CN114168457A - Method and system for generating test scene of automatic driving automobile - Google Patents
Method and system for generating test scene of automatic driving automobile Download PDFInfo
- Publication number
- CN114168457A CN114168457A CN202111416121.0A CN202111416121A CN114168457A CN 114168457 A CN114168457 A CN 114168457A CN 202111416121 A CN202111416121 A CN 202111416121A CN 114168457 A CN114168457 A CN 114168457A
- Authority
- CN
- China
- Prior art keywords
- scene
- test
- vehicle
- traffic
- parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012360 testing method Methods 0.000 title claims abstract description 190
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012216 screening Methods 0.000 claims abstract description 39
- 230000006870 function Effects 0.000 claims abstract description 31
- 238000013507 mapping Methods 0.000 claims abstract description 31
- 238000013142 basic testing Methods 0.000 claims abstract description 23
- 238000004088 simulation Methods 0.000 claims abstract description 23
- 230000007613 environmental effect Effects 0.000 claims abstract description 9
- 238000005070 sampling Methods 0.000 claims abstract description 5
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 230000009471 action Effects 0.000 claims description 10
- 230000003068 static effect Effects 0.000 claims description 7
- 230000004888 barrier function Effects 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 3
- 230000008447 perception Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 230000002452 interceptive effect Effects 0.000 claims description 2
- 230000008859 change Effects 0.000 description 4
- 239000002184 metal Substances 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000002829 reductive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3684—Test management for test design, e.g. generating new test cases
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
A method for generating a test scene of an automatic driving automobile comprises the following steps: s1, determining a test scene related to the test requirement according to the test requirement of the automatic driving simulation, designing a motion mode of the test main vehicle, analyzing possible motion states of other traffic participants in the test scene, and determining a function scene based on a hierarchical model; s2, combining the possible movement directions of the main vehicle with the possible movement directions of the environmental interference vehicle, adding movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups; s3, screening out a basic test scene group which has test value and can cover the whole test scene and functions according to the screening rule of the scene, mapping the basic test scene group to a parameter space to generate a logic scene, and S4, sampling the parameter space of the scene through scene keywords to generate a specific scene file in an OpenSCENARIO format; and completing the automatic driving test case, and verifying the validity of the test case through simulation.
Description
Technical Field
The invention relates to the technical field of automatic driving simulation, in particular to a method and a system for generating a test scene of an automatic driving automobile.
Background
The traditional intelligent driving road testing method is expensive and time-consuming, and has a high risk coefficient. And the test can be only carried out on a specific road section, and in addition, special scenes such as extreme weather, sensor failure and partial road section damage can not be tested and repeated. The existing automatic driving test technology can solve the defects of the traditional intelligent road test. For the test work of the automatic driving simulation technology built based on the scene in the existing automatic driving test technology, the safety and the repeatability of the test can be improved. The scene generation mode in the existing simulation test of automatic driving has low efficiency.
Disclosure of Invention
In view of this, the invention provides a method and a system for generating a test scene of an automatic driving automobile.
A method for generating a test scene of an automatic driving automobile comprises the following steps:
s1, determining a test scene related to the test requirement according to the test requirement of the automatic driving simulation, designing a motion mode of the test main vehicle, analyzing possible motion states of other traffic participants in the test scene, and determining a function scene based on a hierarchical model;
s2, combining the possible movement directions of the main vehicle with the possible movement directions of the environmental interference vehicle, adding movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups;
s3, screening basic test scene groups which have test values and can cover the whole test scene and functions according to the screening rules of the scenes, and mapping the basic test scene groups to a parameter space to generate a logic scene;
s4, sampling the parameter space of the scene through the scene keywords to generate a specific scene file in OpenSCENARIO format; and completing an automatic driving test case based on an artificial potential field method and a model predictive control algorithm, and verifying the validity of the test case through simulation.
In the method for generating the test scenario of the autonomous vehicle according to the present invention,
the road types in the test scene comprise straight roads, curved roads, ramp road sections, intersections, rotary islands and high-speed access ramps of urban roads, high-speed roads and the like;
the test scene comprises the daytime, sunny days and rainy days;
the traffic facilities in the test scene are road traffic signs and road marking lines;
the traffic participants in the test scenario are vehicles and other traffic participants.
In the method for generating the test scenario of the autonomous vehicle according to the present invention,
designing a motion mode of the test main vehicle, and then analyzing possible motion states of other traffic participants in a test scene, wherein the motion modes comprise:
analyzing possible combinations of relative positions and movement directions of the host vehicle and the traffic participants around the host vehicle aiming at the specified road traffic environment, and determining the relative position range of the host vehicle and the traffic participants around the host vehicle; and determining the possible movement direction of the main vehicle and the movement direction of each interference vehicle influencing the movement direction of the main vehicle according to the test target of the main vehicle control and perception function aiming at the complex scene group.
