Method, System, Medium and Device for Constructing Road End Data of City

Abstract

The present invention provides a method, system, medium and device for constructing roadside data of a city, including: Step S1: Collect urban traffic data based on roadside devices, and perform data classification and fragmented extraction using event rule automation to obtain event data; Step S2: Construct an autonomous driving simulation dynamic scene based on the preprocessed event data; Step S3: Construct a scene feature library according to the key working condition parameters and key traffic flow feature parameters of the dynamic scene, and construct a simulation test scene library. The present invention can effectively improve the utilization rate of roadside data in local autonomous vehicle-road collaboration and enhance the data value.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving testing, and specifically, to a method, system, medium, and device for constructing roadside data about cities. Background Art

[0002] Currently, building an autonomous driving scenario library based on roadside data of the vehicle network has become the key data foundation for testing before the commercial implementation of autonomous driving. And the processing of massive data from roadside devices 24 hours a day, 7 days a week is the most critical problem currently faced, that is, how to obtain valuable data from massive data, how to transform the data into effective test scenarios, and how to drive autonomous driving simulation testing and evaluation through the data flywheel.

[0003] In the published patents on roadside data construction methods, most only focus on how to generate scenarios from roadside data, ignoring how to define the rules for effective scenarios from the perspective of testing, how to extract scenario features based on the produced scenarios, and how to build a city-specific scenario library to realize the data flywheel.

[0004] Patent document CN117036733A discloses a method for extracting characteristic events of urban road scenarios. By summarizing common and reliable event categories in urban road scenarios, a simulation data set containing category labels corresponding to characteristic events is constructed to train a classification model for obtaining road scenario characteristic events. However, this invention does not propose the rule definition of effective scenarios from the perspective of testing. Summary of the Invention

[0005] Aiming at the deficiencies in the prior art, the purpose of the present invention is to provide a method, system, medium, and device for constructing roadside data about cities.

[0006] According to a method for constructing roadside data about cities provided by the present invention, it includes:

[0007] Step S1: Based on the roadside device to collect urban traffic data, use event rule automation to classify and fragment the data to obtain event data;

[0008] Step S2: Based on the preprocessed event data, construct an autonomous driving simulation dynamic scenario;

[0009] Step S3: According to the key working condition parameters and key traffic flow characteristic parameters of the dynamic scenario, construct a scenario feature library and a simulation test scenario library.

[0010] Preferably, in the step S1:

[0011] Step S1.1: Based on the urban traffic data collected by the roadside device, including:

[0012] Step S1.1.1: Classify and divide the points and coverage areas of roadside devices, determine the coverage of local urban road conditions, and set the road condition type and road condition weight for the roadside device points;

[0013] Step S1.1.2: Define the content of target-level urban traffic data, including the target recognition time, target unique identification code, target type, target size, target speed, and target location;

[0014] Step S1.2: Event rules, including:

[0015] Step S1.2.1: Define event types, including working condition type events, traffic flow type events, and autonomous vehicle operation type events;

[0016] Step S1.2.2: Define the rules corresponding to event types. The rules include: time range rule, coverage range rule, and valid event type rule;

[0017] Step S1.2.3: Other event characteristics, including collection time, traffic characteristics, and types of traffic participants;

[0018] Step S1.3: Classification and fragmented extraction of data, including:

[0019] Step S1.3.1: Classify, identify, and extract data according to event types and save them as event data;

[0020] Step S1.3.2: Different event types correspond to different data time ranges and data coverage ranges;

[0021] Event data, including: collection time, target unique identification code, type, location, size, orientation, and speed.

[0022] Preferably, in the said Step S2:

[0023] Step S2.1: Preprocessing of event data, including: data format standardization, data temporalization, trajectory smoothing, speed smoothing, target jump tracking and fusion, deletion of invalid target data, and target size matching;

[0024] Step S2.2: Construction of dynamic scenarios. Use data tools and, based on the simulation test platform and scene production process specifications, process the preprocessed event data into simulation scenarios and corresponding tagged scenarios, including: automatic scene generation, secondary scene processing, and scene rationality verification.

[0025] Preferably, in the said Step S3:

[0026] Step S3.1: Scene feature extraction. Use data tools to extract scene features, and conduct statistics and analysis on the scene features. Combine with road condition features to form parameter settings, parameter distributions, and parameter combination configurations, providing a basis for the authenticity design of the simulation test logic scene, including: extraction of traffic participant behavior features, extraction of macroscopic traffic flow features, extraction of key working condition features, and extraction of key road condition features;

[0027] Step S3.2: Construction of the urban simulation test scene library; Based on the scene feature library of real roadside data, construct the urban test scene library, including: construction of the scene library, construction of the scene evaluation plan, construction of the test evaluation report, test plan, and scene configuration;

[0028] Step S3.2.1: The construction of the scene library includes: the urban working condition scene library and the urban traffic flow scene library.

[0029] According to a roadside data construction system for cities provided by the present invention, execute the above-mentioned roadside data construction method for cities, including:

[0030] Event rule module: Quantitatively define event rules as mathematical relationships and formulas of target-level data;

[0031] Data classification and fragmentation extraction module: Process roadside target-level data using the event rule module and automated data extraction algorithms to obtain event data;

[0032] Event data preprocessing module: Process event data using automated data preprocessing algorithms to obtain normalized event data;

[0033] Dynamic scene construction module: Process the preprocessed normalized event data using automated scene construction algorithms to obtain a dynamic scene;

[0034] Dynamic scene key working condition parameter acquisition module: Process the dynamic scene using automated working condition parameter statistical algorithms to obtain the preset key working condition parameters of the dynamic scene;

[0035] Dynamic scene key traffic flow feature parameter acquisition module: Process the dynamic scene using automated traffic feature statistical algorithms to obtain the preset key traffic flow feature parameters of the dynamic scene;

[0036] Urban typical working condition scene library construction module: Construct the preset urban typical working condition scene library according to the preset key working condition parameters of the dynamic scene;

[0037] Urban traffic flow scene library construction module: Construct the urban traffic flow scene library according to the preset key traffic flow feature parameters of the dynamic scene.

[0038] Preferably, in the urban typical working condition scene library construction module:

[0039] The pre-set typical working condition scenario library of the city stores various typical working condition scenarios of the city, including following scenarios, cut-in and cut-out scenarios, lateral offset scenarios of the target vehicle, scenarios of vulnerable traffic participants crossing, obstacle scenarios, intersection scenarios, and special scenarios.

[0040] Preferably, the scenarios of vulnerable traffic participants crossing include: zebra crossing crossing scenarios, non-zebra crossing crossing scenarios, and scenarios emerging after being blocked;

[0041] Obstacle scenarios include: illegal parking and occupying road scenarios, construction and occupying road scenarios;

[0042] Intersection scenarios include: intersection merging scenarios, traffic scenarios passing under different states of traffic lights;

[0043] Special scenarios include: scenarios of mixed traffic of motor vehicles and vulnerable traffic participants, slow-moving scenarios.

[0044] Preferably, in the city traffic flow scenario library construction module:

[0045] The city traffic flow scenario library stores scenarios with different times and different traffic characteristics, where the traffic characteristics cover traffic flow, traffic density, and average vehicle speed, including low density and low flow, and high density and high flow.

[0046] A computer-readable storage medium storing a computer program, when the computer program is executed by a processor, implements the steps of the method for constructing road-end data of the city as described above.

[0047] A device for constructing road-end data of a city according to the present invention includes: a controller;

[0048] The controller includes the computer-readable storage medium storing the computer program, when the computer program is executed by a processor, implements the steps of the method for constructing road-end data of the city as described above; or, the controller includes the system for constructing road-end data of the city as described above.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention truly collects urban traffic data through roadside devices, and performs regular classification and fragmentation extraction on the urban traffic data. According to the key working condition parameters and key traffic flow characteristic parameters of the dynamic scenario, a typical urban working condition scenario library and an urban traffic flow scenario library that are more in line with the local urban characteristics are constructed, serving as the test database for the landing application of autonomous driving with local urban characteristics and the test database for the local urban autonomous driving road test management specification, improving the response ability of the autonomous driving system to local characteristic working conditions and promoting the commercial landing application of the autonomous driving system in the local area. At the same time, it can also effectively improve the utilization rate of roadside data in the local autonomous driving vehicle-road collaboration and enhance the data value. Brief Description of the Drawings

[0051] By reading the following detailed description of non-limiting embodiments with reference to the accompanying drawings, other features, objects, and advantages of the present invention will become more apparent:

[0052] Figure 1 It is a schematic diagram of the construction method of the present invention;

[0053] Figure 2 It is a schematic structural diagram of the construction device of the present invention (the part within the dashed box). Detailed Embodiments

[0054] The present invention will be described in detail below with reference to specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that those of ordinary skill in the art can make several changes and improvements without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0055] Embodiment 1:

[0056] Aiming at the development trend of the existing scenario library and the deficiencies of the existing technology patents, the present invention proposes a method, device, equipment, and storage medium for constructing roadside data in the city, starting from the rule definition of effective scenarios, to the preprocessing of data and the construction of scenarios, and finally proposing a city-specific test scenario library and a scenario test evaluation plan based on scenario characteristics, forming a complete data flywheel and test closed-loop based on real roadside data. The present invention improves the utilization efficiency of a large amount of roadside data and the automation degree of simulation scenario production based on roadside data, thereby improving the authenticity and complexity of the simulation scenario. Using the simulation scenario constructed based on real data for intelligent networked vehicle simulation testing effectively replaces part of the road test and improves the safety guarantee of urban road driving when facing commercial landing applications.

[0057] According to a method for constructing roadside data in the city provided by the present invention, as Figure 1 - Figure 2 shown, it includes:

[0058] Step S1: Based on the urban traffic data collected by roadside devices, use event rules to automatically classify and fragment the data to obtain event data;

[0059] Specifically, in the said step S1:

[0060] Step S1.1: Based on the urban traffic data collected by roadside devices, including:

[0061] Step S1.1.1: Classify and divide the positions and coverage ranges of roadside devices, determine the coverage of local urban road conditions, and set road condition types and road condition weights for the positions of roadside devices;

[0062] Set road condition types for the positions of roadside devices according to road elements, form road condition type codes, and assign attributes to the positions of roadside devices. Road elements include: the number of lanes, the central separation method, and the number of roads merging at intersections (for example, 4 roads merge at a crossroads). The road condition type code is encoded according to

"intersection code" "number of lanes in each of the four directions of east, south, west, and north" "central separation method code in each of the four directions of east, south, west, and north"

[0063] Intersection code table

[0064]

[0065] Central separation method code table

[0066]

[0067] When the road condition type codes assigned to the intersection position devices are the same encoding value, or can achieve consistent encoding through rotation (for example, for a crossroads with four lanes in the east-west direction and two lanes in the north-south direction 44242111, and a crossroads with four lanes in the north-south direction and two lanes in the east-west direction 424241111, they can be made consistent by rotating 90 degrees, that is, the digits are shifted backward. These two intersections can be regarded as having the same encoding value and the same road condition characteristics), it indicates that the road condition characteristics of this intersection are highly consistent and can be regarded as intersections under the same type of intersection.

[0068] By counting the intersection type code values of the position devices in the entire area and according to the distribution of the number of intersection characteristics in the area, a regional road condition weight table can be formed according to the regional characteristics, and the road condition weights can be taken into account when analyzing the regional traffic flow characteristics.

[0069] The calculation formula for road condition weight is: the number of points with the same type of coding value / the total number of points. For example, when there are 20 intersections in the project area this time, and 3 of them are crossroads with four lanes in the east-west direction and two lanes in the north-south direction (44242111), which have the same coding value. Therefore, the weight of the road condition obtained by the roadside equipment at the intersection with the coding value of 44242111 is 4 / 20 = 20%.

[0070] Step S1.1.2: Define the content of the target-level urban traffic data, including the target recognition time, the unique identification code of the target, the target type, the target size, the target speed, and the target location;

[0071] Step S1.2: Event rules, including:

[0072] Step S1.2.1: Define event types, including working condition type events, traffic flow type events, and autonomous driving vehicle operation type events;

[0073] Step S1.2.2: Define the rules corresponding to the event types, and the rules include: time range rule, coverage range rule, and valid event type rule;

[0074] Step S1.2.3: Other event characteristics, including collection time, traffic characteristics, and types of traffic participants;

[0075] Step S1.3: Classification and fragmented extraction of data, including:

[0076] Step S1.3.1: Classify, identify, and extract data according to event types and save them as event data;

[0077] Step S1.3.2: Different event types correspond to different data time ranges and data coverage ranges;

[0078] Step S1.4: Event data, including: collection time, unique identification code of the target, type, location, size, orientation, and speed.

[0079] Step S2: Construct an autonomous driving simulation dynamic scenario based on the preprocessed event data;

[0080] Specifically, in the step S2:

[0081] Step S2.1: Preprocessing of event data, including: data format standardization, data time series, trajectory smoothing, speed smoothing, target jump tracking and fusion, deletion of invalid target data, and target size matching;

[0082] Step S2.2: Construction of dynamic scenarios. Use a data tool (since roadside data is regularized data, such as in JSON format, and dynamic scenarios are also regularized data, such as in OpenSCENARIO format, so according to personal ability and habits, the roadside data can be written into the dynamic scenario by writing Python scripts or C++ scripts). According to the simulation test platform and scenario production process specifications, process the preprocessed event data into a simulation scenario and label the corresponding scenario, including: automatic scenario generation, secondary scenario processing, and scenario rationality verification.

[0083] Among them, the data tool realizes writing roadside data into the dynamic scenario by writing Python scripts or C++ scripts. The specific process includes: Data collection: Real-time acquisition of data from sensors, databases, or other data sources; Data processing: Cleaning, transforming, and formatting the collected data for subsequent use; Scenario construction: Constructing a dynamic scenario based on the processed data, including drawing graphics, updating UI elements, or simulating physical behaviors; Real-time update: Regularly or triggered by events to update the scenario to keep it synchronized with real-time data; User interaction: Processing user input, including mouse clicks and keyboard keystrokes, to change the scenario or obtain more information.

[0084] Step S3: Construct a scenario feature library based on the key working condition parameters and key traffic flow characteristic parameters of the dynamic scenario, and construct a simulation test scenario library.

[0085] Specifically, in the said Step S3:

[0086] Step S3.1: Scenario feature extraction. Use a data tool (write the required Python script or C++ script according to personal ability in accordance with the above rules and requirements), extract scenario features, statistically analyze the scenario features, and combine with road condition features to form parameter settings, parameter distributions, and parameter combination configurations, providing a basis for the authenticity design of the simulation test logic scenario, including: extraction of traffic participant behavior characteristics, extraction of macroscopic traffic flow characteristics, extraction of key working condition characteristics, and extraction of key road condition characteristics;

[0087] Among them, the data tool writes roadside data into the dynamic scene by writing Python scripts or C++ scripts; specifically including: data preprocessing: using Python libraries including Pandas, NumPy or C++ standard libraries to clean, format and transform the collected data; data visualization: using libraries including Matplotlib, Folium, Pyecharts to visualize the processed data into a dynamic scene; data writing into the dynamic scene: using web frameworks including Flask, Django or real-time data push technologies including WebSocket to push the processed data to the front end; the front end uses technologies including JavaScript, HTML5 Canvas to dynamically render the data as a scene;

[0088] Among them, feature selection needs to select the most suitable feature extraction algorithms according to specific application scenarios, including SIFT, SURF; before feature extraction, image preprocessing needs to be performed on the image, including denoising, enhancing contrast, adjusting the size, and improving the effect of feature extraction; using the visualization function to display key points and descriptors to deeply understand the process and results of feature extraction;

[0089] Step S3.2: Construction of the urban simulation test scenario library; based on the scenario feature library of real roadside data, construct the urban test scenario library, including: construction of the scenario library, construction of the scenario evaluation plan, construction of the test evaluation report, test plan and scenario configuration;

[0090] Step S3.2.1: The construction of the scenario library includes: the urban driving condition scenario library and the urban traffic flow scenario library.

[0091] Embodiment 2:

[0092] Embodiment 2 is a preferred example of Embodiment 1 to more specifically illustrate the present invention.

[0093] The present invention also provides a roadside data construction system for cities, and the roadside data construction system for cities can be implemented by executing the process steps of the roadside data construction method for cities, that is, those skilled in the art can understand the roadside data construction method for cities as the preferred implementation manner of the roadside data construction system for cities.

[0094] According to a roadside data construction system for cities provided by the present invention, executing the above-mentioned roadside data construction method for cities includes:

[0095] Event rule module: Quantitatively define event rules as mathematical relationships and formulas of target-level data;

[0096] The event rule module includes: condition selection of the affected vehicle (i.e., the host vehicle) in the event, condition selection of the key influencing vehicle (i.e., the key target vehicle) in the event, and event rule data screening. Among them, the condition selection of the host vehicle and the key target vehicle is not mandatory and can be used in combination with the event rules.

[0097] Condition selection of the affected vehicle (i.e., the host vehicle) in the event: The conditions include the vehicle type of the host vehicle, the minimum duration requirement for the host vehicle to exist in the data segment, the minimum mileage requirement for the host vehicle to travel in the data segment, and the minimum proportion requirement for the non-stationary time of the host vehicle in the data segment. These four conditions can all be customized according to the test requirements. The default values are that the vehicle type of the host vehicle is a sedan, the minimum duration requirement for the host vehicle to exist is 15 seconds, the minimum mileage requirement for the host vehicle to travel is 100 meters, and the minimum proportion of the non-stationary duration of the host vehicle is 50%. All the targets in the data segment are screened according to the conditions. Each unique ID is regarded as 1 target. The vehicle that meets the conditions can be selected as the host vehicle. When the condition selection of the host vehicle is checked and set, each condition can be set individually or jointly. Only the host vehicle that meets the conditions can be transferred to the next event rule data screening.

[0098] Condition selection of the key influencing vehicle (i.e., the key target vehicle) in the event: The conditions include the vehicle type of the target vehicle, the minimum duration requirement for the target vehicle to exist in the data segment, and the minimum mileage requirement for the target vehicle to travel in the data segment. These three conditions can all be customized according to the test requirements. The default values are that the vehicle type of the target vehicle is a sedan, the minimum duration requirement for the target vehicle to exist is 5 seconds, and the minimum mileage requirement for the target vehicle to travel is 20 meters. All the targets in the data segment are screened according to the conditions. Each unique ID is regarded as 1 target. The vehicle that meets the conditions can be selected as the key target vehicle. When the condition selection of the key target vehicle is checked and set, each condition can be set individually or jointly. Only the key target vehicle that meets the conditions can be transferred to the next event rule data screening.

[0099] Event rule data screening includes: congestion in the driving area of the host vehicle, existence of a specific target vehicle close to the host vehicle, and change in the external input conditions of the host vehicle.

[0100] The basic principle of event rule data screening is: The vehicles that meet the condition selection of the host vehicle are used as the host vehicle one by one for data processing. Data processing requires converting the coordinates of other vehicles into the coordinates of the host vehicle coordinate system. The xy values in the host vehicle coordinate system are used as the calculation basis for the horizontal and vertical movement relationships relative to the host vehicle (x corresponds to the horizontal distance, y corresponds to the vertical distance). At the same time, the speed also needs to be converted into the speed components in the host vehicle coordinate system (vx represents the horizontal speed, vy represents the vertical speed). All subsequent data calculations of the event rules are based on the host vehicle coordinate system.

[0101] (1)Congestion in the driving area of the host vehicle: Taking the coordinates of the host vehicle as the center of the bottom side of a rectangle, with the given length TJ_Length and width TJ_Width of the rectangle for driving area determination, this rectangle is the future forward driving area of the host vehicle. Given the time TJ_Duration for maintaining the calculation state, determine whether the xy values of the target vehicle in the host vehicle coordinate system fall within the rectangular driving area and satisfy the condition that the falling time ≥ 0.5 * TJ_Duration. If satisfied, then the target vehicle is in the front driving area of the host vehicle. Calculate the number of target vehicles that meet the conditions and determine whether the number meets the vehicle threshold for congestion conditions. The recommended calculation method for the vehicle congestion threshold is as follows: According to the standard of a sedan with a length of 5 meters and a width of 2 meters, the number of vehicles that can exist longitudinally is TJ_Length / 5, and the number of vehicles that can exist horizontally is TJ_Width / 2. Considering vehicle parking gaps, etc., the threshold is recommended to be set as [0.5 * (TJ_Length / 5)] * [0.5 * TJ_Width / 2]. Slide the window for each host vehicle according to the given TJ_Duration for the entire data segment. When the number of target vehicles in a certain TJ_Duration segment of a certain host vehicle ≥ the vehicle congestion threshold, then this time segment of this host vehicle is the data segment extracted by the event rule.

[0102] (2)There is a specific target vehicle relatively close to the host vehicle: Taking the coordinates of the host vehicle as the center, calculate the Euclidean distance from the target object to the host vehicle, and give the setting of the minimum Euclidean distance and the setting of the target object type. When a certain frame of the target vehicle meets the requirement of the minimum Euclidean distance, then the two seconds before and after this time point of this target vehicle are the data segments that need to be extracted by the event rule.

[0103] (3)Change in the external input conditions of the host vehicle: According to the lateral / longitudinal speed of the target object in the host vehicle coordinate system mentioned above, subtract the speed of the host vehicle to convert it into the relative lateral / longitudinal speed, and the lateral / longitudinal distance of the target vehicle in the host vehicle coordinate system is the relative lateral / longitudinal distance. According to the given change rates of lateral and longitudinal distances and the change rates of lateral and longitudinal speeds (the default recommended values are the relative lateral distance change rate of 1 m / s, the relative longitudinal distance change rate of 2 m / s, and the relative lateral / longitudinal speed change rate of 1.5 m / s 2 ), calculate the state change situation of the target vehicle in each frame. When the target vehicle meets the condition of the change rate in a continuous given number of frames, then this time period of this target vehicle is the data segment that needs to be extracted by the event rule.

[0104] Data classification and fragmentation extraction module: Use the event rule module and the automated data extraction algorithm to process the road-end target-level data to obtain event data;

[0105] Event data preprocessing module: Use the automated data preprocessing algorithm to process the event data to obtain normalized event data;

[0106] Dynamic scene construction module: Processes the preprocessed normalized event data using an automated scene construction algorithm to obtain a dynamic scene;

[0107] Dynamic scene key operating condition parameter acquisition module: Processes the dynamic scene using an automated operating condition parameter statistical algorithm to obtain the preset key operating condition parameters of the dynamic scene;

[0108] Dynamic scene key traffic flow feature parameter acquisition module: Processes the dynamic scene using an automated traffic feature statistical algorithm to obtain the preset key traffic flow feature parameters of the dynamic scene;

[0109] Urban typical operating condition scene library construction module: Constructs the preset urban typical operating condition scene library according to the preset key operating condition parameters of the dynamic scene;

[0110] Urban traffic flow scene library construction module: Constructs the urban traffic flow scene library according to the preset key traffic flow feature parameters of the dynamic scene.

[0111] Specifically, in the urban typical operating condition scene library construction module:

[0112] The preset urban typical operating condition scene library stores various preset urban typical operating condition scenes such as following scenes, cut-in and cut-out scenes, target vehicle lateral offset scenes, vulnerable traffic participant crossing scenes, obstacle scenes, intersection scenes, and special scenes.

[0113] Specifically, the vulnerable traffic participant crossing scene includes: zebra crossing crossing scene, non-zebra crossing crossing scene, and scene appearing after being blocked;

[0114] The obstacle scene includes: illegal parking and occupying lane scene, construction and occupying lane scene;

[0115] The intersection scene includes: intersection merging scene, traffic signal different state passing scene;

[0116] The special scene includes: mixed traffic scene of motor vehicles and vulnerable traffic participants, slow-moving scene.

[0117] Specifically, in the urban traffic flow scene library construction module:

[0118] The urban traffic flow scene library stores scenes with different times and different traffic characteristics, where the traffic characteristics cover traffic flow, traffic density, and average vehicle speed, including low density and low flow and high density and high flow.

[0119] According to a computer-readable storage medium storing a computer program provided by the present invention, when the computer program is executed by a processor, the steps of the method for constructing roadside data of a city are implemented.

[0120] A road-end data construction device for a city provided by the present invention includes: a controller;

[0121] The controller includes the computer-readable storage medium storing the computer program, and when the computer program is executed by the processor, the steps of the road-end data construction method for the city are implemented; or, the controller includes the road-end data construction system for the city.

[0122] Embodiment 3:

[0123] Embodiment 3 is a preferred example of Embodiment 1 to more specifically illustrate the present invention.

[0124] The present invention adopts the following technical solutions:

[0125] In a first aspect, the present invention provides a method for constructing road-end data for a city. Among them, the method for constructing a dynamic scenario of the urban road-end data includes:

[0126] 1. Based on the urban traffic data collected by the road-end device, data classification and fragmentation extraction are automatically performed using event rules to obtain event data.

[0127] 1.1. Based on the urban traffic data collected by the road-end device, including:

[0128] 1.1.1. Classify and divide the positions and coverage ranges of the road-end devices, determine the coverage of typical road conditions in the local city, and set road condition types and road condition weights for the road-end device positions.

[0129] 1.1.2. Define the content of the target-level urban traffic data, including at least the target recognition time (such as UNIX time), the unique identification code of the target, the target type, the target size, the target speed, and the target position.

[0130] 1.2. Event rules, including:

[0131] 1.2.1. Define event types. Include working condition type events, traffic flow type events, and autonomous driving vehicle operation type events.

[0132] 1.2.2. Define the rules corresponding to the event types. The rules include: time range rule, coverage range rule, and valid event type rule.

[0133] 1.2.3. Other event characteristics. Include collection time, traffic characteristics (traffic flow, traffic density), and traffic participant types.

[0134] 1.3. Data classification and fragmentation extraction, including:

[0135] 1.3.1. Classify, identify, and extract data according to event types and save them as event data.

[0136] 1.3.2. Different event types correspond to different data time ranges and data coverage ranges.

[0137] 1.4. Event data includes: acquisition time, unique identification code of the target, type, location, size, orientation, speed.

[0138] 2. Construct an autonomous driving simulation dynamic scenario based on the preprocessed event data.

[0139] 2.1. Preprocessing of event data includes: data format standardization, data serialization, trajectory smoothing, speed smoothing, target jump tracking and fusion, deletion of invalid target data, and target size matching.

[0140] 2.2. Construction of the dynamic scenario. Using specific data tools, according to the simulation test platform and scenario production process specifications, process the preprocessed event data into a scenario available for simulation and label the corresponding scenario. It includes: automatic scenario generation, secondary scenario processing, and scenario rationality verification.

[0141] 3. Construct a scenario feature library based on the key working condition parameters and key traffic flow characteristic parameters of the dynamic scenario, and construct a simulation test scenario library with urban characteristics.

[0142] 3.1. Scenario feature extraction. Using specific data tools, extract scenario features, statistically analyze the scenario features, and combine with road condition features to form real and reasonable parameter settings, parameter distributions, and parameter combination configurations, providing a basis for the authenticity design of the simulation test logic scenario. It includes: extraction of traffic participant behavior characteristics, extraction of macroscopic traffic flow characteristics, extraction of key working condition characteristics, extraction of key road condition characteristics, etc.

[0143] 3.2. Construction of a simulation test scenario library with urban characteristics. Based on the scenario feature library of real roadside data, construct a test scenario library with urban characteristics to facilitate the commercial implementation of intelligent connected vehicles in the local area. It includes: construction of the scenario library, construction of the scenario evaluation plan, construction of the test evaluation report, test plan and scenario configuration, etc.

[0144] 3.2.1. The construction of the scenario library includes: an urban typical working condition scenario library and an urban traffic flow scenario library.

[0145] In the second aspect, the present invention provides a roadside data construction device for a city. Among them, the dynamic scenario construction device for the urban roadside data includes:

[0146] 1. An event rule module, configured to quantitatively define event rules as mathematical relationships and formulas of target-level data.

[0147] 2. A data classification and fragmentation extraction module, configured to process roadside target-level data by using an event rule module and an automated data extraction algorithm to obtain event data.

[0148] 3. An event data preprocessing module, configured to process the event data by using an automated data preprocessing algorithm to obtain normalized event data.

[0149] 4. A dynamic scenario construction module, configured to process the preprocessed normalized event data by using an automated scenario construction algorithm to obtain a dynamic scenario.

[0150] 5. A dynamic scenario key working condition parameter acquisition module, configured to process the dynamic scenario by using an automated working condition parameter statistical algorithm to obtain dynamic scenario key working condition parameters.

[0151] 6. A dynamic scenario key traffic flow feature parameter acquisition module, configured to process the dynamic scenario by using an automated traffic feature statistical algorithm to obtain dynamic scenario key traffic flow feature parameters.

[0152] 7. An urban typical working condition scenario library construction module, configured to construct an urban typical working condition scenario library according to the dynamic scenario key working condition parameters; the urban typical working condition scenario library stores various urban typical working condition scenarios such as a following scenario, a cut-in and cut-out scenario, a target vehicle lateral offset scenario, a vulnerable traffic participant crossing scenario (a zebra crossing crossing scenario, a non-zebra crossing crossing scenario, an appearance scenario after being blocked), an obstacle scenario (an illegal parking and road occupation scenario, a construction road occupation scenario, etc.), an intersection scenario (an intersection merging scenario, a traffic signal different state passing scenario), a special scenario (a mixed traffic scenario of motor vehicles and vulnerable traffic participants, a slow-moving scenario), etc.

[0153] 8. An urban traffic flow scenario library construction module, configured to construct an urban traffic flow scenario library according to the dynamic scenario key traffic flow feature parameters; the urban traffic flow scenario library stores scenarios with different times and different traffic features, where the traffic features cover traffic flow, traffic density, and average vehicle speed, such as low density and low flow (mostly occurring at night), high density and high flow (mostly occurring during the morning and evening rush hours on weekdays).

[0154] In addition, to achieve the above object, the present invention also proposes a scenario construction device, where the scenario construction device includes: a memory, a processor, and a scenario construction program stored on the memory and executable on the processor, and when the scenario construction program is executed by the processor, the above-mentioned scenario construction method is implemented.

[0155] In addition, to achieve the above object, the present invention further provides a storage medium, on which a scene construction program is stored. When the scene construction program is executed by a processor, the above-mentioned scene construction method is implemented.

[0156] The method and device proposed by this method are both implemented by writing in the Python language. During the implementation process, we adopted Python 3.9 as the interpreter and used PyCharm as the main running software. To implement this method, we need the following Python packages: pandas, numpy, xml.etree.ElementTree, scipy, math, and pyproj. These packages provide us with functions such as data processing, mathematical calculations, and geographic information processing.

[0157] Those skilled in the art know that in addition to implementing the system and its various devices, modules, and units provided by the present invention in the form of pure computer-readable program code, the method steps can be logically programmed to enable the system and its various devices, modules, and units provided by the present invention to be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers, etc. to achieve the same function. Therefore, the system and its various devices, modules, and units provided by the present invention can be regarded as a kind of hardware component, and the devices, modules, and units included therein for implementing various functions can also be regarded as the structure within the hardware component; the devices, modules, and units for implementing various functions can also be regarded as either software modules for implementing the method or the structure within the hardware component.

[0158] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for constructing road-end data of a city, characterized in that, Including: Step S1: Based on the urban traffic data collected by roadside devices, use event rules to automatically classify and fragment the data to obtain event data; Step S2: Based on the preprocessed event data, construct a dynamic scenario for autonomous driving simulation; Step S3: According to the key working condition parameters and key traffic flow characteristic parameters of the dynamic scenario, construct a scenario feature library and a simulation test scenario library; In the said Step S1: Step S1.1: Based on the urban traffic data collected by roadside devices, including: Step S1.1.1: Classify and divide the positions and coverage ranges of roadside devices, determine the coverage of the local urban road conditions, and set road condition types and road condition weights for the roadside device positions; Set road condition types for roadside device positions according to road elements, form road condition type codes, and assign attributes to roadside device positions; the road elements include: the number of lanes, the central separation method, and the number of roads joining at intersections; The road condition type code is encoded according to "intersection code", "the number of lanes in each of the four directions of east, south, west, and north", and "the code of the central separation method in each of the four directions of east, south, west, and north"; By statistically analyzing the intersection type code values of the point devices in the entire area, according to the distribution of the number of intersection characteristics in the area, form a regional road condition weight table according to the regional characteristics; The road condition weight calculation formula is: the number of point positions with the same type of code value / the total number of point positions; Step S1.1.2: Define the content of the target-level urban traffic data, including the target recognition time, the unique identification code of the target, the target type, the target size, the target speed, and the target position; Step S1.2: Event rules, including: Step S1.2.1: Define event types, including working condition type events, traffic flow type events, and autonomous driving vehicle operation type events; Step S1.2.2: Define the rules corresponding to the event types, and the rules include: time range rules, coverage range rules, and valid event type rules; Step S1.2.3: Other event characteristics, including collection time, traffic characteristics, and types of traffic participants; Step S1.3: Classification and fragmentation extraction of data, including: Step S1.3.1: Classify, identify, and extract data according to event types and save it as event data; Step S1.3.2: Different event types correspond to different data time ranges and data coverage ranges; Step S1.4: Event data, including: collection time, unique identification code of the target, type, position, size, orientation, and speed; In the said Step S2: Step S2.1: Preprocessing of event data, including: data format standardization, data time sequencing, trajectory smoothing, speed smoothing, target jump tracking and fusion, deletion of invalid target data, and target size matching; Step S2.2: Construction of the dynamic scenario, using data tools, according to the simulation test platform and scenario production process specifications, process the preprocessed event data into a simulation scenario and corresponding label the scenario, including: scenario automatic generation, scenario secondary processing, and scenario rationality verification; Among them, the data tool realizes writing roadside data into the dynamic scenario by writing Python scripts or C++ scripts. The specific process includes: Data collection: Real-time acquisition of data from sensors, databases or other data sources; Data processing: Cleaning, transforming and formatting the collected data for subsequent use; Scenario construction: Constructing a dynamic scenario based on the processed data, including drawing graphics, updating UI elements or simulating physical behaviors; Real-time update: Regularly or triggered by events to update the scenario to keep it synchronized with the real-time data; User interaction: Processing user input, including mouse clicks and keyboard key presses, to change the scenario or obtain more information; In the step S3: Step S3.1: Scene feature extraction. Using the data tool, extract scene features, and perform statistics and analysis on the scene features. Combining with road condition features, form parameter settings, parameter distributions and parameter combination configurations, providing a basis for the authenticity design of the simulation test logic scenario, including: Extraction of traffic participant behavior features, extraction of macroscopic traffic flow features, extraction of key operating condition features, extraction of key road condition features; Among them, the data tool realizes writing roadside data into the dynamic scenario by writing Python scripts or C++ scripts; specifically including: Data preprocessing: Using Python libraries including Pandas, NumPy or C++ standard libraries to clean, format and transform the collected data; Data visualization: Using libraries including Matplotlib, Folium, Pyecharts to visualize the processed data as a dynamic scenario; Writing data into the dynamic scenario: Using web frameworks including Flask, Django or real-time data push technologies including WebSocket to push the processed data to the front end; The front end uses technologies including JavaScript, HTML5 Canvas to dynamically render the data as a scenario; Among them, feature selection needs to select the most suitable feature extraction algorithm according to the specific application scenario, including SIFT, SURF; Before feature extraction, image preprocessing needs to be performed on the image, including denoising, enhancing contrast, and adjusting the size to improve the effect of feature extraction; Using the visualization function, display key points and descriptors to deeply understand the process and results of feature extraction; Step S3.2: Construction of the urban simulation test scenario library; Based on the scenario feature library of real roadside data, construct the urban test scenario library, including: Construction of the scenario library, construction of the scenario evaluation plan, construction of the test evaluation report, test plan and scenario configuration; Step S3.2.1: The construction of the scenario library includes: Urban operating condition scenario library and urban traffic flow scenario library.

2. A road-end data construction system for a city, characterized in that, Execute the method for constructing roadside data of a city described in claim 1, including: Event rule module: Quantitatively define event rules as mathematical relationships and formulas of target-level data; Data classification and fragmentation extraction module: Use the event rule module and automated data extraction algorithms to process roadside target-level data to obtain event data; Event data preprocessing module: Processes event data using an automated data preprocessing algorithm to obtain normalized event data; Dynamic scenario construction module: Processes the preprocessed normalized event data using an automated scenario construction algorithm to obtain a dynamic scenario; Dynamic scenario key operating condition parameter acquisition module: Processes the dynamic scenario using an automated operating condition parameter statistical algorithm to obtain the key operating condition parameters preset for the dynamic scenario; Dynamic scenario key traffic flow feature parameter acquisition module: Processes the dynamic scenario using an automated traffic feature statistical algorithm to obtain the key traffic flow feature parameters preset for the dynamic scenario; Urban typical operating condition scenario library construction module: Constructs an urban preset typical operating condition scenario library based on the key operating condition parameters preset for the dynamic scenario; Urban traffic flow scenario library construction module: Constructs an urban traffic flow scenario library based on the key traffic flow feature parameters preset for the dynamic scenario.

3. The road-end data construction system for cities according to claim 2, characterized in that, In the urban typical operating condition scenario library construction module: The urban preset typical operating condition scenario library stores various urban preset typical operating condition scenarios such as following scenarios, cut-in and cut-out scenarios, target vehicle lateral offset scenarios, vulnerable traffic participant crossing scenarios, obstacle scenarios, intersection scenarios, and special scenarios.

4. The road-end data construction system for a city according to claim 3, wherein: The vulnerable traffic participant crossing scenarios include: zebra crossing crossing scenarios, non-zebra crossing crossing scenarios, and scenarios that appear after being blocked; The obstacle scenarios include: illegal parking and occupying the road scenarios, construction and occupying the road scenarios; The intersection scenarios include: intersection merging scenarios, traffic signal different state passing scenarios; The special scenarios include: scenarios of mixed traffic of motor vehicles and vulnerable traffic participants, slow-moving scenarios.

5. The road-end data construction system for cities according to claim 2, characterized in that, In the urban traffic flow scenario library construction module: The urban traffic flow scenario library stores scenarios with different times and different traffic characteristics, where the traffic characteristics cover traffic flow, traffic density, and average vehicle speed, including low density and low flow, and high density and high flow.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the road-end data construction method for a city according to claim 1.

7. A road-end data construction device for a city, characterized in that, Including: A controller; The controller includes the computer-readable storage medium storing the computer program according to claim 6, and when the computer program is executed by a processor, it implements the steps of the road-end data construction method for a city according to claim 1; or, the controller includes the road-end data construction system for a city according to any one of claims 2 to 5.

Images