CN115879323B - Automatic driving simulation test method, electronic equipment and computer readable storage medium - Google Patents
Automatic driving simulation test method, electronic equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN115879323B CN115879323B CN202310051366.0A CN202310051366A CN115879323B CN 115879323 B CN115879323 B CN 115879323B CN 202310051366 A CN202310051366 A CN 202310051366A CN 115879323 B CN115879323 B CN 115879323B
- Authority
- CN
- China
- Prior art keywords
- simulation
- test
- simulation test
- vehicle
- automatic driving
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 304
- 238000010998 test method Methods 0.000 title claims abstract description 9
- 238000012360 testing method Methods 0.000 claims abstract description 251
- 238000000034 method Methods 0.000 claims abstract description 38
- 230000009471 action Effects 0.000 claims description 12
- 230000008447 perception Effects 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 7
- 238000005516 engineering process Methods 0.000 abstract description 6
- 230000006872 improvement Effects 0.000 abstract description 5
- 238000004891 communication Methods 0.000 description 20
- 238000010586 diagram Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000013507 mapping Methods 0.000 description 4
- 230000036544 posture Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 230000033772 system development Effects 0.000 description 3
- 206010039203 Road traffic accident Diseases 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000006798 recombination Effects 0.000 description 2
- 238000005215 recombination Methods 0.000 description 2
- 230000008521 reorganization Effects 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000013142 basic testing Methods 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008602 contraction Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000593 degrading effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000033001 locomotion Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The application provides an automatic driving simulation test method, electronic equipment and a computer readable storage medium, which realize large-scale cloud simulation without using a containerization technology and Kubernetes, and the specific method comprises the following steps: when the calculation complexity of the simulation test task is too high and the simulation frequency which can be realized by a single server cannot meet the requirement, the simulation test task is disassembled to a plurality of servers to cooperatively operate, and the output result is synchronously transmitted to the same automatic driving system. Through the improvement, the cloud simulation realizes the utilization rate of server resources approaching 100%, and the support of hardware on-loop simulation and whole-vehicle on-loop simulation can be realized no matter how complex the test task is.
Description
Technical Field
The application relates to the technical field of automatic driving, in particular to an automatic driving simulation test method, electronic equipment and a computer readable storage medium.
Background
At present, a container-based scheme is mostly adopted in cloud simulation for an auxiliary driving system or an automatic driving system, each instance is operated in one container, one server can operate a plurality of instances, and the containers are scheduled through Kubernetes, so that operations such as scheduling are realized. However, the above operation may cause overhead of resources such as CPU, further reducing the test efficiency.
Disclosure of Invention
An object of the embodiments of the present application is to provide an autopilot simulation test method, an electronic device, and a computer readable storage medium, which are used for solving the problems in the prior art that a containerization technology and Kubernetes are applied to simulation tests of an auxiliary driving system and an autopilot system, so that complex simulation tasks are difficult to disassemble, additional overhead is caused, and the utilization rate of server resources is reduced.
The embodiment of the application provides an automatic driving simulation test method, which comprises the following steps:
acquiring a simulation test task of a target automatic driving system, and determining simulation frequency and characteristic information of the simulation test task, wherein the characteristic information is used for representing the complexity of the simulation test task;
dividing the simulation test task into a plurality of simulation test subtasks according to the characteristic information of the simulation test task;
and synchronously distributing each simulation test subtask and the vehicle gesture corresponding to the target automatic driving system in the simulation test subtask to a plurality of servers, so that the plurality of servers can synchronously execute the distributed simulation test subtasks according to the simulation frequency, and performing simulation test according to the vehicle gesture of the target automatic driving system.
In the technical scheme, the large-scale cloud simulation is realized without using a containerization technology and Kubernetes, and the method specifically comprises the following steps: when the calculation complexity of the simulation test task is too high and the simulation frequency which can be realized by a single server cannot meet the requirement, the simulation test task is disassembled to a plurality of servers to cooperatively operate, and the output result is synchronously transmitted to the same automatic driving system. Through the improvement, the cloud simulation realizes the utilization rate of server resources approaching 100%, and the support of hardware on-loop simulation and whole-vehicle on-loop simulation can be realized no matter how complex the test task is.
In some alternative embodiments, the feature information of the simulated test task includes a simulation complexity of the test vehicle and/or a simulation complexity of the test scenario.
In the above technical solution, the feature information of the simulation test task includes simulation complexity of the test vehicle, for example, when splitting the simulation test task, the corresponding simulation test subtask may be split according to each sensor or positioning device of the test vehicle. The feature information of the simulation test task further comprises simulation complexity of the test scene, for example, when the simulation test task is split, the corresponding simulation test subtasks can be split respectively according to NPC vehicles and traffic participants (such as traffic lights and the like) in the test scene.
In some optional embodiments, obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes:
acquiring automatic driving system information of a test vehicle in a simulation test task, and determining a vehicle dynamics model solver corresponding to the automatic driving system of the test vehicle;
and determining the simulation complexity of the test vehicle according to the vehicle dynamics model solver.
In the above technical solution, the simulation complexity of the test vehicle is related to the vehicle dynamics model solver, and the relationship mapping tables of the simulation complexity of the test vehicle corresponding to different vehicle dynamics model solvers can be stored in advance, and when the simulation test task is disassembled, the relationship mapping tables are queried to obtain the simulation complexity of the test vehicle.
In some optional embodiments, obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes:
acquiring perception module information of a test vehicle in a simulation test task, and determining the type and the number of the perception modules;
and determining the simulation complexity of the test vehicle according to the type and the number of the sensing modules.
According to the technical scheme, the sensing module comprises the sensor and the positioning equipment, and the simulation complexity of the test vehicle can be determined according to the types and the quantity of the sensor and the positioning equipment, wherein the more the types and the quantity of the sensor and the positioning equipment are, the higher the simulation complexity of the test vehicle is, and the simulation test subtasks corresponding to the sensing module in the same type set quantity can be distributed to the same server when the simulation test tasks are disassembled.
In some optional embodiments, obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes:
acquiring scene elements of a test scene in a simulation test task, and determining action types corresponding to the scene elements;
and determining the simulation complexity of the test scene according to the action type corresponding to the scene element.
In the above technical solution, NPC vehicles, traffic participants, maps, and the like other than the test vehicle are collectively referred to as scene elements in the test scene, so that the simulation complexity of the test scene can be determined according to the action types corresponding to the scene elements.
In some alternative embodiments, dividing the simulation test task into a plurality of simulation test subtasks according to the feature information of the simulation test task includes:
dividing simulation test subtasks corresponding to the same type of sensing modules in the simulation test tasks into a server according to the types of different sensing modules in the simulation test tasks.
In some optional embodiments, after synchronously distributing each simulation test subtask and the vehicle gesture corresponding to the target autopilot system in the simulation test subtask to a plurality of servers, the method further includes:
setting simulation frequencies of different sensing modules, sending corresponding simulation frequencies to a plurality of servers, and ensuring that the servers are in a time synchronous state.
In some alternative embodiments, the multiple servers may synchronously execute the assigned simulation test subtasks according to the simulation frequency, and perform the simulation test according to the vehicle pose of the target autopilot system, including:
after each server executes the distributed simulation test subtasks, sensing data of at least one sensing module are obtained;
the method comprises the steps that sensing data obtained by simulating each server according to the simulation frequency of different sensing modules are respectively sent to corresponding input ends of an automatic driving system of a test vehicle;
and updating the posture of the test vehicle according to the control signal output by the automatic driving system.
In some alternative embodiments, the method is applied to hardware in-loop simulation or whole vehicle in-loop simulation, and the simulation frequency is sent to a plurality of servers, and the method comprises the following steps:
setting corresponding simulation frequencies according to the working frequency of a sensing module of an automatic driving system of a test vehicle, determining a server capable of realizing the corresponding simulation frequencies, and sending the corresponding simulation frequencies to a plurality of servers.
According to the technical scheme, when the calculation complexity of the simulation test task is too high and the simulation frequency which can be achieved by a single server cannot meet the requirements, the test task is disassembled to a plurality of servers to cooperatively operate, the output result is synchronously transmitted to the same automatic driving system, and the true frequency is set to be the working frequency of a sensor of the automatic driving system, so that hardware on-loop simulation and whole-vehicle on-loop simulation can be supported.
An electronic device provided in an embodiment of the present application includes: a processor and a memory storing machine-readable instructions executable by the processor, which when executed by the processor, perform a method as any one of the above.
A computer readable storage medium provided by an embodiment of the present application, on which a computer program is stored, which when executed by a processor performs a method as described in any of the above.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of steps of an automatic driving simulation test method according to an embodiment of the present application;
FIG. 2 is a functional block diagram of an autopilot simulation test system according to an embodiment of the present application;
FIG. 3 is a schematic diagram of an autopilot simulation test system according to an embodiment of the present application;
FIG. 4 is a schematic diagram of an autopilot simulation test system according to another embodiment of the present application;
fig. 5 is a schematic diagram of a possible structure of an electronic device according to an embodiment of the present application.
Icon: 100-user interface, 101-simulator cluster, 1011-simulator, 102-data bus, 103-adapter cluster, 1031-signal converter, 1032-model solver, 104-system under test, 105-simulation task scheduler, 21-processor, 22-memory, 23-communication interface, 24-communication bus.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
First, the following terms mentioned in one or more embodiments of the present application are annotated:
cloud simulation: a large-scale simulation test program is performed in a simulator cluster in a dynamic scheduling and deployment mode. The performance indexes of cloud simulation mainly comprise: flexibility, e.g., expansion, contraction time consumption; server resource utilization, such as CPU utilization, GPU utilization, storage utilization, etc.
Examples: cloud simulation consists of a series of instances, each running on a server. A test task or a part of the test task is performed. The number of the examples can be increased or reduced through operations such as capacity expansion and capacity shrinkage, so that the cloud simulation can simultaneously execute a large number of testing tasks on a large number of examples according to the needs, and the testing efficiency is improved.
The tested system comprises: the test object of the cloud simulation program comprises an automatic driving system and an auxiliary driving system.
Simulation frequency: the cloud simulation program generates an update frequency of the simulation data for the automated driving system. For example, the simulation frequency of a vehicle dynamics model is typically on the order of 100 Hz.
And a perception module: the electronic or optical device with sensing and/or positioning functions and the like is arranged in the automatic driving system, and can output sensing information such as images, videos, two-dimensional or three-dimensional point clouds, traffic participant information lists and the like of road environments and/or positioning information such as positions, speeds, postures, accelerations and the like of carriers where the device is positioned. Typical real devices include cameras, millimeter wave radars, lidars, global navigation systems (GNSS), inertial Measurement Units (IMU), and the like.
Operating frequency of the sensing module: frame rate of output data from a sensor or positioning device, etc. For example, the operating frequency of a camera is typically in the range of 20-60 Hz.
Closed loop simulation: the simulation method generates real and/or virtual sensor data and positioning equipment data according to information such as a map, a test scene, a test environment, the state of a tested vehicle, the state of an NPC vehicle and the like, and data acquired in an actual traffic environment, and sends the real and/or virtual sensor data and positioning equipment data to a tested object, the data output by the tested object is sent to a simulation or real vehicle to obtain the position and posture change of the vehicle, new data of a real or virtual sensor and/or positioning equipment are generated according to the change, the data are circularly reciprocated in this way, and qualitative and quantitative evaluation is continuously carried out on the output of the tested object and the state of the vehicle.
The model simulates in the loop: a closed loop simulation method, wherein the object to be tested is usually an algorithm and/or model of a software aided driving or automatic driving system.
Software in loop simulation: a closed loop simulation method is characterized in that the object to be tested is usually software of a software aided driving or automatic driving system.
The processor simulates in the loop: a closed loop simulation method, wherein the object to be tested is usually software of a software aided driving or automatic driving system on a target platform (usually a domain controller).
Hardware in loop simulation: a closed-loop simulation method is characterized in that a detected object is usually a domain controller, and the simulation frequency is equal to the working frequency of a device installed in an automatic driving system.
The whole vehicle is simulated in a ring: the object to be tested is usually a vehicle, and the simulation frequency is equal to the working frequency of devices installed in an automatic driving system.
At present, a container-based scheme is mostly adopted in cloud simulation for an auxiliary driving system or an automatic driving system, each instance is operated in one container, one server can operate a plurality of instances, and the containers are scheduled through Kubernetes, so that operations such as scheduling are realized. The key technical indexes of cloud simulation comprise overhead of a system and maximum simulation frequency. The overhead of the system comprises the utilization rate of resources such as a CPU (Central processing Unit) and a storage device of the server; the overhead of the system determines the cost of cloud simulation: the smaller the overhead, the lower the cost. When the maximum simulation frequency is not less than the working frequency of a device of the automatic driving system, the cloud simulation can realize hardware on-loop simulation and whole-vehicle on-loop simulation; otherwise, cloud simulation can only realize software on-loop simulation and model on-loop simulation.
The containerization technology and the Kubernetes are suitable for cloud computing application scenes with low computational complexity and extremely large concurrency. However, the characteristics of the simulation test application of the auxiliary driving system and the automatic driving system are contrary to the above: the computational complexity is particularly high and the number of complications is not particularly large. The calculation complexity of the simulation test is proportional to the number of devices, and the number of devices of the vehicle is rapidly increased along with the improvement of the intelligent degree. For example, high-level autopilot vehicles typically have over 10 cameras and millimeter wave radars, as well as multiple lidars, which makes the simulation frequency of test task instances running on a single server typically much lower than the operating frequency of the device, severely degrading test efficiency, and failing to support hardware-in-loop simulation and whole-vehicle-in-loop simulation. In addition, the containerization technology also causes the use cost of resources such as a CPU and the like, and further reduces the test efficiency. These defects can be generalized to the following two points:
1. the complex simulation task is difficult to disassemble, and a test case cannot be cooperatively executed by using a plurality of servers, so that the simulation frequency is generally lower than the working frequency of a device, and the requirements of hardware on-loop simulation and whole-vehicle on-loop simulation cannot be met;
2. causing additional overhead and reducing the utilization rate of server resources. For example, each container has its own file system, reducing the utilization of the storage device; the containers are isolated from each other and virtualize hardware resources, so that the utilization rate of the CPU and the GPU is reduced.
Referring to fig. 1, fig. 1 is a flowchart of steps of an automatic driving simulation test method provided in an embodiment of the present application, which specifically includes:
step S1, acquiring a simulation test task of a target automatic driving system, and determining simulation frequency and characteristic information of the simulation test task, wherein the characteristic information is used for representing complexity of the simulation test task;
s2, dividing the simulation test task into a plurality of simulation test subtasks according to the characteristic information of the simulation test task;
and S3, synchronously distributing each simulation test subtask and the vehicle gesture corresponding to the target automatic driving system in the simulation test subtask to a plurality of servers, so that the plurality of servers can synchronously execute the distributed simulation test subtasks according to the simulation frequency, and performing simulation test according to the vehicle gesture of the target automatic driving system.
In the embodiment of the application, the large-scale cloud simulation is realized without using a containerization technology and Kubernetes, and the specific method comprises the following steps: when the calculation complexity of the simulation test task is too high and the simulation frequency which can be realized by a single server cannot meet the requirement, the simulation test task is disassembled to a plurality of servers to cooperatively operate, and the output result is synchronously transmitted to the same automatic driving system. Through the improvement, the cloud simulation realizes the utilization rate of server resources approaching 100%, and the support of hardware on-loop simulation and whole-vehicle on-loop simulation can be realized no matter how complex the test task is.
In some alternative embodiments, the feature information of the simulated test task includes a simulation complexity of the test vehicle and/or a simulation complexity of the test scenario. In this embodiment, the feature information of the simulation test task includes simulation complexity of the test vehicle, for example, when splitting the simulation test task, the corresponding simulation test subtask may be split according to each sensor or positioning device of the test vehicle. The feature information of the simulation test task further comprises simulation complexity of the test scene, for example, when the simulation test task is split, the corresponding simulation test subtasks can be split respectively according to NPC vehicles and traffic participants (such as traffic lights and the like) in the test scene.
In some optional embodiments, obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes: acquiring automatic driving system information of a test vehicle in a simulation test task, and determining a vehicle dynamics model solver corresponding to the automatic driving system of the test vehicle; and determining the simulation complexity of the test vehicle according to the vehicle dynamics model solver.
In the embodiment of the application, the simulation complexity of the test vehicle is related to the vehicle dynamics model solver, and the relationship mapping tables of the simulation complexity of the test vehicle corresponding to different vehicle dynamics model solvers can be stored in advance, and when the simulation test task is disassembled in practice, the relationship mapping tables are queried to obtain the simulation complexity of the test vehicle.
In some optional embodiments, obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes: acquiring perception module information of a test vehicle in a simulation test task, and determining the type and the number of the perception modules; and determining the simulation complexity of the test vehicle according to the type and the number of the sensing modules.
In the embodiment of the application, the sensing module comprises a sensor and positioning equipment, and the simulation complexity of the test vehicle can be determined according to the types and the quantity of the sensor and the positioning equipment, wherein the more the types and the quantity of the sensor and the positioning equipment are, the higher the simulation complexity of the test vehicle is, and the simulation test subtasks corresponding to the sensing module in the same type set quantity can be distributed to the same server when the simulation test tasks are disassembled.
In some optional embodiments, obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes: acquiring scene elements of a test scene in a simulation test task, and determining action types corresponding to the scene elements; and determining the simulation complexity of the test scene according to the action type corresponding to the scene element.
In the embodiment of the application, NPC vehicles, traffic participants, maps and the like except for the test vehicles are collectively referred to as scene elements in the test scene, so that the simulation complexity of the test scene can be determined according to the action types corresponding to the scene elements.
The elements in the simulation test are mainly divided into two parts, namely a test vehicle element and a scene element; the scene elements in turn cover: static environmental elements, dynamic environmental elements, traffic participant elements, weather elements, and the like. Testing the elements of the vehicle: the method refers to testing basic attributes, position information, motion state information and driving task information of the vehicle. Static environmental elements: an object in a stationary state in a traffic environment, comprising: roads, traffic facilities, surrounding landscapes, obstructions, and the like. Dynamic environment elements: refers to elements in a traffic environment that are dynamically changing, including: facility and communication environment information is dynamically indicated. Traffic participant element: object information influencing the decision-making planning of the vehicle in an automatic driving test scene comprises the following steps: vehicles, pedestrians, and animals. Weather elements: including information such as ambient temperature, lighting conditions, and weather conditions in the driving scene.
Three phases that the autopilot system development needs to go through: concept phase, system development phase, test phase; with the gradual penetration of the system development process, the abstract degree requirement of the test scenes is continuously reduced, but the number requirement of the test scenes is continuously increased. The logical scene can be generated by converting the structured functional scene into a range of parameters, which can be defined by a data driven method. Each logical scene may be converted to a specific scene by selecting a specific value from the parameter range.
Wherein, the functional scene: describing entities in a scene area and relations among the entities through operation scenes of the most abstract level of semantic description, namely through language scene symbols; the functional scene is used for project definition, risk analysis and risk assessment of the concept stage; in the testing process, it is often necessary to convert the functional scene into a logical scene and into a data format usable in the corresponding simulation environment.
Logic scenario: expressing the relation between the entity characteristics and the entities by defining the parameter range of the variables in the state space; the logic scene is a further detailed description of the functional scene based on state space variables and is used for generating requirements in a project development stage; any number of specific scenes can be derived for each logical scene with a continuous range of values.
The specific scene is as follows: the relation between the entities is explicitly described by determining the specific value of each parameter in the state space, and the test scene is described in detail in the state space; the specific scene can be directly converted into a test case; to convert a specific scenario into a test case, the expected performance of the tested object and the need for related test facilities need to be increased.
Classifying according to the source of the test scene data: natural driving scene, dangerous working condition scene, standard regulation scene and parameter recombination scene. Wherein, natural driving scene: the data is the most basic data source under the real natural driving state scene of the automobile; all-round information such as people-vehicle-environment-task where the automatic driving vehicle is located is contained; the natural driving scene can provide multidimensional information such as vehicle data, driver behaviors, road environments and the like, and is a full test scene for proving the effectiveness of automatic driving. Dangerous working condition scene: the data mainly comes from a traffic accident database, and is the key for verifying the safety and reliability of the automatic driving control strategy; the dangerous working condition scene mainly covers three major types of scenes of severe weather environment and complex road traffic and typical traffic accidents, and is a necessary test scene for proving the effectiveness of automatic driving. Standard regulatory scenarios: the data mainly originate from the existing standards, evaluation rules and the like, such as ISO, NHTSA, E-NCAP, C-NCAP and the like, and the existing automatic driving functions are subjected to test rules by the evaluation rules; the standard legal test scenario is a basic test scenario that the autopilot function must meet during the development and certification phases. Parameter recombination scene: the data is derived from the existing scene database resources, and the corresponding types of scenes are randomly generated or automatically recombined by carrying out parameterization setting on the existing simulation scenes; the parameter reorganization scene expands the boundary of the parameter reorganization scene by carrying out different permutation and combination and traversing values on static elements, dynamic elements, driver behavior elements and the like; the test blind area of the automatic driving function is effectively covered, and the test scene is effectively supplemented for unknown working conditions.
In some alternative embodiments, dividing the simulation test task into a plurality of simulation test subtasks according to the feature information of the simulation test task includes: dividing simulation test subtasks corresponding to the same type of sensing modules in the simulation test tasks into a server according to the types of different sensing modules in the simulation test tasks.
In some optional embodiments, after synchronously distributing each simulation test subtask and the vehicle gesture corresponding to the target autopilot system in the simulation test subtask to a plurality of servers, the method further includes: setting simulation frequencies of different sensing modules, sending corresponding simulation frequencies to a plurality of servers, and ensuring that the servers are in a time synchronous state.
In some alternative embodiments, the multiple servers may synchronously execute the assigned simulation test subtasks according to the simulation frequency, and perform the simulation test according to the vehicle pose of the target autopilot system, including: after each server executes the distributed simulation test subtasks, sensing data of at least one sensing module are obtained; the method comprises the steps that sensing data obtained by simulating each server according to the simulation frequency of different sensing modules are respectively sent to corresponding input ends of an automatic driving system of a test vehicle; and updating the posture of the test vehicle according to the control signal output by the automatic driving system.
In some alternative embodiments, the method is applied to hardware in-loop simulation or whole vehicle in-loop simulation, and the simulation frequency is sent to a plurality of servers, and the method comprises the following steps: setting corresponding simulation frequencies according to the working frequency of a sensing module of an automatic driving system of a test vehicle, determining a server capable of realizing the corresponding simulation frequencies, and sending the corresponding simulation frequencies to a plurality of servers.
In the embodiment of the application, when the calculation complexity of the simulation test task is too high and the simulation frequency which can be realized by a single server cannot meet the requirement, the test task is disassembled to a plurality of servers to cooperatively operate, the output result is synchronously transmitted to the same automatic driving system, and the true frequency is set to be the working frequency of a sensor of the automatic driving system, so that the hardware on-loop simulation and the whole-vehicle on-loop simulation can be supported.
Referring to fig. 2, fig. 2 is a functional block diagram of an autopilot simulation test system according to an embodiment of the present application, where the system includes a user interface 100, a simulator cluster 101, a data bus 102, an adapter cluster 103, a tested system 104, a simulation task scheduler 105, and other functional blocks. The environment can be used for application scenes such as model on-loop simulation, software on-loop simulation, processor on-loop simulation, hardware on-loop simulation, whole vehicle on-loop simulation and the like.
The user interface 100 is: a software program running on the user device with a graphical interface and/or a text interface in which a user of the cloud simulation may submit a simulation test task and view and process the results of the simulation test.
The simulator cluster 101 is composed of a plurality of simulator software instances running on a single server or a plurality of servers, and can send data such as simulation results to the user interface 100; each simulator software instance runs on a server and comprises the functions of map analysis and rendering, simulation scene construction, sensor modeling program, simulation result transmission and the like; the communication delay between emulators is typically on the order of 1 millisecond and the communication bandwidth is on the order of 1-10 Gbps. Wherein the simulator cluster 101 comprises K simulators 1011, K being a positive integer greater than 0, each simulator 1011 being a certain simulator software instance located in the simulator cluster 101.
The data bus 102 is used to connect the data interfaces of the different emulators and the different signal converters. When the emulator cluster and the signal translator are located in the same server, the data bus 102 is implemented by an interprocess communication mechanism, such as pipes, message queues, etc. When the emulator cluster and signal converter are located in multiple servers, the data bus 102 is comprised of a high-speed switch, communication cable, and communication protocol, with a communication bandwidth typically on the order of 1-10 Gbps, and with very low communication delay and packet loss rate; common communication protocols include ethernet, infiniBand (IB), and the like.
The adapter cluster 103 is used to adapt the data interface of the auxiliary or autopilot system. Software and/or hardware devices for modeling the dynamics of an autopilot system have a data interface and are connected to the data bus 102. Adapter cluster 103 is used for software-in-loop, model-in-loop, and processor-in-loop simulation, and is typically comprised of software programs running on one or more servers; the adapter cluster 103 is used for hardware-in-loop and whole-vehicle-in-loop simulation, and is typically composed of a series of hardware devices and installed in a cabinet.
The adapter cluster 103 comprises N signal converters 1031, N being a positive integer greater than 0. Each signal sensor 1031 includes communication and signal conversion functions, which receives data (i.e., simulation results) sent by a server in the simulator cluster through the data bus 102, and converts the data into data required by the tested system 104 in real time, such as data structures in software programs, bus data, electrical signals, optical signals, etc.; these signals appear to the system under test 104 as not differing from the signals sent by the real sensor and the positioning device, thus enabling real-time high-fidelity simulation of the sensor.
The number of signal converters N is equal to the total number of all sensors and positioning devices that all autopilot systems have, N may be greater than, equal to, or less than K; the data connection between the K servers and the N signal converters is defined by the simulation task scheduler 105.
The form of the data output by the signal converter to the system under test 104 depends on the nature of the simulation test task. For example, for model-in-loop, software-in-loop, and processor-in-loop testing, the form of data is typically a structure, array, and/or object that conforms to some definition, etc.; for hardware-in-loop and whole-vehicle-in-loop testing, the data is typically in the form of bus data, electrical signals, and/or optical signals, etc.
There may be M systems under test 104, i.e., the autopilot systems and the model solver are in one-to-one correspondence, and each system under test 104 is an auxiliary or autopilot system under test, which typically consists of a software program and/or a domain controller and its corresponding peripherals, power supplies, and software programs running in the domain controller.
An autopilot system may contain L sensors or positioning devices etc. that need to be emulated, requiring L signal converters and 1 model solver in the adapter cluster to be adapted to.
The simulation task scheduler 105 is a software program running in a server, which receives test tasks submitted by a user through the user interface 100, analyzes the computational complexity of the tasks, disassembles the simulation test tasks if necessary, and distributes the disassembled test tasks to the corresponding servers in the simulator cluster.
When the operating frequency of all the functional modules in the system is strictly equal to the operating frequency of the device, the system can be used for hardware on-loop simulation and whole-vehicle on-loop simulation; otherwise, the system can only be used for model-in-loop simulation, software-in-loop simulation and processor-in-loop simulation.
In a specific embodiment, please refer to fig. 3, fig. 3 is a schematic diagram of an autopilot simulation test system provided in an embodiment of the present application, wherein the autopilot system has 3 sensors, 3 servers and 3 signal converters. The simulation task dispatcher receives a simulation test task, then disassembles the simulation test task into 3 test tasks corresponding to different sensors, distributes the test tasks of each sensor to corresponding servers in the simulation task dispatcher for simulation, converts each simulation result into a signal adapting to a data input interface of the corresponding sensor through a corresponding signal converter, and inputs the signal into the automatic driving system.
In another embodiment, referring to fig. 4, fig. 4 is a schematic diagram of an autopilot simulation test system according to another embodiment of the present application, wherein the autopilot system has 3 sensors, 3 servers and 3 signal converters. The simulation task scheduler receives 3 simulation test tasks, and distributes each simulation test task to a corresponding server for simulation, and the simulation result of each server comprises the simulation results of 3 sensors, so that each server needs to respectively send the simulation results of 3 sensors to a corresponding signal converter, and each signal sensor is only adapted to one sensor interface of the automatic driving system.
Fig. 5 shows a possible structure of the electronic device provided in the embodiment of the present application. Referring to fig. 5, the electronic device includes: processor 21, memory 22, and communication interface 23, which are interconnected and communicate with each other by a communication bus 24 and/or other forms of connection mechanisms (not shown).
The Memory 22 includes one or more (Only one is shown in the figure), which may be, but is not limited to, a random access Memory (Random AccessMemory, RAM), a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an erasable programmable Read Only Memory (ErasableProgrammable Read-Only Memory, EPROM), an electrically erasable programmable Read Only Memory (ElectricErasable Programmable Read-Only Memory, EEPROM), and the like. The processor 21 and possibly other components may access the memory 22, read and/or write data therein.
The processor 21 comprises one or more (only one shown) which may be an integrated circuit chip having signal processing capabilities. The processor 21 may be a general-purpose processor, including a central processing unit (CentralProcessing Unit, CPU), a micro control unit (Micro ControllerUnit, MCU), a network processor (Network Processor, NP), or other conventional processor; but may also be a special purpose processor including a Neural Network Processor (NPU), a graphics processor (GraphicsProcessing Unit GPU), a digital signal processor (Digital Signal Processor DSP), an application specific integrated circuit (ApplicationSpecific Integrated Circuits ASIC), a field programmable gate array (FieldProgrammable Gate Array FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. Also, when the number of processors 21 is plural, some of them may be general-purpose processors, and the other may be special-purpose processors.
The communication interface 23 includes one or more (only one shown) that may be used to communicate directly or indirectly with other devices for data interaction. The communication interface 23 may include an interface for wired and/or wireless communication.
One or more computer readable instructions may be stored in memory 22 that may be read and executed by processor 21 to implement the methods provided by embodiments of the present application.
It will be appreciated that the configuration shown in fig. 5 is merely illustrative, and that the electronic device may also include more or fewer components than shown in fig. 5, or have a different configuration than shown in fig. 5. The components shown in fig. 5 may be implemented in hardware, software, or a combination thereof. The electronic device may be a physical device such as a PC, a notebook, a tablet, a cell phone, a server, an embedded device, etc., or may be a virtual device such as a virtual machine, a virtualized container, etc. The electronic device is not limited to a single device, and may be a combination of a plurality of devices or a cluster of a large number of devices.
The present embodiments also provide a computer readable storage medium having computer readable instructions stored thereon, which when read and executed by a processor of a computer, perform the method provided by the embodiments of the present application. For example, the computer readable storage medium may be implemented as the memory 22 in the electronic device of FIG. 5.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
Further, the units described as separate units may or may not be physically separate, and units displayed as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
Furthermore, functional modules in various embodiments of the present application may be integrated together to form a single portion, or each module may exist alone, or two or more modules may be integrated to form a single portion.
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.
The foregoing is merely exemplary embodiments of the present application and is not intended to limit the scope of the present application, and various modifications and variations may be suggested to one skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
Claims (11)
1. An automated driving simulation test method, the method comprising:
acquiring a simulation test task of a target automatic driving system, and determining simulation frequency and characteristic information of the simulation test task, wherein the characteristic information is used for representing the complexity of the simulation test task; the complexity of the simulation test task can be determined according to at least one of the type and the number of the perception modules of the test vehicle, a vehicle dynamics model corresponding to an automatic driving system of the test vehicle and the action type of a scene element of the test scene;
dividing the simulation test task into a plurality of simulation test subtasks according to the characteristic information of the simulation test task;
and synchronously distributing each simulation test subtask and the vehicle gesture corresponding to the target automatic driving system in the simulation test task to a plurality of servers, so that the plurality of servers can synchronously execute the distributed simulation test subtasks according to the simulation frequency, and performing simulation test according to the vehicle gesture of the target automatic driving system.
2. The method of claim 1, wherein the feature information of the simulated test task comprises a simulation complexity of a test vehicle and/or a simulation complexity of a test scenario.
3. The method of claim 2, wherein the obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, comprises:
acquiring automatic driving system information of a test vehicle in the simulation test task, and determining a vehicle dynamics model solver corresponding to the automatic driving system of the test vehicle;
and determining the simulation complexity of the test vehicle according to the vehicle dynamics model solver.
4. The method of claim 2, wherein the obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, comprises:
acquiring perception module information of a test vehicle in the simulation test task, and determining the type and the number of the perception modules;
and determining the simulation complexity of the test vehicle according to the type and the number of the sensing modules.
5. The method according to any one of claims 2-4, wherein the obtaining a simulation test task of the target autopilot system, determining feature information of the simulation test task, includes:
acquiring scene elements of a test scene in the simulation test task, and determining action types corresponding to the scene elements;
and determining the simulation complexity of the test scene according to the action type corresponding to the scene element.
6. The method of claim 4, wherein dividing the simulation test task into a plurality of simulation test subtasks according to the feature information of the simulation test task, comprises:
dividing simulation test subtasks corresponding to the same type of sensing modules in the simulation test tasks into a server according to the types of different sensing modules in the simulation test tasks.
7. The method of claim 6, wherein after synchronously distributing each simulation test subtask and the vehicle gesture corresponding to the target autopilot system in the simulation test subtask to the plurality of servers, further comprises:
setting simulation frequencies of different sensing modules, sending corresponding simulation frequencies to the servers, and ensuring that the servers are in a time synchronization state.
8. The method of claim 7, wherein the plurality of servers are operable to synchronously execute the assigned simulation test subtasks at the simulation frequency, and wherein performing the simulation test based on the vehicle pose of the target autopilot system comprises:
after each server executes the distributed simulation test subtasks, sensing data of at least one sensing module are obtained;
the sensing data obtained by simulating the different sensing modules according to the simulation frequencies of the different sensing modules are respectively sent to the corresponding input ends of an automatic driving system of the test vehicle;
and updating the posture of the test vehicle according to the control signal output by the automatic driving system.
9. The method according to any one of claims 7-8, wherein the sending the corresponding simulation frequencies to the plurality of servers when the method is applied to hardware-in-loop simulation or whole-vehicle-in-loop simulation comprises:
setting corresponding simulation frequencies according to the working frequency of the sensing module of the automatic driving system of the test vehicle, determining a server capable of realizing the corresponding simulation frequencies, and sending the corresponding simulation frequencies to a plurality of servers.
10. An electronic device, comprising: a processor and a memory storing machine-readable instructions executable by the processor, which when executed by the processor, perform the method of any of claims 1-9.
11. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when executed by a processor, performs the method according to any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310051366.0A CN115879323B (en) | 2023-02-02 | 2023-02-02 | Automatic driving simulation test method, electronic equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310051366.0A CN115879323B (en) | 2023-02-02 | 2023-02-02 | Automatic driving simulation test method, electronic equipment and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115879323A CN115879323A (en) | 2023-03-31 |
| CN115879323B true CN115879323B (en) | 2023-05-23 |
Family
ID=85758632
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310051366.0A Active CN115879323B (en) | 2023-02-02 | 2023-02-02 | Automatic driving simulation test method, electronic equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115879323B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116126740B (en) * | 2023-04-18 | 2023-08-04 | 小米汽车科技有限公司 | Model in-loop test method and device, computer equipment and storage medium |
| CN117271374A (en) * | 2023-11-20 | 2023-12-22 | 北京开源芯片研究院 | Simulation test method, device and equipment for chip and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115185205A (en) * | 2022-08-30 | 2022-10-14 | 北斗星通智联科技有限责任公司 | Intelligent driving simulation method, system, terminal device and readable storage medium |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102016220670A1 (en) * | 2015-11-06 | 2017-05-11 | Ford Global Technologies, Llc | Method and system for testing software for autonomous vehicles |
| CN106200612B (en) * | 2016-07-07 | 2019-01-22 | 百度在线网络技术（北京）有限公司 | For testing the method and system of vehicle |
| US10902165B2 (en) * | 2018-01-09 | 2021-01-26 | The Charles Stark Draper Laboratory, Inc. | Deployable development platform for autonomous vehicle (DDPAV) |
| CN112148481B (en) * | 2020-09-10 | 2022-11-22 | 苏州浪潮智能科技有限公司 | Method, system, equipment and medium for executing simulation test task |
| CN114692289A (en) * | 2020-12-31 | 2022-07-01 | 华为技术有限公司 | Automatic driving algorithm testing method and related equipment |
| CN112926224B (en) * | 2021-03-30 | 2024-02-02 | 深圳安途智行科技有限公司 | Event-based simulation method and computer equipment |
| CN113848749A (en) * | 2021-08-31 | 2021-12-28 | 际络科技(上海)有限公司 | Automatic driving simulation test system, method, electronic device and storage medium |
| CN114442507A (en) * | 2021-12-17 | 2022-05-06 | 际络科技(上海)有限公司 | Vehicle in-loop automatic driving simulation test method and system based on frequency control |
| CN114356757A (en) * | 2021-12-22 | 2022-04-15 | 重庆长安汽车股份有限公司 | Test condition configuration method based on limited automatic driving simulation scene |
| CN114936156A (en) * | 2022-05-19 | 2022-08-23 | 国汽智控(北京)科技有限公司 | Method and device for processing automatic driving simulation task |
| CN115016323A (en) * | 2022-06-21 | 2022-09-06 | 际络科技(上海)有限公司 | Automatic driving simulation test system and method |
-
2023
- 2023-02-02 CN CN202310051366.0A patent/CN115879323B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115185205A (en) * | 2022-08-30 | 2022-10-14 | 北斗星通智联科技有限责任公司 | Intelligent driving simulation method, system, terminal device and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115879323A (en) | 2023-03-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115879323B (en) | Automatic driving simulation test method, electronic equipment and computer readable storage medium | |
| US10635844B1 (en) | Methods and systems for simulating vision sensor detection at medium fidelity | |
| CN108345290B (en) | Method for testing at least one control device function of at least one control device | |
| US20190130056A1 (en) | Deterministic Simulation Framework for Autonomous Vehicle Testing | |
| CN110991040A (en) | Complex product collaborative simulation environment construction method | |
| CN113419518B (en) | VIL test platform based on VTS | |
| KR102606423B1 (en) | Method, apparatus, and device for testing traffic flow monitoring system | |
| CN115185205A (en) | Intelligent driving simulation method, system, terminal device and readable storage medium | |
| Kim et al. | Collaborative visualization of a warfare simulation using a commercial game engine | |
| Lucas et al. | Remote simulation to evaluate real-time traffic control strategies | |
| CN115384526A (en) | Debugging system and debugging method | |
| CN116403174A (en) | End-to-end automatic driving method, system, simulation system and storage medium | |
| CN114707242A (en) | Universal airborne data bus monitoring and testing software framework design method | |
| Feilhauer et al. | A multi-domain simulation approach to validate Advanced Driver Assistance Systems | |
| US20210173705A1 (en) | Method and apparatus for software isolation and security utilizing multi-soc orchestration | |
| Noth et al. | An integrated architecture for the development and assessment of adas | |
| JP7441981B2 (en) | System and method for providing autonomous driving simulation architecture with switchable models | |
| US11772681B2 (en) | Method and apparatus for processing autonomous driving simulation data, and electronic device | |
| Chaudron et al. | Real-time aircraft simulation using HLA standard | |
| CN112306520B (en) | Container-based vehicle networking simulation test method and electronic equipment | |
| US20230409383A1 (en) | Simulation system and method | |
| Bauer et al. | Usage of Operating System Kernel Paradigms to Build Modern Simulation Platforms for Autonomous Driving | |
| Gaska et al. | Leveraging driverless car investment in next generation integrated modular avionics (IMA) | |
| CN117873910A (en) | Rapid prototyping system, method, apparatus, storage medium, and product for intelligent algorithms | |
| Harris et al. | A research tool for the power and performance analysis of sensor-based mobile robots |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: Room 533, 5th Floor, Building A3A4, Phase I, Zhong'an Chuanggu Science and Technology Park, No. 900 Wangjiang West Road, High tech Zone, Hefei City, Anhui Province, 230031 Patentee after: Anhui Xinxin Science and Technology Innovation Information Technology Co.,Ltd. Address before: 2nd Floor, Building B2, Yunhui Valley, No. 156, Tiangu 8th Road, Software New Town, Yuhua Street Office, High-tech Zone, Xi'an City, Shaanxi Province 710000 Patentee before: Xi'an Xinxin Information Technology Co.,Ltd. |
|
| CP03 | Change of name, title or address |