In the method for generating the test scenario of the autonomous vehicle according to the present invention,
determining a functional scene based on a hierarchical model according to the movement mode of the main vehicle and the movement states of the rest traffic participants, wherein the functional scene comprises the following steps:
sequentially arranging a basic road, a traffic infrastructure, a temporary static barrier, a traffic participant, an environment and an electronic communication facility according to the 1 st layer to the 6 th layer of the determined hierarchical model;
aiming at the test requirement, the necessary inputs of the test requirement are the road topology structure of the layer 1, the traffic environment signal provided by the layer 2 and the information of the traffic participants in the layer 4, wherein the information of the traffic participants in the layer 4 comprises the position and action information of the static and dynamic traffic participants including the test main vehicle.
In the method for generating the test scenario of the autonomous vehicle according to the present invention,
combining the possible movement direction permutation and combination of the possible movement directions of the main vehicle with the environment interference vehicle, adding necessary movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups, wherein the method comprises the following steps:
combining all possible motion directions determined by the main vehicle function with all possible motion directions of interfering vehicles according to the principle, adopting a PICT combination test tool during combination, adding necessary motion constraint conditions, selecting a parameter combination coverage standard to automatically generate all combined scene groups, designing a scene screening rule, screening out scenes with test values from the combined scene groups according to the scene screening rule, and forming a normal driving working condition scene group and a pre-collision working condition scene group, namely forming a basic scene library.
In the method for generating the test scene of the automatic driving automobile, a basic test scene group which has test value and can cover the whole test scene and function is screened out according to the screening rule of the scene, and the method comprises the following steps:
after the test scenes are generated by combining the possible motion directions of the main vehicle and all the traffic participants, similar scenes and scenes which do not exist in reality exist, the test scenes generated by the PICT are screened so as to obtain valuable test scenes, and the screening principle is set as follows: 1. scenes that cannot be realized in real traffic; 2. similar scenes exist, and only scenes with large influence on the main vehicle are reserved; 3. only one scene is reserved for scenes with the same influence on the movement of the main vehicle, and a basic test scene group which has test value and can cover the whole test scene and functions is screened out according to the screening principle of the test scene.
In the method for generating the test scene of the automatic driving automobile, a basic test scene group which has test value and can cover the whole test scene and function is screened out and mapped to a parameter space to generate a logic scene, and the method comprises the following steps:
extracting road topological structure, traffic signals, self-vehicle and other traffic participant function scene keywords to map into corresponding parameters according to the screened scene library so as to construct a parameter space and realize the generation of a logic scene; the parameter type obtained by mapping the functional scene keyword is called a key parameter type, other default parameter types required by the subsequent generation of a specific scene are called non-key parameter types, and the mapping of the functional scene to a parameter space is realized to generate a logic scene.
In the method for generating the test scene of the automatic driving automobile, the parameter space of the scene is sampled by the scene keyword to generate the specific scene file in the OpenSCENARIO format, which comprises the following steps:
and automatically generating a specific scene file in an Open-SCENARIO format based on the functional scene keywords by analyzing the mapping relation between the logic scene parameter space and the OpenSCENARIO format.
In the method for generating the test scene of the automatic driving automobile, the step of analyzing the mapping relation between the logic scene parameter space and the OpenSCENARIO format comprises the following steps:
analyzing an OpenSCENARIO format, and dividing the OpenSCENARIO format into a directory, an object and a storyboard 3-layer structure, wherein the directory corresponds to a road network file and a vehicle directory in a logic scene, the object corresponds to a vehicle, a pedestrian and a controller in a logic space, and the storyboard corresponds to the initialization of the logic scene, a storyboard sequence and an ending condition;
distinguishing key parameters and non-key parameters, and using the mapped non-key parameters as default values to simplify parameter space; key parameters corresponding to the road topological structure and the traffic signal are assigned in a road network file, namely Opendrive;
and determining the parameter range of each parameter type according to the test requirement to obtain a complete parameter space.
In the method for generating the test scene of the automatic driving automobile, the specific scene file in the Open-SCENARIO format is automatically generated based on the functional scene keywords, and the method comprises the following steps:
and the mapping relation adopts a modularized XML document object model to compile an automatic script, the automatic script samples the constructed parameter space based on the functional scene keywords, and generates corresponding OpenSCE-NARIO content by adopting XML DOM according to the mapping relation between the parameter space and the OpenSCENARIO format, so as to obtain a complete specific scene file in the OpenSCENARIO format.
In the method for generating the test scenario of the autonomous vehicle according to the present invention,
completing an automatic driving test case based on an artificial potential field method and a model predictive control algorithm, and verifying the validity of the test case through simulation comprises the following steps:
and designing an automatic driving test case based on an artificial potential field and a model predictive control algorithm on the completed test scene, completing simulation test verification, and verifying the validity of the test case.
The invention also provides a system for generating the test scene of the automatic driving automobile, which is realized by the method for generating the test scene of the automatic driving automobile.
The beneficial technical effects are as follows: compared with the prior art, the method and the system for generating the test scene of the automatic driving automobile can quickly build a basic test scene group which has test value and meets various functions by utilizing a combined reasoning method and a scene screening rule. And then determining a functional scene, a logic scene and a specific scene based on the hierarchical model, and completing the construction of the scene. The problems that an existing scene building mode is low in efficiency, incomplete in coverage and the like are solved.
Drawings
Fig. 1 is a flowchart of a method for generating a test scenario of an autonomous vehicle according to an embodiment of the present invention.
Fig. 2 is a schematic diagram illustrating a type of a straight road in a functional scene during a scene building process according to an embodiment of the present invention.
Detailed Description
As shown in fig. 1, in an embodiment of the present invention, an embodiment of the present application discloses a method for generating an auto-driving vehicle test scenario, including the following steps:
s1, determining a test scene related to the test requirement according to the test requirement of the automatic driving simulation, designing a motion mode of the test main vehicle, analyzing possible motion states of other traffic participants in the test scene, and determining a function scene based on a hierarchical model;
s2, combining the possible movement directions of the main vehicle with the possible movement directions of the environmental interference vehicle, adding movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups;
s3, screening basic test scene groups which have test values and can cover the whole test scene and functions according to the screening rules of the scenes, and mapping the basic test scene groups to a parameter space to generate a logic scene;
s4, sampling the parameter space of the scene through the scene keywords to generate a specific scene file in OpenSCENARIO format; and completing an automatic driving test case based on an artificial potential field method and a model predictive control algorithm, and verifying the validity of the test case through simulation.
Preferably, the first and second electrodes are formed of a metal,
the road types in the test scene comprise straight roads, curved roads, ramp road sections, intersections, rotary islands and high-speed access ramps of urban roads, high-speed roads and the like;
the test scene comprises the daytime, sunny days and rainy days;
the traffic facilities in the test scene are road traffic signs and road marking lines;
the traffic participants in the test scenario are vehicles and other traffic participants.
Preferably, the first and second electrodes are formed of a metal,
designing a motion mode of the test main vehicle, and then analyzing possible motion states of other traffic participants in a test scene, wherein the motion modes comprise:
analyzing possible combinations of relative positions and movement directions of the host vehicle and the traffic participants around the host vehicle aiming at the specified road traffic environment, and determining the relative position range of the host vehicle and the traffic participants around the host vehicle; and determining the possible movement direction of the main vehicle and the movement direction of each interference vehicle influencing the movement direction of the main vehicle according to the test target of the main vehicle control and perception function aiming at the complex scene group.
Preferably, the first and second electrodes are formed of a metal,
determining a functional scene based on a hierarchical model according to the movement mode of the main vehicle and the movement states of the rest traffic participants, wherein the functional scene comprises the following steps:
sequentially arranging a basic road, a traffic infrastructure, a temporary static barrier, a traffic participant, an environment and an electronic communication facility according to the 1 st layer to the 6 th layer of the determined hierarchical model;
aiming at the test requirement, the necessary inputs of the test requirement are the road topology structure of the layer 1, the traffic environment signal provided by the layer 2 and the information of the traffic participants in the layer 4, wherein the information of the traffic participants in the layer 4 comprises the position and action information of the static and dynamic traffic participants including the test main vehicle.
Preferably, the first and second electrodes are formed of a metal,
combining the possible movement direction permutation and combination of the possible movement directions of the main vehicle with the environment interference vehicle, adding necessary movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups, wherein the method comprises the following steps:
combining all possible motion directions determined by the main vehicle function with all possible motion directions of all interference vehicles (including the situation that any interference vehicle does not exist) according to the above, adopting a PICT (central insulated gate bipolar transistor) combined test tool during combination, adding necessary motion constraint conditions, selecting a parameter combination coverage standard to automatically generate all combined scene groups, designing a scene screening rule, screening out scenes with test values from the combined scene groups according to the scene screening rule, and forming a normal driving working condition scene group and a pre-collision working condition scene group, namely forming a basic scene library.
Preferably, screening out a basic test scenario group which has test value and can cover the whole test scenario and function according to the screening rule of the scenario, and comprises the following steps:
after the test scenes are generated by combining the possible motion directions of the main car and all the traffic participants, similar scenes and scenes which do not exist in reality exist, the test scenes generated by the paired test case generators PICT are screened so as to obtain valuable test scenes, and the screening principle is set as follows: 1. scenes that cannot be realized in real traffic; 2. similar scenes exist, and only scenes with large influence on the main vehicle are reserved; 3. only one scene is reserved for scenes with the same influence on the movement of the main vehicle, and a basic test scene group which has test value and can cover the whole test scene and functions is screened out according to the screening principle of the test scene. The screening rules 1 to 3 are "or" relationships, and any one of the screening rules 1 to 3 may be satisfied.
Preferably, screening out a basic test scenario group which has test value and can cover the whole test scenario and function, and mapping the basic test scenario group to a parameter space to generate a logic scenario, wherein the basic test scenario group comprises:
extracting road topological structure, traffic signals, self-vehicle and other traffic participant function scene keywords to map into corresponding parameters according to the screened scene library so as to construct a parameter space and realize the generation of a logic scene; the parameter type obtained by mapping the functional scene keyword is called a key parameter type, other default parameter types required by the subsequent generation of a specific scene are called non-key parameter types, and the mapping of the functional scene to a parameter space is realized to generate a logic scene.
Preferably, the parameter space of the scene is sampled by the scene keyword to generate a specific scene file in an OpenSCENARIO format, including:
and automatically generating a specific scene file in an Open-SCENARIO format based on the functional scene keywords by analyzing the mapping relation between the logic scene parameter space and the OpenSCENARIO format.
In the method for generating the test scene of the automatic driving automobile, the step of analyzing the mapping relation between the logic scene parameter space and the OpenSCENARIO format comprises the following steps:
analyzing an OpenSCENARIO format, and dividing the OpenSCENARIO format into a directory, an object and a storyboard 3-layer structure, wherein the directory corresponds to a road network file and a vehicle directory in a logic scene, the object corresponds to a vehicle, a pedestrian and a controller in a logic space, and the storyboard corresponds to the initialization of the logic scene, a storyboard sequence and an ending condition;
distinguishing key parameters and non-key parameters, and using the mapped non-key parameters as default values to simplify parameter space; key parameters corresponding to the road topological structure and the traffic signal are assigned in a road network file, namely Opendrive;
and determining the parameter range of each parameter type according to the test requirement to obtain a complete parameter space.
Preferably, the specific scene file in the Open-SCENARIO format is automatically generated based on the functional scene keyword, and comprises:
and the mapping relation adopts a modularized XML document object model to compile an automatic script, the automatic script samples the constructed parameter space based on the functional scene keywords, and generates corresponding OpenSCE-NARIO content by adopting XML DOM according to the mapping relation between the parameter space and the OpenSCENARIO format, so as to obtain a complete specific scene file in the OpenSCENARIO format.
In the method for generating the test scenario of the autonomous vehicle according to the present invention,
completing an automatic driving test case based on an artificial potential field method and a model predictive control algorithm, and verifying the validity of the test case through simulation comprises the following steps:
and designing an automatic driving test case based on an artificial potential field and a model predictive control algorithm on the completed test scene, completing simulation test verification, and verifying the validity of the test case.
The embodiment of the invention also provides a system for generating the test scene of the automatic driving automobile, which is realized by the method for generating the test scene of the automatic driving automobile.
The beneficial technical effects are as follows: compared with the prior art, the method and the system for generating the test scene of the automatic driving automobile can quickly build a basic test scene group which has test value and meets various functions by using a combined reasoning method and a scene screening rule. And then determining a functional scene, a logic scene and a specific scene based on the hierarchical model, and completing the construction of the scene. The problems that an existing scene building mode is low in efficiency, incomplete in coverage and the like are solved.
In a preferred embodiment, a test scenario for the autopilot simulation is determined based on the autopilot simulation test requirements. Determining the road type of a scene, then determining the movement mode of a main vehicle according to the determined road scene type, analyzing the possible movement states of other traffic participants in the scene, and determining a functional scene based on a hierarchical model, specifically comprising:
determining the scene requirement of the simulation test according to the test requirement of the simulation of the automatic driving, wherein the road type of the determined scene can be straight roads, curved roads, ramp road sections, intersections, roundabouts, expressway entrance and exit ramps and the like of urban roads, expressway and the like; the time of day is day, sunny and rainy and the like; the traffic facilities are road traffic signs and road marking lines; traffic participants are vehicles and other traffic participants.
Then, the movement mode of the main vehicle is designed according to the type of the road, and the movement modes of the main vehicle and the environment interference vehicle are 9 possible movement states of straight running, left-right lane changing, left-right deviation running, turning around and reversing, reversing and parking. According to the possible motion state of the vehicle, a function model is built based on a hierarchical model, the function model is divided into 1-4 layers, namely a topological structure of a road, a traffic environment signal, various sensor data and information (static position information and action information) of traffic participants including the vehicle.
In a preferred embodiment, the possible movement directions of the main vehicle are arranged and combined in combination with the possible movement directions of the environmental interference vehicle according to the functional scenes, necessary movement constraint conditions are added, parameter combination coverage standards are selected to automatically generate all combined scene groups, scene screening rules are designed, and scenes with test values are screened from the combined scene groups according to the scene screening rules, so that the scenes form a normal driving condition scene group and a pre-collision condition scene group, namely a basic scene library is formed, and the method specifically comprises the following steps:
according to the functional scenario, as shown in fig. 2, the type of road is determined to be a straight lane of three lanes of an urban road, the host vehicle and the preceding environmental vehicles a1, a2, and A3. There are 9 possible directions of movement of the vehicle (straight, left-right lane change, left-right turning, left-right yaw, reverse, and stop), and for the movement of the main vehicle, the a1 vehicle may move in 7 possible directions of straight, left-right lane change, left-right yaw, and stationary. Similarly, a2 and A3 are adjacent lanes, which have no influence on the movement of the main vehicle when lane changing and deflecting of the outer lane are completed, so that the possible movement directions of a2 and A3 are 7 cases of straight running, lane changing left and right, deflecting left and right, and standing still. Considering the situation that the environmental interference vehicle does not exist, 4096 combined basic scenes can be achieved. The invention only exemplifies straight roads, and also road types such as straight roads, curved roads, ramp road sections, intersections, roundabouts, high-speed entrance and exit ramps and the like of urban roads, high-speed roads and the like.
In a preferred embodiment, a scene screening rule is designed, and scenes with test value are screened from the combined scene group according to the rule, so that the scenes form a normal driving condition scene group and a pre-collision condition scene group, namely, a basic scene library is formed, and the method specifically comprises the following steps:
designing a rule for scene screening, wherein the rule for scene screening is as follows: a scene which cannot be realized in real traffic; similar scenes exist, and only scenes with large influence on the main vehicle are reserved; and thirdly, only one scene is reserved for scenes with the same influence on the movement of the main vehicle. When the main vehicle travels straight, 6 movement directions of the main vehicle and an environmental vehicle A1 are considered, and 3 directions of A2 and A3 form a scene, and meanwhile, according to the actual situation, the situations that A1, A2 and A3 do not exist are considered, and 14 valuable basic scenes are screened out by using a PICT combined test case generation tool. When the main vehicle changes lane to the left, 6 movement directions of the main vehicle and the environmental vehicle A1, and 3 directions of A2 and A3 form a scene, and meanwhile, according to the actual situation, the situation that A1, A2 and A3 do not exist is considered, and a scene construction method is adopted to obtain 16 basic scenes with test value. Similarly, there are 16 valuable test scenarios when the main vehicle makes a left-right yaw. Therefore, the test scene of the straight-going main vehicle is reduced to the analysis of the rest road types and straight roads. Therefore, 78 test scenes with test values are screened out from the basic scene library according to the screening principle of the test scenes, and the basic test scene library which can cover the whole test scene and functions can be obtained.
In a preferred embodiment, mapping to a parameter space according to a screened basic test scenario library to generate a logic scenario specifically includes:
and extracting functional scene keywords of road topological structures, traffic signals, own vehicles, other traffic participants and the like according to the screened scene library and mapping the keywords into corresponding parameters. The topological structure of the road is of a lane type, a straight road, a curve, a ramp and the like so as to construct a parameter space. The actions of the self-vehicle and each traffic participant are divided into atomic actions such as acceleration, deceleration, uniform speed, left lane change, right lane change, left turn, right turn and the like. The relevant action models of lane changing and turning are based on tracks and time, then action triggering conditions are designed, and the sub-parameters of the action triggering conditions are divided into time-based, event-based and space triggering. And realizing the generation of the logic scene. The parameter type obtained by mapping the functional scene keyword is called a key parameter type, other default parameter types required by the subsequent generation of a specific scene are called non-key parameter types, and the mapping of the functional scene to a parameter space is realized to generate a logic scene.
In a preferred embodiment, according to the mapping relationship between the analysis logic scene parameter space and the OpenSCENARIO format, a specific scene file in the Open-SCENARIO format is automatically generated based on the functional scene keyword, and the specific scene file specifically includes:
the opensceenario format is analyzed and divided into directory, object, storyboard 3-layer structures. And then mapping non-critical OpenENARIO contents, such as vehicle directories, controllers and the like, into non-critical parameters in a parameter space corresponding to the logic parameters, and taking the non-critical OpenENARIO contents as default values in the parameter space to simplify the conversion process, thereby finally obtaining the mapping relation between the OpenENARIO contents and the keywords and the types of the parameters. The key parameters, the road topology structure and the key parameters corresponding to the traffic signals are assigned in a road network file, namely Opendrive, the initialization of OpenSCENARIO corresponds to the initial states (initial positions and initial speeds) of the self-vehicle and other traffic participants, and the story sequence corresponds to the action parameters of the self-vehicle and other traffic participants. And determining the parameter range of each parameter type according to the test requirement to obtain a complete parameter space. And generating corresponding OpenSCE-NARIO content by adopting XML DOM according to the mapping relation between the parameter space and the OpenSCENARIO format to obtain a complete specific scene file in the OpenSCENARIO format.
In the preferred embodiment, the automatic driving test case based on the artificial potential field method and the model predictive control algorithm is completed, and the validity of the test case is verified through simulation. The method specifically comprises the following steps:
and establishing an automatic driving vehicle decision, planning and control algorithm model under the established specific test scene, completing an automatic driving test case based on unified modeling of an artificial potential field and a model predictive control algorithm, and verifying the effectiveness of the test scene establishment through simulation.
In the above, the method and system for generating the test scenario of the autonomous driving vehicle provided by the embodiment of the present invention are not limited to the specific implementation manner, and for those skilled in the art, various other corresponding changes and modifications may be made according to the technical concept of the present invention, and all of these changes and modifications shall fall within the protection scope of the claims of the present invention.
Claims (12)
1. A method for generating a test scene of an automatic driving automobile is characterized by comprising the following steps:
s1, determining a test scene related to the test requirement according to the test requirement of the automatic driving simulation, designing a motion mode of the test main vehicle, analyzing possible motion states of other traffic participants in the test scene, and determining a function scene based on a hierarchical model;
s2, combining the possible movement directions of the main vehicle with the possible movement directions of the environmental interference vehicle, adding movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups;
s3, screening basic test scene groups which have test values and can cover the whole test scene and functions according to the screening rules of the scenes, and mapping the basic test scene groups to a parameter space to generate a logic scene;
s4, sampling the parameter space of the scene through the scene keywords to generate a specific scene file in OpenSCENARIO format; and completing an automatic driving test case based on an artificial potential field method and a model predictive control algorithm, and verifying the validity of the test case through simulation.
2. The method of generating an autopilot vehicle test scenario of claim 1,
the road types in the test scene comprise straight roads, curved roads, ramp road sections, intersections, rotary islands and high-speed access ramps of urban roads, high-speed roads and the like;
the test scene comprises the daytime, sunny days and rainy days;
the traffic facilities in the test scene are road traffic signs and road marking lines;
the traffic participants in the test scenario are vehicles and other traffic participants.
3. The method of generating an autopilot vehicle test scenario of claim 1,
designing a motion mode of the test main vehicle, and then analyzing possible motion states of other traffic participants in a test scene, wherein the motion modes comprise:
analyzing possible combinations of relative positions and movement directions of the host vehicle and the traffic participants around the host vehicle aiming at the specified road traffic environment, and determining the relative position range of the host vehicle and the traffic participants around the host vehicle; and determining the possible movement direction of the main vehicle and the movement direction of each interference vehicle influencing the movement direction of the main vehicle according to the test target of the main vehicle control and perception function aiming at the complex scene group.
4. The method of generating an autopilot vehicle test scenario of claim 3,
determining a functional scene based on a hierarchical model according to the movement mode of the main vehicle and the movement states of the rest traffic participants, wherein the functional scene comprises the following steps:
sequentially arranging a basic road, a traffic infrastructure, a temporary static barrier, a traffic participant, an environment and an electronic communication facility according to the 1 st layer to the 6 th layer of the determined hierarchical model;
aiming at the test requirement, the necessary inputs of the test requirement are the road topology structure of the layer 1, the traffic environment signal provided by the layer 2 and the information of the traffic participants in the layer 4, wherein the information of the traffic participants in the layer 4 comprises the position and action information of the static and dynamic traffic participants including the test main vehicle.
5. The method of generating an autopilot vehicle test scenario of claim 3,
combining the possible movement direction permutation and combination of the possible movement directions of the main vehicle with the environment interference vehicle, adding necessary movement constraint conditions, and selecting a parameter combination coverage standard to automatically generate all combined scene groups, wherein the method comprises the following steps:
combining all possible motion directions determined by the main vehicle function with all possible motion directions of interfering vehicles according to the principle, adopting a PICT combination test tool during combination, adding necessary motion constraint conditions, selecting a parameter combination coverage standard to automatically generate all combined scene groups, designing a scene screening rule, screening out scenes with test values from the combined scene groups according to the scene screening rule, and forming a normal driving working condition scene group and a pre-collision working condition scene group, namely forming a basic scene library.
6. The method of claim 1, wherein screening a group of basic test scenarios that can cover the entire test scenario and function and that have a test value according to the screening rules of the scenarios comprises:
after the test scenes are generated by combining the possible motion directions of the main vehicle and all the traffic participants, similar scenes and scenes which do not exist in reality exist, the test scenes generated by the PICT are screened so as to obtain valuable test scenes, and the screening principle is set as follows: 1. scenes that cannot be realized in real traffic; 2. similar scenes exist, and only scenes with large influence on the main vehicle are reserved; 3. only one scene is reserved for scenes with the same influence on the movement of the main vehicle, and a basic test scene group which has test value and can cover the whole test scene and functions is screened out according to the screening principle of the test scene.
7. The method for generating the test scenario of the autonomous vehicle of claim 1, wherein a group of basic test scenarios that have test value and can cover the whole test scenario and function is screened out and mapped to a parameter space to generate a logic scenario, comprising:
extracting road topological structure, traffic signals, self-vehicle and other traffic participant function scene keywords to map into corresponding parameters according to the screened scene library so as to construct a parameter space and realize the generation of a logic scene; the parameter type obtained by mapping the functional scene keyword is called a key parameter type, other default parameter types required by the subsequent generation of a specific scene are called non-key parameter types, and the mapping of the functional scene to a parameter space is realized to generate a logic scene.
8. The method of claim 1, wherein the generating of the specific scene file in the opensceenario format by sampling the parameter space of the scene through the scene keyword comprises:
and automatically generating a specific scene file in an Open-SCENARIO format based on the functional scene keywords by analyzing the mapping relation between the logic scene parameter space and the OpenSCENARIO format.
9. The method of claim 8, wherein analyzing the mapping between the logical scene parameter space and the OpenSCENARIO format comprises:
analyzing an OpenSCENARIO format, and dividing the OpenSCENARIO format into a directory, an object and a storyboard 3-layer structure, wherein the directory corresponds to a road network file and a vehicle directory in a logic scene, the object corresponds to a vehicle, a pedestrian and a controller in a logic space, and the storyboard corresponds to the initialization of the logic scene, a storyboard sequence and an ending condition;
distinguishing key parameters and non-key parameters, and using the mapped non-key parameters as default values to simplify parameter space; key parameters corresponding to the road topological structure and the traffic signal are assigned in a road network file, namely Opendrive;
and determining the parameter range of each parameter type according to the test requirement to obtain a complete parameter space.
10. The method of claim 8, wherein the automatically generating specific scene files in Open-SCENARIO format based on the functional scene keywords comprises:
and the mapping relation adopts a modularized XML document object model to compile an automatic script, the automatic script samples the constructed parameter space based on the functional scene keywords, and generates corresponding OpenSCE-NARIO content by adopting XML DOM according to the mapping relation between the parameter space and the OpenSCENARIO format, so as to obtain a complete specific scene file in the OpenSCENARIO format.
11. The method of generating an autopilot test scenario of any one of claims 1-10,
completing an automatic driving test case based on an artificial potential field method and a model predictive control algorithm, and verifying the validity of the test case through simulation comprises the following steps:
and designing an automatic driving test case based on an artificial potential field and a model predictive control algorithm on the completed test scene, completing simulation test verification, and verifying the validity of the test case.
12. An automatic driving automobile test scenario generation system, which is implemented by the automatic driving automobile test scenario generation method according to any one of claims 1 to 11.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111416121.0A CN114168457A (en) | 2021-11-25 | 2021-11-25 | Method and system for generating test scene of automatic driving automobile |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111416121.0A CN114168457A (en) | 2021-11-25 | 2021-11-25 | Method and system for generating test scene of automatic driving automobile |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114168457A true CN114168457A (en) | 2022-03-11 |
Family
ID=80480818
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111416121.0A Pending CN114168457A (en) | 2021-11-25 | 2021-11-25 | Method and system for generating test scene of automatic driving automobile |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114168457A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115080388A (en) * | 2022-06-08 | 2022-09-20 | 中国科学院软件研究所 | Simulation test scene generation method for automatic driving system |
| CN115114729A (en) * | 2022-06-21 | 2022-09-27 | 重庆长安汽车股份有限公司 | Design method of user case of driving assistance system and storage medium |
| CN115257891A (en) * | 2022-05-27 | 2022-11-01 | 浙江众合科技股份有限公司 | CBTC scene testing method based on key position extraction and random position fusion |
| CN115687164A (en) * | 2023-01-05 | 2023-02-03 | 中汽智联技术有限公司 | Test case generalization screening method, equipment and storage medium |
| CN115880930A (en) * | 2023-01-19 | 2023-03-31 | 禾多科技（北京）有限公司 | Navigation route creation failure warning method, device, electronic device and medium |
| CN115906539A (en) * | 2023-01-10 | 2023-04-04 | 联友智连科技有限公司 | System for realizing automatic driving simulation based on Django |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110597711A (en) * | 2019-08-26 | 2019-12-20 | 湖南大学 | Automatic driving test case generation method based on scene and task |
| CN111841012A (en) * | 2020-06-23 | 2020-10-30 | 北京航空航天大学 | Automatic driving simulation system and test resource library construction method thereof |
-
2021
- 2021-11-25 CN CN202111416121.0A patent/CN114168457A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110597711A (en) * | 2019-08-26 | 2019-12-20 | 湖南大学 | Automatic driving test case generation method based on scene and task |
| CN111841012A (en) * | 2020-06-23 | 2020-10-30 | 北京航空航天大学 | Automatic driving simulation system and test resource library construction method thereof |
Non-Patent Citations (4)
| Title |
|---|
| 徐杨等: "无人车辆轨迹规划与跟踪控制的统一建模方法", 自动化学报, vol. 45, no. 4, 30 April 2019 (2019-04-30), pages 799 - 807 * |
| 李浩然等: "基于硬件在环的APF和MPC统一建模算法测试研究", 研究与应用, no. 10, 31 October 2021 (2021-10-31), pages 14 - 21 * |
| 舒红等: "自动驾驶汽车基础测试场景构建研究", 中国公路学报, vol. 32, no. 11, 30 November 2019 (2019-11-30), pages 245 - 254 * |
| 陈军毅等: "面向决策规划系统测试的具体场景自动化生成方法", 汽车技术, no. 10, 31 October 2020 (2020-10-31), pages 45 - 50 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115257891A (en) * | 2022-05-27 | 2022-11-01 | 浙江众合科技股份有限公司 | CBTC scene testing method based on key position extraction and random position fusion |
| CN115257891B (en) * | 2022-05-27 | 2024-06-04 | 浙江众合科技股份有限公司 | CBTC scene test method based on integration of key position extraction and random position |
| CN115080388A (en) * | 2022-06-08 | 2022-09-20 | 中国科学院软件研究所 | Simulation test scene generation method for automatic driving system |
| CN115114729A (en) * | 2022-06-21 | 2022-09-27 | 重庆长安汽车股份有限公司 | Design method of user case of driving assistance system and storage medium |
| CN115114729B (en) * | 2022-06-21 | 2024-06-04 | 重庆长安汽车股份有限公司 | Design method of driving assistance system user case and storage medium |
| CN115687164A (en) * | 2023-01-05 | 2023-02-03 | 中汽智联技术有限公司 | Test case generalization screening method, equipment and storage medium |
| CN115687164B (en) * | 2023-01-05 | 2023-04-14 | 中汽智联技术有限公司 | Test case generalization screening method, equipment and storage medium |
| CN115906539A (en) * | 2023-01-10 | 2023-04-04 | 联友智连科技有限公司 | System for realizing automatic driving simulation based on Django |
| CN115880930A (en) * | 2023-01-19 | 2023-03-31 | 禾多科技（北京）有限公司 | Navigation route creation failure warning method, device, electronic device and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114168457A (en) | Method and system for generating test scene of automatic driving automobile | |
| CN111061167B (en) | Intelligent network connection demonstration area-oriented mixed reality automatic driving test method and virtual test platform | |
| Li et al. | Survey on artificial intelligence for vehicles | |
| Queiroz et al. | GeoScenario: An open DSL for autonomous driving scenario representation | |
| CN115843347A (en) | Generating autonomous vehicle simulation data from recorded data | |
| CN103164579B (en) | Method for achieving interactive integration experimental platform of microscopic traffic simulator and driving simulator | |
| CN106198049A (en) | Real vehicles is at ring test system and method | |
| CN112115600A (en) | Simulation system for automatically driving vehicle | |
| CN114495527A (en) | Method and system for collaborative optimization of vehicle paths at internet intersection in mixed traffic environment | |
| Gao et al. | Hardware and software architecture of intelligent vehicles and road verification in typical traffic scenarios | |
| CN112035951A (en) | Simulation platform and simulation method for automatic driving algorithm verification | |
| CN113511215B (en) | Hybrid automatic driving decision method, device and computer storage medium | |
| CN109753623A (en) | A kind of analysis of automatic driving vehicle multi-test scene and number simplifying method | |
| CN108508761A (en) | A kind of AEB algorithm function verification methods based on CarMaker simulated environment | |
| CN115523934A (en) | Vehicle track prediction method and system based on deep learning | |
| CN115795808A (en) | Automatic driving decision dangerous scene generation method, system, equipment and medium | |
| CN111983934B (en) | Unmanned vehicle simulation test case generation method and system | |
| Schütt et al. | An application of scenario exploration to find new scenarios for the development and testing of automated driving systems in urban scenarios | |
| Damak et al. | Operational context-based design method of autonomous vehicles logical architectures | |
| CN117667699A (en) | Intelligent networking automobile test scene generation method based on knowledge graph | |
| CN117950399A (en) | Automatic obstacle avoidance decision method and system based on multi-mode knowledge graph | |
| Medrano-Berumen et al. | Development of a validation regime for an autonomous campus shuttle | |
| JP7414434B2 (en) | Recognition model distribution system and recognition model update method | |
| DE102021201177A1 (en) | Computer-implemented method and computer program for generating routes for an automated driving system | |
| Li | A scenario-based development framework for autonomous driving |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |