Automatic driving simulation test scene random generation method and system

Abstract

The application provides an automatic driving simulation test scene random automatic generation method and system, comprising: collecting real map data and real scene data in a real road environment by using a data acquisition device; analyzing the real map data to obtain a plurality of basic map elements, counting the occurrence probability and occurrence position of each basic map element, and obtaining a map element set; analyzing the real scene data to obtain a plurality of basic scene elements, counting the occurrence probability and occurrence position of each basic scene element, and obtaining a scene element set; obtaining map configuration information and scene configuration information; automatically generating a target map according to the map configuration information; automatically generating a target scene according to the scene configuration information; and automatically driving simulation testing a target vehicle in the target scene. The application can quickly and randomly generate a large number of test maps and test scenes, realize automatic construction of a simulation scene, and does not need manual editing.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving simulation testing technology, and more specifically, to a method and system for automatically generating random autonomous driving simulation testing scenarios. Background Technology

[0002] Before being officially put into public operation, autonomous vehicles must undergo testing to verify the normality and completeness of their autonomous driving functions. Testing autonomous vehicles is a challenging engineering task. While real-road testing is feasible, it carries high costs and risks. Therefore, software simulation is often used for autonomous vehicle testing. The effectiveness of autonomous driving simulation testing depends on the richness and realism of the test scenarios. Only test scenarios with broad coverage and high realism can guarantee the reliability and efficiency of autonomous driving simulation testing.

[0003] Currently, the construction of autonomous driving simulation test scenarios mostly relies on replaying real-world scenarios or manual random editing, resulting in a limited number of scenarios and low construction efficiency.

[0004] Patent CN112789619B discloses a method for constructing a simulation scene, a simulation method, and a device. The method includes: acquiring an instruction indicating a starting point and a destination; acquiring, based on the instruction and map information, the driving route of a vehicle from the starting point to the destination and road information along the route; selecting at least one data segment from a plurality of pre-stored data segments based on the road information, the plurality of data segments being determined based on data collected by a sensor device; stitching the at least one data segment together according to a first preset rule to obtain a stitched data segment; and generating a simulation scene based on the stitched data segment. However, the solution provided by this patent is entirely based on collected actual scene segments, lacking the flexibility and granularity to construct arbitrary scenes. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for automatically generating randomized autonomous driving simulation test scenarios.

[0006] According to the present invention, a method and system for automatically generating randomized autonomous driving simulation test scenarios are provided, the scheme of which is as follows:

[0007] Firstly, a method for automatically generating randomized autonomous driving simulation test scenarios is provided, the method comprising:

[0008] Step S1: Use data acquisition equipment to collect real map data and real scene data in a real road environment;

[0009] Step S2: Analyze the real map data to obtain multiple basic map elements, and calculate the occurrence probability and occurrence location of each basic map element to obtain a set of map elements with probability information and location information; Analyze the real scene data to obtain multiple basic scene elements, and calculate the occurrence probability and occurrence location of each basic scene element to obtain a set of scene elements with probability information and location information.

[0010] Step S3: Obtain map configuration information and scene configuration information;

[0011] Step S4: Based on the map configuration information, randomly select basic map elements that meet the quantity requirements from the map element set, and combine the selected basic map elements according to the first preset rule to automatically generate the target map;

[0012] Step S5: Based on the scene configuration information, randomly select basic scene elements that meet the quantity requirements from the scene element set, and place the selected basic scene elements in the target map according to the second preset rule to automatically generate the target scene;

[0013] Step S6: Conduct autonomous driving simulation tests on the target vehicle in the target scenario.

[0014] Preferably, the basic map elements include lanes, intersections, traffic markings, traffic signs, traffic lights, protective barriers, and bus stops; the lanes are minimum driving lanes, i.e., single lanes, including straight lanes, curved lanes, and transition lanes; the intersections include three-way intersections, four-way intersections, and multi-way intersections; each road includes at least one lane, and the transition lanes are used to connect two roads with different numbers of lanes;

[0015] The probability of a certain basic map element appearing is obtained by dividing the number of a certain basic map element by the total number of all basic map elements. The locations where basic map elements appear include intersections, roadsides, roadside, laneside, and lane center. The map element set contains all basic map elements as well as the probability and location information of each basic map element.

[0016] The basic scene elements include traffic participants and non-traffic participants; traffic participants include motor vehicles, non-motor vehicles and pedestrians, and non-traffic participants include ancillary facilities and obstacles; obstacles include man-made roadblocks and accidental obstacles, and accidental obstacles include large tree branches, large bricks and rubble, and large pieces of garbage;

[0017] The probability of occurrence of a certain basic scene element is obtained by dividing the number of a certain basic scene element by the total number of all basic scene elements. The occurrence locations of basic scene elements include roadside, road center, lane center, and traffic markings. The scene element set contains all basic scene elements as well as probability information and location information for each basic scene element. The scene element set also contains the starting point, ending point, trajectory, and speed information of each traffic participant.

[0018] Preferably, the map configuration information includes the map size and the number of each basic map element;

[0019] The method for determining the number of each basic map element is to first determine the number of intersections, and then, based on the probability information of the map element set, use the probability ratio of each basic map element to determine the number of other basic map elements in turn, or to select them completely randomly.

[0020] The scenario configuration information includes the number of each basic scenario element and the degree of rule compliance for each traffic participant;

[0021] The quantity of each basic map element includes the quantity of each type of traffic participant and the quantity of each type of non-traffic participant. The method for determining the quantity of each type of traffic participant is to first determine the quantity of motor vehicles, and then, based on the probability information of the scene element set, determine the quantity of non-motor vehicles and pedestrians sequentially using the probability ratio of motor vehicles, non-motor vehicles, and pedestrians, or select them completely randomly. The method for determining the quantity of each type of non-traffic participant is to first determine the quantity of ancillary facilities, and then, based on the probability information of the scene element set, determine the quantity of obstacles using the probability ratio of ancillary facilities and obstacles, or select them completely randomly.

[0022] The degree of compliance with the rules for each type of traffic participant ranges from 0 to 1, representing the ratio of the number of participants who comply with the rules to the total number of participants, and indicating the probability that each type of traffic participant complies with the rules.

[0023] Preferably, the first preset rule includes:

[0024] Step T1: Create a blank map based on the map size information;

[0025] Step T2: Randomly place all intersections sequentially on the blank map;

[0026] Step T3: Connect adjacent intersections using lanes to form a road network;

[0027] Step T4: Place traffic markings, traffic signs, traffic lights, isolation and protection facilities, and bus stops to form the target map.

[0028] Preferably, step T2 includes: multiple intersections are arranged horizontally, vertically, or irregularly; adjacent intersections are guaranteed to have at least one opening facing each other; if they cannot face each other, they are guaranteed to be smoothly connected in terms of distance and angle; the distance between two adjacent intersections follows a probability distribution or is determined randomly.

[0029] Step T3 includes: if two adjacent intersections have openings facing each other and the same number of lanes, then connect them with a straight lane; if two adjacent intersections have openings facing each other but different numbers of lanes, then connect them with a straight lane and a transition lane; if two adjacent intersections have openings not facing each other but the same number of lanes, then connect them with a curved lane; if two adjacent intersections have openings not facing each other but different numbers of lanes, then connect them with a curved lane and a transition lane.

[0030] Step T4 includes: placing traffic markings at intersections and on lanes; placing traffic signs at intersections and on the roadside, the traffic signs being suspended above the lanes; placing traffic lights at intersections; placing isolation and protective facilities in the middle of the road and on the roadside; and placing bus stops on the roadside.

[0031] The placement of all basic map elements must conform to the corresponding location information, and the placement of later basic map elements must not conflict with the placement of earlier basic map elements.

[0032] Preferably, the second preset rule includes:

[0033] All motor vehicles, non-motor vehicles, pedestrians, ancillary facilities and obstacles are randomly placed in the target map in sequence, and the placement order of different basic scene elements can be arbitrarily selected;

[0034] The placement of all basic scene elements must conform to the corresponding position information, and the placement of later basic scene elements must not conflict with the placement of earlier basic scene elements.

[0035] Based on the start point, end point, trajectory, and speed information of each traffic participant contained in the scene element set, determine the start point, end point, trajectory, and speed of each motor vehicle, each non-motor vehicle, and each pedestrian in the target map;

[0036] Based on the degree of compliance with the rules by each type of traffic participant, a corresponding number of motor vehicles, non-motor vehicles, and pedestrians are selected and instructed to perform behaviors that violate traffic rules at a preset time, as a sudden unexpected situation.

[0037] Preferably, step S6 includes:

[0038] Step S6.1: Run the target scenario in the simulation environment;

[0039] Step S6.2: Place the target vehicle in the target scene and set the start and end points;

[0040] Step S6.3: Instruct the target vehicle to drive from the starting point to the destination in an autonomous driving mode;

[0041] Step S6.4: Test the autonomous driving performance of the target vehicle based on its performance during driving.

[0042] Secondly, a system for automatically generating randomized autonomous driving simulation test scenarios is provided, the system comprising:

[0043] Data acquisition module: Uses data acquisition equipment to collect real map data and real scene data in a real road environment;

[0044] Element parsing module: Parses the real map data to obtain multiple basic map elements, calculates the occurrence probability and occurrence location of each basic map element, and obtains a set of map elements with probability information and location information; Parses the real scene data to obtain multiple basic scene elements, calculates the occurrence probability and occurrence location of each basic scene element, and obtains a set of scene elements with probability information and location information.

[0045] Configuration information module: Retrieves map configuration information and scene configuration information;

[0046] Map generation module: Based on the map configuration information, randomly selects basic map elements that meet the quantity requirements from the map element set, and combines the selected basic map elements according to the first preset rule to automatically generate the target map;

[0047] Scene generation module: Based on the scene configuration information, randomly selects basic scene elements that meet the quantity requirements from the scene element set, and places the selected basic scene elements in the target map according to the second preset rule to automatically generate the target scene;

[0048] Simulation testing module: Performs autonomous driving simulation tests on the target vehicle in the target scenario.

[0049] Preferably, the basic map elements include lanes, intersections, traffic markings, traffic signs, traffic lights, protective barriers, and bus stops; the lanes are minimum driving lanes, i.e., single lanes, including straight lanes, curved lanes, and transition lanes; the intersections include three-way intersections, four-way intersections, and multi-way intersections; each road includes at least one lane, and the transition lanes are used to connect two roads with different numbers of lanes;

[0050] The probability of a certain basic map element appearing is obtained by dividing the number of a certain basic map element by the total number of all basic map elements. The locations where basic map elements appear include intersections, roadsides, roadside, laneside, and lane center. The map element set contains all basic map elements as well as the probability and location information of each basic map element.

[0051] The basic scene elements include traffic participants and non-traffic participants; traffic participants include motor vehicles, non-motor vehicles and pedestrians, and non-traffic participants include ancillary facilities and obstacles; obstacles include man-made roadblocks and accidental obstacles, and accidental obstacles include large tree branches, large bricks and rubble, and large pieces of garbage;

[0052] The probability of occurrence of a certain basic scene element is obtained by dividing the number of a certain basic scene element by the total number of all basic scene elements. The occurrence locations of basic scene elements include roadside, road center, lane center, and traffic markings. The scene element set contains all basic scene elements as well as probability information and location information for each basic scene element. The scene element set also contains the starting point, ending point, trajectory, and speed information of each traffic participant.

[0053] Preferably, the map configuration information includes the map size and the number of each basic map element;

[0054] The method for determining the number of each basic map element is to first determine the number of intersections, and then, based on the probability information of the map element set, use the probability ratio of each basic map element to determine the number of other basic map elements in turn, or to select them completely randomly.

[0055] The scenario configuration information includes the number of each basic scenario element and the degree of rule compliance for each traffic participant;

[0056] The quantity of each basic map element includes the quantity of each type of traffic participant and the quantity of each type of non-traffic participant. The method for determining the quantity of each type of traffic participant is to first determine the quantity of motor vehicles, and then, based on the probability information of the scene element set, determine the quantity of non-motor vehicles and pedestrians sequentially using the probability ratio of motor vehicles, non-motor vehicles, and pedestrians, or select them completely randomly. The method for determining the quantity of each type of non-traffic participant is to first determine the quantity of ancillary facilities, and then, based on the probability information of the scene element set, determine the quantity of obstacles using the probability ratio of ancillary facilities and obstacles, or select them completely randomly.

[0057] The degree of compliance with the rules for each type of traffic participant ranges from 0 to 1, representing the ratio of the number of participants who comply with the rules to the total number of participants, and indicating the probability that each type of traffic participant complies with the rules.

[0058] The first preset rule includes:

[0059] Step T1: Create a blank map based on the map size information;

[0060] Step T2: Randomly place all intersections in the blank map in sequence; specifically, multiple intersections are arranged horizontally, vertically, or irregularly; adjacent intersections are guaranteed to have at least one opening facing each other, and if they cannot face each other, they are guaranteed to be smoothly connected in terms of distance and angle; the distance between two adjacent intersections follows a probability distribution or is determined randomly.

[0061] Step T3: Connect adjacent intersections using lanes to form a road network; specifically, if two adjacent intersections have openings facing each other and the same number of lanes, connect them with straight lanes; if two adjacent intersections have openings facing each other but different numbers of lanes, connect them with straight lanes and transition lanes; if two adjacent intersections have openings not facing each other but the same number of lanes, connect them with curved lanes; if two adjacent intersections have openings not facing each other but different numbers of lanes, connect them with curved lanes and transition lanes.

[0062] Step T4: Place traffic markings, traffic signs, traffic lights, protective barriers, and bus stops to form a target map; specifically, place traffic markings at intersections and on lanes; place traffic signs at intersections and on the roadside, with the traffic signs suspended above the lanes; place traffic lights at intersections; place protective barriers in the middle of the road and on the roadside; and place bus stops on the roadside.

[0063] The placement of all basic map elements must conform to the corresponding location information, and the placement of later basic map elements must not conflict with the placement of earlier basic map elements.

[0064] The second preset rule includes:

[0065] All motor vehicles, non-motor vehicles, pedestrians, ancillary facilities and obstacles are randomly placed in the target map in sequence, and the placement order of different basic scene elements can be arbitrarily selected;

[0066] The placement of all basic scene elements must conform to the corresponding position information, and the placement of later basic scene elements must not conflict with the placement of earlier basic scene elements.

[0067] Based on the start point, end point, trajectory, and speed information of each traffic participant contained in the scene element set, determine the start point, end point, trajectory, and speed of each motor vehicle, each non-motor vehicle, and each pedestrian in the target map;

[0068] Based on the degree of compliance with the rules by each type of traffic participant, a corresponding number of motor vehicles, non-motor vehicles, and pedestrians are selected and instructed to perform behaviors that violate traffic rules at a preset time, as a sudden unexpected situation.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. Based on map elements and scene elements extracted from actual test data, this invention can randomly and quickly generate a large number of test maps and test scenes. These test scenes are not limited to the restoration or splicing of actual test scenes, but have high flexibility and can be arbitrarily combined and generated according to actual test needs, which greatly enriches the test scene library and has a wider coverage.

[0071] 2. This invention is based on an automated program, which realizes the automated construction of simulation test scenarios without the need for a lot of manual editing, thus saving manpower costs. Attached Figure Description

[0072] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0073] Figure 1 This is an overall flowchart of the method provided in the embodiments of the present invention;

[0074] Figure 2 This is a flowchart illustrating the specific process of the first preset rule in an embodiment of the present invention.

[0075] Figure 3 This is a flowchart of step S6 in an embodiment of the present invention. Detailed Implementation

[0076] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0077] One embodiment of the present invention provides a method for automatically generating randomized autonomous driving simulation test scenarios, such as... Figure 1 As shown, the specific process is as follows:

[0078] Step S1: Use data acquisition equipment to collect real map data and real scene data in a real road environment.

[0079] Step S2: Analyze the real map data to obtain multiple basic map elements, and calculate the occurrence probability and occurrence location of each basic map element to obtain a set of map elements with probability information and location information; Analyze the real scene data to obtain multiple basic scene elements, and calculate the occurrence probability and occurrence location of each basic scene element to obtain a set of scene elements with probability information and location information.

[0080] The basic map elements include lanes, intersections, traffic markings, traffic signs, traffic lights, isolation and protection facilities, and bus stops; lanes are the minimum driving lanes, i.e., single lanes, including straight lanes, curved lanes, and transition lanes; intersections include three-way intersections, four-way intersections, and multi-way intersections; each road includes at least one lane, and transition lanes are used to connect two roads with different numbers of lanes.

[0081] Traffic markings include instruction markings, prohibition markings, and warning markings; traffic signs include warning signs, prohibition signs, instruction signs, guide signs, work signs, notice signs, and auxiliary signs; traffic lights include motor vehicle traffic lights, non-motor vehicle traffic lights, and pedestrian traffic lights; and isolation and protection facilities include separation facilities, crash barriers, and anti-glare facilities.

[0082] The probability of a certain basic map element appearing is obtained by dividing the number of a certain basic map element by the total number of all basic map elements. The locations where basic map elements appear include intersections, roadsides, roadside, laneside, and lane center. The map element set contains all basic map elements as well as the probability and location information of each basic map element.

[0083] The basic scene elements include traffic participants and non-traffic participants; traffic participants include motor vehicles, non-motor vehicles and pedestrians, and non-traffic participants include ancillary facilities and obstacles; obstacles include man-made roadblocks and accidental obstacles, and accidental obstacles include thick tree branches, large bricks and rubble, and large pieces of garbage.

[0084] Motor vehicles include cars, buses, motorcycles, vans, buses, small trucks, large trucks, dump trucks, construction vehicles, and agricultural vehicles; non-motor vehicles include bicycles, electric bicycles, tricycles, and animal-drawn vehicles; ancillary facilities include roadside trees, fire hydrants, billboards, and garbage bins.

[0085] The probability of a certain basic scene element is obtained by dividing the number of a certain basic scene element by the total number of all basic scene elements. The occurrence locations of basic scene elements include roadside, road center, lane center, and traffic markings. The scene element set contains all basic scene elements as well as the probability and location information of each basic scene element. The scene element set also contains the starting point, ending point, trajectory, and speed information of each traffic participant.

[0086] Step S3: Obtain map configuration information and scene configuration information.

[0087] Map configuration information includes the size of the map and the number of each basic map element.

[0088] The method for determining the number of each basic map element is to first determine the number of intersections, and then, based on the probability information of the map element set, use the probability ratio of each basic map element to determine the number of other basic map elements in turn, or to select them completely randomly.

[0089] Scene configuration information includes the number of each basic scene element and the degree of rule compliance for each traffic participant.

[0090] The number of each basic map element includes the number of each type of traffic participant and the number of each type of non-traffic participant. The method for determining the number of each type of traffic participant is to first determine the number of motor vehicles, and then, based on the probability information of the scene element set, determine the number of non-motor vehicles and pedestrians in turn using the probability ratio of motor vehicles, non-motor vehicles and pedestrians, or select them completely randomly. The method for determining the number of each type of non-traffic participant is to first determine the number of ancillary facilities, and then, based on the probability information of the scene element set, determine the number of obstacles using the probability ratio of ancillary facilities and obstacles, or select them completely randomly.

[0091] The compliance level for each type of traffic participant ranges from 0 to 1, representing the ratio of the number of rule-compliant participants to the total number of participants for that type of traffic participant, indicating the probability of rule compliance for that type of traffic participant. The closer the compliance level is to 0, the greater the probability and proportion of that type of traffic participant violating traffic rules.

[0092] Step S4: Based on the map configuration information, randomly select basic map elements that meet the quantity requirements from the map element set, and combine the selected basic map elements according to the first preset rule to automatically generate the target map.

[0093] Specifically, such as Figure 2 As shown, the first preset rule includes:

[0094] Step T1: Create a blank map based on the map size information;

[0095] Step T2: Randomly place all intersections on a blank map in sequence.

[0096] Multiple intersections can be arranged horizontally, vertically, or irregularly; adjacent intersections should ideally have at least one opening facing each other directly. If they cannot face each other directly, they should be smoothly connected by distance and angle; the distance between two adjacent intersections follows a probability distribution or is determined randomly.

[0097] Step T3: Connect adjacent intersections using lanes to form a road network.

[0098] If two adjacent intersections have openings facing each other and the same number of lanes, they are connected by a straight lane; if two adjacent intersections have openings facing each other but different numbers of lanes, they are connected by a straight lane and a transition lane; if two adjacent intersections have openings not facing each other but the same number of lanes, they are connected by a curved lane; if two adjacent intersections have openings not facing each other but different numbers of lanes, they are connected by a curved lane and a transition lane.

[0099] Step T4: Place traffic markings, traffic signs, traffic lights, isolation and protection facilities, and bus stops to form the target map.

[0100] Place traffic markings at intersections and on lanes; place traffic signs at intersections and on the roadside, with the signs suspended above the lanes; place traffic lights at intersections; place protective barriers in the middle of the road and on the roadside; and place bus stops on the roadside.

[0101] The placement of all basic map elements must conform to the corresponding location information, and the placement of later basic map elements must not conflict with the placement of earlier basic map elements.

[0102] Step S5: Based on the scene configuration information, randomly select basic scene elements that meet the quantity requirements from the scene element set, and place the selected basic scene elements in the target map according to the second preset rule to automatically generate the target scene.

[0103] The second preset rule includes:

[0104] All motor vehicles, non-motor vehicles, pedestrians, ancillary facilities and obstacles are randomly placed in the target map in sequence, and the placement order of different basic scene elements can be arbitrarily selected;

[0105] The placement of all basic scene elements must conform to the corresponding position information, and the placement of later basic scene elements must not conflict with the placement of earlier basic scene elements.

[0106] Based on the start point, end point, trajectory, and speed information of each traffic participant contained in the scene element set, determine the start point, end point, trajectory, and speed of each motor vehicle, each non-motor vehicle, and each pedestrian in the target map;

[0107] Based on the degree of compliance with traffic rules by each type of traffic participant, a corresponding number of motor vehicles, non-motor vehicles, and pedestrians are selected and instructed to engage in behaviors that violate traffic rules at preset times, as a sudden unexpected situation.

[0108] Step S6: Conduct autonomous driving simulation tests on the target vehicle in the target scenario.

[0109] Among them, such as Figure 3 As shown, step S6 specifically includes:

[0110] Step S6.1: Run the target scenario in the simulation environment;

[0111] Step S6.2: Place the target vehicle in the target scene and set the start and end points;

[0112] Step S6.3: Instruct the target vehicle to drive from the starting point to the destination in an autonomous driving mode;

[0113] Step S6.4: Test the autonomous driving performance of the target vehicle based on its performance during driving.

[0114] Another embodiment of the present invention provides an automatic random generation system for autonomous driving simulation test scenarios. The system includes a data acquisition module, an element parsing module, a configuration information module, a map generation module, a scene generation module, and a simulation test module.

[0115] The data acquisition module uses data acquisition equipment to collect real map data and real scene data in a real road environment.

[0116] The element parsing module parses the real map data to obtain multiple basic map elements, calculates the occurrence probability and occurrence location of each basic map element, and obtains a set of map elements with probability information and location information; it also parses the real scene data to obtain multiple basic scene elements, calculates the occurrence probability and occurrence location of each basic scene element, and obtains a set of scene elements with probability information and location information.

[0117] The basic map elements include lanes, intersections, traffic markings, traffic signs, traffic lights, protective barriers, and bus stops; the lanes are minimum driving lanes, i.e., single lanes, including straight lanes, curved lanes, and transition lanes; the intersections include three-way intersections, four-way intersections, and multi-way intersections; each road includes at least one lane, and the transition lanes are used to connect two roads with different numbers of lanes;

[0118] The probability of a certain basic map element appearing is obtained by dividing the number of a certain basic map element by the total number of all basic map elements. The locations where basic map elements appear include intersections, roadsides, roadside, laneside, and lane center. The map element set contains all basic map elements as well as the probability and location information of each basic map element.

[0119] The basic scene elements include traffic participants and non-traffic participants; traffic participants include motor vehicles, non-motor vehicles and pedestrians, and non-traffic participants include ancillary facilities and obstacles; obstacles include man-made roadblocks and accidental obstacles, and accidental obstacles include large tree branches, large bricks and rubble, and large pieces of garbage;

[0120] The probability of occurrence of a certain basic scene element is obtained by dividing the number of a certain basic scene element by the total number of all basic scene elements. The occurrence locations of basic scene elements include roadside, road center, lane center, and traffic markings. The scene element set contains all basic scene elements as well as probability information and location information for each basic scene element. The scene element set also contains the starting point, ending point, trajectory, and speed information of each traffic participant.

[0121] The configuration information module obtains map configuration information and scene configuration information.

[0122] The map configuration information includes the map size and the number of each basic map element;

[0123] The method for determining the number of each basic map element is to first determine the number of intersections, and then, based on the probability information of the map element set, use the probability ratio of each basic map element to determine the number of other basic map elements in turn, or to select them completely randomly.

[0124] The scenario configuration information includes the number of each basic scenario element and the degree of rule compliance for each traffic participant;

[0125] The quantity of each basic map element includes the quantity of each type of traffic participant and the quantity of each type of non-traffic participant. The method for determining the quantity of each type of traffic participant is to first determine the quantity of motor vehicles, and then, based on the probability information of the scene element set, determine the quantity of non-motor vehicles and pedestrians sequentially using the probability ratio of motor vehicles, non-motor vehicles, and pedestrians, or select them completely randomly. The method for determining the quantity of each type of non-traffic participant is to first determine the quantity of ancillary facilities, and then, based on the probability information of the scene element set, determine the quantity of obstacles using the probability ratio of ancillary facilities and obstacles, or select them completely randomly.

[0126] The degree of compliance with the rules for each type of traffic participant ranges from 0 to 1, representing the ratio of the number of participants who comply with the rules to the total number of participants, and indicating the probability that each type of traffic participant complies with the rules.

[0127] The map generation module randomly selects basic map elements that meet the quantity requirements from the map element set according to the map configuration information, and combines the selected basic map elements according to the first preset rule to automatically generate the target map.

[0128] The first preset rule includes:

[0129] Step T1: Create a blank map based on the map size information;

[0130] Step T2: Randomly place all intersections in the blank map in sequence; specifically, multiple intersections are arranged horizontally, vertically, or irregularly; adjacent intersections are guaranteed to have at least one opening facing each other, and if they cannot face each other, they are guaranteed to be smoothly connected in terms of distance and angle; the distance between two adjacent intersections follows a probability distribution or is determined randomly.

[0131] Step T3: Connect adjacent intersections using lanes to form a road network; specifically, if two adjacent intersections have openings facing each other and the same number of lanes, connect them with straight lanes; if two adjacent intersections have openings facing each other but different numbers of lanes, connect them with straight lanes and transition lanes; if two adjacent intersections have openings not facing each other but the same number of lanes, connect them with curved lanes; if two adjacent intersections have openings not facing each other but different numbers of lanes, connect them with curved lanes and transition lanes.

[0132] Step T4: Place traffic markings, traffic signs, traffic lights, protective barriers, and bus stops to form a target map; specifically, place traffic markings at intersections and on lanes; place traffic signs at intersections and on the roadside, with the traffic signs suspended above the lanes; place traffic lights at intersections; place protective barriers in the middle of the road and on the roadside; and place bus stops on the roadside.

[0133] The placement of all basic map elements must conform to the corresponding location information, and the placement of later basic map elements must not conflict with the placement of earlier basic map elements.

[0134] The scene generation module randomly selects basic scene elements that meet the quantity requirements from the scene element set according to the scene configuration information, and places the selected basic scene elements in the target map according to the second preset rule to automatically generate the target scene.

[0135] The second preset rule includes:

[0136] All motor vehicles, non-motor vehicles, pedestrians, ancillary facilities and obstacles are randomly placed in the target map in sequence, and the placement order of different basic scene elements can be arbitrarily selected;

[0137] The placement of all basic scene elements must conform to the corresponding position information, and the placement of later basic scene elements must not conflict with the placement of earlier basic scene elements.

[0138] Based on the start point, end point, trajectory, and speed information of each traffic participant contained in the scene element set, determine the start point, end point, trajectory, and speed of each motor vehicle, each non-motor vehicle, and each pedestrian in the target map;

[0139] Based on the degree of compliance with the rules by each type of traffic participant, a corresponding number of motor vehicles, non-motor vehicles, and pedestrians are selected and instructed to perform behaviors that violate traffic rules at a preset time, as a sudden unexpected situation.

[0140] The simulation testing module performs autonomous driving simulation tests on the target vehicle in the target scenario. Specifically, the target scenario is first run in the simulation environment, then the target vehicle is placed in the target scenario, a start point and an end point are set, and the target vehicle is instructed to drive from the start point to the end point in an autonomous driving mode. Finally, the autonomous driving performance of the target vehicle is tested based on its performance during the driving process.

[0141] The autonomous driving simulation test scenario random automatic generation method and system provided in this invention solves the problem of how to quickly and automatically generate a large number of simulation test scenarios.

[0142] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0143] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, 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. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for automatically generating randomized simulation test scenarios for autonomous driving, characterized in that, include: Step S1: Use data acquisition equipment to collect real map data and real scene data in a real road environment; Step S2: Analyze the real map data to obtain multiple basic map elements, and count the occurrence probability and occurrence location of each basic map element to obtain a set of map elements with probability information and location information; The real scene data is parsed to obtain multiple basic scene elements. The occurrence probability and occurrence position of each basic scene element are statistically analyzed to obtain a set of scene elements with probability information and position information. Step S3: Obtain map configuration information and scene configuration information; Step S4: Based on the map configuration information, randomly select basic map elements that meet the quantity requirements from the map element set, and combine the selected basic map elements according to the first preset rule to automatically generate the target map; Step S5: Based on the scene configuration information, randomly select basic scene elements that meet the quantity requirements from the scene element set, and place the selected basic scene elements in the target map according to the second preset rule to automatically generate the target scene; Step S6: Conduct autonomous driving simulation tests on the target vehicle in the target scenario; The basic map elements include lanes, intersections, traffic markings, traffic signs, traffic lights, protective barriers, and bus stops; the lanes are minimum driving lanes, i.e., single lanes, including straight lanes, curved lanes, and transition lanes; the intersections include three-way intersections, four-way intersections, and multi-way intersections; each road includes at least one lane, and the transition lanes are used to connect two roads with different numbers of lanes; The probability of a certain basic map element appearing is obtained by dividing the number of a certain basic map element by the total number of all basic map elements. The locations where basic map elements appear include intersections, roadsides, roadside, laneside, and lane center. The map element set contains all basic map elements as well as the probability and location information of each basic map element. The basic scene elements include traffic participants and non-traffic participants; traffic participants include motor vehicles, non-motor vehicles and pedestrians, and non-traffic participants include ancillary facilities and obstacles; obstacles include man-made roadblocks and accidental obstacles, and accidental obstacles include large tree branches, large bricks and rubble, and large pieces of garbage; The probability of a certain basic scene element is obtained by dividing the number of a certain basic scene element by the total number of all basic scene elements. The occurrence locations of basic scene elements include roadside, road center, lane center, and traffic markings. The scene element set contains all basic scene elements as well as probability and location information for each basic scene element. The scene element set also contains the starting point, ending point, trajectory, and speed information of each traffic participant. The map configuration information includes the map size and the number of each basic map element; The method for determining the number of each basic map element is to first determine the number of intersections, and then, based on the probability information of the map element set, use the probability ratio of each basic map element to determine the number of other basic map elements in turn. The scenario configuration information includes the number of each basic scenario element and the degree of rule compliance for each traffic participant; The quantity of each basic map element includes the quantity of each traffic participant and the quantity of each non-traffic participant; the method for determining the quantity of each traffic participant is to first determine the quantity of motor vehicles, and then, based on the probability information of the scene element set, use the probability ratio of motor vehicles, non-motor vehicles and pedestrians to determine the quantity of non-motor vehicles and pedestrians in turn. The method for determining the number of each type of non-traffic participant is to first determine the number of ancillary facilities, and then, based on the probability information of the scene element set, determine the number of obstacles by using the probability ratio of ancillary facilities to obstacles; The degree of compliance with the rules for each type of traffic participant ranges from 0 to 1, representing the ratio of the number of participants who comply with the rules to the total number of participants, and indicating the probability that each type of traffic participant complies with the rules.

2. The method for automatically generating randomized autonomous driving simulation test scenarios according to claim 1, characterized in that, The first preset rule includes: Step T1: Create a blank map based on the map size information; Step T2: Randomly place all intersections sequentially on the blank map; Step T3: Connect adjacent intersections using lanes to form a road network; Step T4: Place traffic markings, traffic signs, traffic lights, isolation and protection facilities, and bus stops to form the target map.

3. The method for automatically generating randomized autonomous driving simulation test scenarios according to claim 2, characterized in that, Step T2 includes: multiple intersections are arranged horizontally, vertically, or irregularly; adjacent intersections are guaranteed to have at least one opening facing each other; if they cannot face each other, they are guaranteed to be smoothly connected in terms of distance and angle; the distance between two adjacent intersections follows a probability distribution or is determined randomly. Step T3 includes: if two adjacent intersections have openings facing each other and the same number of lanes, then connect them with a straight lane; if two adjacent intersections have openings facing each other but different numbers of lanes, then connect them with a straight lane and a transition lane; if two adjacent intersections have openings not facing each other but the same number of lanes, then connect them with a curved lane; if two adjacent intersections have openings not facing each other but different numbers of lanes, then connect them with a curved lane and a transition lane. Step T4 includes: placing traffic markings at intersections and on lanes; placing traffic signs at intersections and on the roadside, the traffic signs being suspended above the lanes; placing traffic lights at intersections; placing isolation and protective facilities in the middle of the road and on the roadside; and placing bus stops on the roadside. The placement of all basic map elements must conform to the corresponding location information, and the placement of later basic map elements must not conflict with the placement of earlier basic map elements.

4. The method for automatically generating randomized autonomous driving simulation test scenarios according to claim 1, characterized in that, The second preset rule includes: All motor vehicles, non-motor vehicles, pedestrians, ancillary facilities and obstacles are randomly placed in the target map in sequence, and the placement order of different basic scene elements can be arbitrarily selected; The placement of all basic scene elements must conform to the corresponding position information, and the placement of later basic scene elements must not conflict with the placement of earlier basic scene elements. Based on the start point, end point, trajectory, and speed information of each traffic participant contained in the scene element set, determine the start point, end point, trajectory, and speed of each motor vehicle, each non-motor vehicle, and each pedestrian in the target map; Based on the degree of compliance with the rules by each type of traffic participant, a corresponding number of motor vehicles, non-motor vehicles, and pedestrians are selected and instructed to perform behaviors that violate traffic rules at a preset time, as a sudden unexpected situation.

5. The method for automatically generating randomized autonomous driving simulation test scenarios according to claim 1, characterized in that, Step S6 includes: Step S6.1: Run the target scenario in the simulation environment; Step S6.2: Place the target vehicle in the target scene and set the start and end points; Step S6.3: Instruct the target vehicle to drive from the starting point to the destination in an autonomous driving mode; Step S6.4: Test the autonomous driving performance of the target vehicle based on its performance during driving.

6. A system for automatically generating randomized simulation test scenarios for autonomous driving, characterized in that, include: Data acquisition module: Uses data acquisition equipment to collect real map data and real scene data in a real road environment; Element parsing module: Parses the real map data to obtain multiple basic map elements, calculates the occurrence probability and occurrence location of each basic map element, and obtains a set of map elements with probability information and location information; The real scene data is parsed to obtain multiple basic scene elements. The occurrence probability and occurrence position of each basic scene element are statistically analyzed to obtain a set of scene elements with probability information and position information. Configuration information module: Retrieves map configuration information and scene configuration information; Map generation module: Based on the map configuration information, randomly selects basic map elements that meet the quantity requirements from the map element set, and combines the selected basic map elements according to the first preset rule to automatically generate the target map; Scene generation module: Based on the scene configuration information, randomly selects basic scene elements that meet the quantity requirements from the scene element set, and places the selected basic scene elements in the target map according to the second preset rule to automatically generate the target scene; Simulation testing module: Performs autonomous driving simulation tests on the target vehicle in the target scenario; The basic map elements include lanes, intersections, traffic markings, traffic signs, traffic lights, protective barriers, and bus stops; the lanes are minimum driving lanes, i.e., single lanes, including straight lanes, curved lanes, and transition lanes; the intersections include three-way intersections, four-way intersections, and multi-way intersections; each road includes at least one lane, and the transition lanes are used to connect two roads with different numbers of lanes; The probability of a certain basic map element appearing is obtained by dividing the number of a certain basic map element by the total number of all basic map elements. The locations where basic map elements appear include intersections, roadsides, roadside, laneside, and lane center. The map element set contains all basic map elements as well as the probability and location information of each basic map element. The basic scene elements include traffic participants and non-traffic participants; traffic participants include motor vehicles, non-motor vehicles and pedestrians, and non-traffic participants include ancillary facilities and obstacles; obstacles include man-made roadblocks and accidental obstacles, and accidental obstacles include large tree branches, large bricks and rubble, and large pieces of garbage; The probability of a certain basic scene element is obtained by dividing the number of a certain basic scene element by the total number of all basic scene elements. The occurrence locations of basic scene elements include roadside, road center, lane center, and traffic markings. The scene element set contains all basic scene elements as well as probability and location information for each basic scene element. The scene element set also contains the starting point, ending point, trajectory, and speed information of each traffic participant. The map configuration information includes the map size and the number of each basic map element; The method for determining the number of each basic map element is to first determine the number of intersections, and then, based on the probability information of the map element set, use the probability ratio of each basic map element to determine the number of other basic map elements in turn. The scenario configuration information includes the number of each basic scenario element and the degree of rule compliance for each traffic participant; The quantity of each basic map element includes the quantity of each traffic participant and the quantity of each non-traffic participant; the method for determining the quantity of each traffic participant is to first determine the quantity of motor vehicles, and then, based on the probability information of the scene element set, use the probability ratio of motor vehicles, non-motor vehicles and pedestrians to determine the quantity of non-motor vehicles and pedestrians in turn. The method for determining the number of each type of non-traffic participant is to first determine the number of ancillary facilities, and then, based on the probability information of the scene element set, determine the number of obstacles by using the probability ratio of ancillary facilities to obstacles; The degree of compliance with the rules for each type of traffic participant ranges from 0 to 1, representing the ratio of the number of participants who comply with the rules to the total number of participants, and indicating the probability that each type of traffic participant complies with the rules.

7. The automatic random generation system for autonomous driving simulation test scenarios according to claim 6, characterized in that, The first preset rule includes: Step T1: Create a blank map based on the map size information; Step T2: Randomly place all intersections in the blank map in sequence; specifically, multiple intersections are arranged horizontally, vertically, or irregularly; adjacent intersections are guaranteed to have at least one opening facing each other, and if they cannot face each other, they are guaranteed to be smoothly connected in terms of distance and angle; the distance between two adjacent intersections follows a probability distribution or is determined randomly. Step T3: Connect adjacent intersections using lanes to form a road network; specifically, if two adjacent intersections have openings facing each other and the same number of lanes, connect them with straight lanes; if two adjacent intersections have openings facing each other but different numbers of lanes, connect them with straight lanes and transition lanes; if two adjacent intersections have openings not facing each other but the same number of lanes, connect them with curved lanes; if two adjacent intersections have openings not facing each other but different numbers of lanes, connect them with curved lanes and transition lanes. Step T4: Place traffic markings, traffic signs, traffic lights, protective barriers, and bus stops to form a target map; specifically, place traffic markings at intersections and on lanes; place traffic signs at intersections and on the roadside, with the traffic signs suspended above the lanes; place traffic lights at intersections; place protective barriers in the middle of the road and on the roadside; and place bus stops on the roadside. The placement of all basic map elements must conform to the corresponding location information, and the placement of later basic map elements must not conflict with the placement of earlier basic map elements.

8. The automatic random generation system for autonomous driving simulation test scenarios according to claim 6, characterized in that, The second preset rule includes: All motor vehicles, non-motor vehicles, pedestrians, ancillary facilities and obstacles are randomly placed in the target map in sequence, and the placement order of different basic scene elements can be arbitrarily selected; The placement of all basic scene elements must conform to the corresponding position information, and the placement of later basic scene elements must not conflict with the placement of earlier basic scene elements. Based on the start point, end point, trajectory, and speed information of each traffic participant contained in the scene element set, determine the start point, end point, trajectory, and speed of each motor vehicle, each non-motor vehicle, and each pedestrian in the target map; Based on the degree of compliance with the rules by each type of traffic participant, a corresponding number of motor vehicles, non-motor vehicles, and pedestrians are selected and instructed to perform behaviors that violate traffic rules at a preset time, as a sudden unexpected situation.

Images