scene based on the scene representation vector by using the generation countermeasure network GAN; introducing a graph neural network GNN to simulate the behavior mode of a vehicle and a pedestrian in a specific traffic scene; adjusting scene parameters in the simulation process by combining a reinforcement learning algorithm; applying the simulation result to a digital twin environment, synchronizing with real-time traffic data, and providing real-time simulation and prediction feedback;
The scene classification identification by using traffic data and scene parameters comprises the steps of adopting a 5G network-based communication protocol model, introducing a graph-based deep learning technology, extracting network structure features from traffic flow and behavior modes, and encoding the structure information of the traffic network into feature vectors by a graph embedding technology;
Training an integrated deep neural network model by utilizing the extracted network structural features and traffic features, classifying scenes into an early peak flow scene, a weekend park people flow scene, a night residence scene, a construction recombination traffic scene and an emergency traffic scene by combining with a GAT graph attention network, feeding back the identification result to a communication protocol model, and optimizing a data transmission strategy;
According to the identification result, formulating and executing an adaptive strategy comprises executing a first adjustment strategy if the adaptive strategy is classified as an early peak traffic scene; if the scene is classified as a weekend park people stream scene, executing a second adjustment strategy; a night house scene, executing a third adjustment strategy; constructing a recombined traffic scene and executing a fourth adjustment strategy; and executing a fifth adjustment strategy in the emergency traffic scene.
2. The unmanned vehicle road cooperative application scenario testing method according to claim 1, wherein: the first adjustment strategy is executed, namely when the early peak flow scene is judged, traffic flow data and intersection type data are monitored in real time, regional congestion conditions are identified through a congestion index identification model, a scheduling strategy is executed based on the congestion conditions, and alternative route information is sent to vehicles through V2X communication; monitoring the strategy effect and adjusting the scheduling strategy according to feedback;
The congestion index identification model is expressed as,
Wherein,Expressed in position/>And time/>Congestion index,/>Representing the spatial position of the observation point,/>Representing the point in time of the observation,/>The starting spatial position of the integral is shown, the starting point of the observation section is shown, 0 is the starting time point of the integral, and the starting time point is the beginning of the analysis period,/>Expressed in position/>And time/>Traffic density of/>Representing location infinitesimal,/>Representing time infinitesimal,/>Representation considering traffic density/>And environmental parameters/>Speed of traffic at time,/>Representing a set of environmental parameters,/>Representing maximum traffic flow speed in absence of congestion,/>Traffic density representing current time position and time,/>Representing the maximum density that causes traffic flow stagnation,/>Representing a non-linear parameter describing the effect of density on speed,/>An adjustment coefficient indicating the influence of environmental factors on the vehicle flow velocity.
3. The unmanned vehicle road cooperative application scenario testing method according to claim 2, wherein: the executing of the second adjustment strategy comprises the steps of constructing a people stream density identification model by utilizing pedestrian flow data and environmental factor data if the people stream scene of the weekend park is judged, and identifying a people stream high density area;
The people stream density recognition model is expressed as,
Wherein,Represents the/>Actual number of people in each zone,/>Represents the/>Area of the monitoring area,/>Representing environmental condition impact,/>Representing the people flow density per unit area,/>Respectively represent the comprehensive influence of environmental conditions, special events and time periods on people stream density,/>Representing the temperature effect, based on the difference between the current temperature and the optimal temperature,/>Indicating the effect of humidity, based on the difference between the current humidity and the optimal humidity,/>Representing weather effects, assigning values according to weather conditions,/>Representing special event effects, assigning values according to the special event effects,/>Representing event type, based on event expected people stream appeal, influencing assignment according to people stream appeal,/>Representing the number of predicted participants,/>Representing time period influence,/>Values representing the influence of different time periods of the day,/>Representing the difference impact value between weekends and weekdays,/>、/>Respectively represent the minimum value and the maximum value of the people flow density in the observation period and are used for normalization processing,/>Representing adjustment coefficient,/>Representing the normalized people stream density;
If it is According to the method, a traffic flow high-density area is judged, when traffic signals are dynamically adjusted according to traffic flow density identification results at park entrances and exits and traffic intersections, the green light time of pedestrian crossing streets is prolonged, the green light time of road sections near park exits is prolonged in a traffic flow peak period, temporary pedestrian crossing street areas are arranged in the traffic flow high-density area to disperse traffic flows, pedestrians are guided to use the designated crossing street areas by using movable railings and temporary marks to avoid randomly crossing the roads, traffic conditions, recommended routes and parking lot vacancy information of park periphery are issued in real time through road side units RSU, electronic information boards and mobile applications, and access suggestions and time selection guides are issued in advance through social media and park official network channels for tourists reserved to visit parks on weekends.
4. The unmanned vehicle road cooperative application scenario testing method according to claim 3, wherein: the third regulation strategy is executed, wherein the third regulation strategy comprises the steps of if a night residence scene is judged, making a preliminary regulation rule for limiting the speed of the vehicle and timing the signal lamps, if the traffic flow is lower than 20% of the average level in the evening, limiting the speed of the vehicle to 30 km/h, reducing the conversion frequency of the signal lamps, and prolonging the duration of the green lamps and the red lamps to 1.5 times of the normal period;
Installing noise monitoring equipment in a residential area, monitoring the night noise level in real time, setting a noise safety threshold as a feedback index of an adjustment strategy, executing a speed and signal lamp timing adjustment rule, monitoring the influence of the speed and the signal lamp timing adjustment rule on the noise level in real time, judging that the preliminary adjustment rule fails to reduce noise pollution if the monitored noise level exceeds the safety threshold, and dynamically adjusting the speed limit and the signal lamp timing according to the noise monitoring result until the noise level falls below the safety threshold;
The fourth adjustment strategy is executed, wherein the fourth adjustment strategy comprises the steps of collecting special event data and construction area information if construction recombination traffic scenes are judged, setting temporary traffic signs and signals, issuing bypass information, informing a driver of the construction information and the bypass route through V2X communication, and setting according to the traffic flow change adjustment signs and signals.
5. The unmanned vehicle road cooperative application scenario testing method according to claim 4, wherein: executing the fifth adjustment strategy comprises processing special event data and traffic flow information in real time if the emergency traffic scene is judged, identifying an accident influence range, implementing emergency traffic control, temporarily closing an accident road section, starting emergency route navigation, providing real-time traffic information for a driver through V2X and recommending an optimal avoidance route.
6. The unmanned vehicle road cooperative application scenario testing method according to claim 5, wherein: the accident influence range identification comprises calculating a congestion index according to traffic flow and speed data through a congestion index identification model; evaluating the people stream density of the accident area and the periphery thereof through a people stream density identification model; the result of the congestion index and the people flow density evaluation is used as the input of a decision tree model to analyze the comprehensive influence of accidents on traffic and people flow, and the accident influence range is identified according to the comprehensive analysis result, wherein the accident influence range comprises road sections influenced by the accidents, the accident influence degree and people flow areas influenced by the accidents;
The recommended optimal avoidance route comprises road sections affected by accidents, accident influence degrees and people flow areas affected by the accidents, and the road sections, the accident influence degrees and the people flow areas are integrated into input parameters of a path planning algorithm; updating the road state of the affected road section in the road network diagram model according to the analysis result of the accident influence range, wherein the road state comprises the increase of the passing cost of the affected road or the setting of the non-passing road; by means of The algorithm combines GANs drawing attention network to carry out dynamic path planning; according to the road network state and traffic information updated in real time, calculating an optimal avoidance route from the current position to the target position;
The said The algorithm is combined with GANs the attention network of the map, and comprises the steps of converting information data of road sections influenced by accidents, accident influence degrees and people flow areas influenced by the accidents into attributes of nodes and edges in a map model, wherein the road sections correspond to the edges in the map, and the accident influence degrees and the people flow areas correspond to the weights of the edges and the nodes; in the road network graph model, dynamically updating the road state of the affected road section according to the analysis result of the accident influence range; and analyzing the updated road network graph model by using GANs graph attention network, and automatically learning the importance of each road section and each intersection by using a graph attention mechanism.
7. A system employing the unmanned vehicle road cooperative application scenario testing method as claimed in any one of claims 1 to 6, comprising: the system comprises a data acquisition module, a scene simulation module, a scene classification module and a scene test module;
The data acquisition module is used for arranging sensors to collect traffic control data, constructing a traffic control scene library, classifying scenes and determining scene parameters based on scene classification results;
the scene simulation module is used for constructing a traffic scene simulation algorithm based on scene parameters in the traffic control scene library and simulating traffic flow and vehicle behaviors in a specific scene;
The scene classification module is used for deploying and verifying a communication protocol model in a simulation environment, training through a machine learning algorithm, and identifying scene classification by utilizing traffic data and scene parameters;
the scene test module is used for making and executing an adaptive strategy according to the identification result, and adjusting the vehicle behavior and the traffic control strategy to adapt to the current traffic scene.
CN202410371229.XA 2024-03-29 2024-03-29 Unmanned vehicle road cooperative application scene test method and system Active CN117975736B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410371229.XA CN117975736B (en) 2024-03-29 2024-03-29 Unmanned vehicle road cooperative application scene test method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410371229.XA CN117975736B (en) 2024-03-29 2024-03-29 Unmanned vehicle road cooperative application scene test method and system

Publications (2)

Publication Number Publication Date
CN117975736A CN117975736A (en) 2024-05-03
CN117975736B true CN117975736B (en) 2024-06-07

Family

ID=90858332

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410371229.XA Active CN117975736B (en) 2024-03-29 2024-03-29 Unmanned vehicle road cooperative application scene test method and system

Country Status (1)

Country Link
CN (1) CN117975736B (en)

Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108646586A (en) * 2018-03-20 2018-10-12 重庆邮电大学 A kind of intelligent network connection automobile assemblage on-orbit, test verification System and method for
CN109714421A (en) * 2018-12-28 2019-05-03 国汽(北京)智能网联汽车研究院有限公司 Intelligent network based on bus or train route collaboration joins automobilism system
CN111260924A (en) * 2020-02-10 2020-06-09 北京中交国通智能交通系统技术有限公司 Traffic intelligent control and service release strategy method adapting to edge calculation
CN111859618A (en) * 2020-06-16 2020-10-30 长安大学 Multi-end in-loop virtual-real combined traffic comprehensive scene simulation test system and method
CN111882924A (en) * 2020-07-28 2020-11-03 上海詹妮建筑设计咨询有限公司 Vehicle testing system, driving behavior judgment control method and accident early warning method
CN113256976A (en) * 2021-05-12 2021-08-13 中移智行网络科技有限公司 Vehicle-road cooperative system, analog simulation method, vehicle-mounted equipment and road side equipment
CN114067561A (en) * 2021-10-25 2022-02-18 东南大学 Virtual reality testing method for urban expressway vehicle-road cooperative active management and control system
CN115116231A (en) * 2022-08-26 2022-09-27 深圳市城市交通规划设计研究中心股份有限公司 Vehicle-road cooperative microscopic simulation system and method, electronic device and storage medium
CN115285143A (en) * 2022-08-03 2022-11-04 东北大学 Automatic driving vehicle navigation method based on scene classification
CN115587463A (en) * 2021-07-06 2023-01-10 上海国际汽车城（集团）有限公司 Automatic driving test scene design method based on scene classification
CN115909783A (en) * 2022-11-24 2023-04-04 同济大学 Lane-level driving assistance method and system based on traffic flow
CN116205024A (en) * 2022-11-09 2023-06-02 吉林大学 Self-adaptive automatic driving dynamic scene general generation method for high-low dimension evaluation scene
CN116229725A (en) * 2023-05-06 2023-06-06 北京市计量检测科学研究院 Traffic control method and system based on simulated traffic scene
CN117056153A (en) * 2022-05-13 2023-11-14 保时捷股份公司 Methods, systems, and computer program products for calibrating and verifying driver assistance systems and/or autopilot systems
CN117290997A (en) * 2023-08-14 2023-12-26 吉林大学 Evaluation method of man-machine co-driving decision system based on digital twin mode
CN117521389A (en) * 2023-11-17 2024-02-06 东南大学 Vehicle perception test method based on vehicle-road collaborative perception simulation platform

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US12039860B2 (en) * 2018-10-16 2024-07-16 Five AI Limited Driving scenarios for autonomous vehicles

Patent Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108646586A (en) * 2018-03-20 2018-10-12 重庆邮电大学 A kind of intelligent network connection automobile assemblage on-orbit, test verification System and method for
CN109714421A (en) * 2018-12-28 2019-05-03 国汽(北京)智能网联汽车研究院有限公司 Intelligent network based on bus or train route collaboration joins automobilism system
CN111260924A (en) * 2020-02-10 2020-06-09 北京中交国通智能交通系统技术有限公司 Traffic intelligent control and service release strategy method adapting to edge calculation
CN111859618A (en) * 2020-06-16 2020-10-30 长安大学 Multi-end in-loop virtual-real combined traffic comprehensive scene simulation test system and method
CN111882924A (en) * 2020-07-28 2020-11-03 上海詹妮建筑设计咨询有限公司 Vehicle testing system, driving behavior judgment control method and accident early warning method
CN113256976A (en) * 2021-05-12 2021-08-13 中移智行网络科技有限公司 Vehicle-road cooperative system, analog simulation method, vehicle-mounted equipment and road side equipment
CN115587463A (en) * 2021-07-06 2023-01-10 上海国际汽车城（集团）有限公司 Automatic driving test scene design method based on scene classification
CN114067561A (en) * 2021-10-25 2022-02-18 东南大学 Virtual reality testing method for urban expressway vehicle-road cooperative active management and control system
CN117056153A (en) * 2022-05-13 2023-11-14 保时捷股份公司 Methods, systems, and computer program products for calibrating and verifying driver assistance systems and/or autopilot systems
CN115285143A (en) * 2022-08-03 2022-11-04 东北大学 Automatic driving vehicle navigation method based on scene classification
CN115116231A (en) * 2022-08-26 2022-09-27 深圳市城市交通规划设计研究中心股份有限公司 Vehicle-road cooperative microscopic simulation system and method, electronic device and storage medium
CN116205024A (en) * 2022-11-09 2023-06-02 吉林大学 Self-adaptive automatic driving dynamic scene general generation method for high-low dimension evaluation scene
CN115909783A (en) * 2022-11-24 2023-04-04 同济大学 Lane-level driving assistance method and system based on traffic flow
CN116229725A (en) * 2023-05-06 2023-06-06 北京市计量检测科学研究院 Traffic control method and system based on simulated traffic scene
CN117290997A (en) * 2023-08-14 2023-12-26 吉林大学 Evaluation method of man-machine co-driving decision system based on digital twin mode
CN117521389A (en) * 2023-11-17 2024-02-06 东南大学 Vehicle perception test method based on vehicle-road collaborative perception simulation platform

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
基于MM-STConv的端到端自动驾驶行为决策模型;赵祥模;连心雨;刘占文;沈超;董鸣;;中国公路学报;20201231(第03期);第174-187页 *
无人驾驶车辆行为决策系统研究;熊璐;康宇宸;张培志;朱辰宇;余卓平;;汽车技术;20180803(第08期);第4-12页 *

Also Published As

Publication number Publication date
CN117975736A (en) 2024-05-03

Similar Documents

Publication Publication Date Title
CN111583639B (en) Road traffic jam early warning method and system
CN109927709B (en) Vehicle driving route working condition determining method, energy management method and system
CN104157139B (en) A kind of traffic congestion Forecasting Methodology and method for visualizing
CN111739284B (en) Traffic signal lamp intelligent timing method based on genetic algorithm optimization fuzzy control
CN104778834B (en) Urban road traffic jam judging method based on vehicle GPS data
CN110956807B (en) Highway flow prediction method based on combination of multi-source data and sliding window
EP3631616A1 (en) Road traffic control system, method, and electronic device
CN109360429A (en) A kind of urban highway traffic dispatching method and system based on simulative optimization
CN109902899A (en) Information generating method and device
JP7504647B2 (en) Computer-implemented method executed by a computer for training a machine learning system for generating a driving curve and/or a driving path of a vehicle, a method for generating a driving curve and/or a driving path of a vehicle, a method for evaluating a driving curve and/or a driving path of a vehicle, a method for identifying route-specific emissions of a drive system of a vehicle, a method for adapting a drive system of a vehicle, as well as a computer program, a machine-readable storage medium, and a computer-implemented machine learning system
CN114495486B (en) Microcosmic traffic flow prediction system and microcosmic traffic flow prediction method based on hierarchical reinforcement learning
CN117351702A (en) Intelligent traffic management method based on adjustment of traffic flow
CN115662113A (en) Signalized intersection people-vehicle game conflict risk assessment and early warning method
CN112016735A (en) Patrol route planning method and system based on traffic violation hotspot prediction and readable storage medium
CN117273964B (en) Intelligent vehicle insurance policy generation system and method for self-adaptive driving data
CN117593167B (en) Intelligent city planning management method and system based on big data
CN117351734A (en) Intelligent regulation and control method and system for vehicle delay
CN118135819A (en) Traffic signal lamp regulation and control method and equipment based on urban road condition
CN117975736B (en) Unmanned vehicle road cooperative application scene test method and system
CN117667699A (en) Intelligent networking automobile test scene generation method based on knowledge graph
Dampage et al. Adaptive & coordinated traffic signal system
KR20210128823A (en) Crossroads LOS Prediction Method Based on Big Data and AI, and Storage Medium Having the Same
Sinha et al. Sustainable time series model for vehicular traffic trends prediction in metropolitan network
Ayson et al. Design of an adaptive traffic light network system through an AIoT-based Analytic model
CN113284338B (en) Method for calculating influence of motor vehicle emergency avoidance no-lamp control pedestrian crossing on traffic flow

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant