How to verify autonomous driving functions based on a VIL system
The method generates a virtual road environment using a digital twin to convert actual vehicle positions, addressing the limitations of conventional VIL systems by enabling flexible and efficient autonomous driving function verification across diverse road conditions.
Patent Information
- Application Number
- JP2023067390
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2022-08-31
- Filing Date
- 2023-04-17
- Publication Date
- 2024-12-09
- Estimated Expiration
- 2043-04-17
AI Technical Summary
Conventional VIL systems for autonomous vehicle function verification are limited by the need to replicate actual road environments, requiring significant time and economic effort to test in different road conditions, and are constrained by the virtual road environment, making it difficult to verify autonomous driving functions in complex scenarios.
A method using a digital twin to generate a virtual road environment, converting the actual position and orientation of the autonomous vehicle into a virtual position and orientation, allowing verification without replicating the actual road, using a simulation engine, rendering module, emulation module, and interlocking module to synchronize sensor data and vehicle position.
Enables autonomous driving function verification in various virtual road environments without changing the driving location, speeding up development and improving convenience by allowing flexible road environment modifications.
Smart Images

Figure 0007596432000012 
Figure 0007596432000013 
Figure 0007596432000014
Abstract
Description
[Technical field]
[0001] The present invention relates to a method for verifying the functions of an autonomous vehicle, and more particularly to a system and method for verifying autonomous driving functions based on Vehicle-In-the-Loop (VIL) that links an autonomous vehicle with a simulator. [Background technology]
[0002] Actual vehicle-based verification is used to verify Advanced Driver Assistance Systems (ADAS) or autonomous driving functions, but since there are limitations to verifying all road conditions, simulator-based verification is also carried out in parallel. The autonomous driving simulator verifies the functionality of the recognition / judgment / control algorithms for autonomous driving through a simulation of a virtual autonomous vehicle running in a virtual road environment. However, since there is a limit to perfectly replicating the performance and status of an autonomous vehicle that is actually running, a VIL system-based autonomous driving function verification method is used, which verifies the simulator in conjunction with an actual autonomous vehicle. The virtual road environment used to verify the autonomous driving function based on the conventional VIL system is a direct replica of the test road environment such as a proving ground (PG) where the autonomous vehicle actually drives. Therefore, in the conventional VIL system, the actual position / attitude of the autonomous vehicle is used as the virtual position / attitude of the virtual autonomous vehicle in the virtual road environment. The verification of autonomous driving functions using such conventional VIL systems has a large restriction on the virtual road environment. In other words, verification in various virtual road environments requires a lot of time and effort because it is necessary to go to the actual roads that are copies of the virtual roads and verify the autonomous driving functions while driving the autonomous vehicle.
[0003] As the level of autonomous driving advances, the need to verify autonomous driving functions in complex road environments also increases. For this reason, with conventional VIL systems, it is necessary to build or change test roads every time a different road environment is required for verification, which is time-consuming and economical. Summary of the Invention [Problem to be solved by the invention]
[0004] Therefore, the present invention has been made in consideration of the above problems, and an object of the present invention is to provide a method for verifying the autonomous driving performance of an autonomous vehicle based on a VIL system, even if the virtual road environment is not a direct replica of an actual test road such as a PG, but a certain target area is created as a virtual environment. [Means for solving the problem]
[0005] To achieve the above object, an autonomous driving simulator according to one embodiment of the present invention includes a road environment DB in which information about a road environment is stored, a simulation engine that generates a virtual road environment using the information stored in the road environment DB, identifies a position and attitude of a virtual autonomous vehicle, generates a virtual NPC (Non Player Character) object, and generates a simulation scenario for verifying autonomous driving performance, a rendering module that renders the virtual road environment according to the scenario generated by the simulation engine, an emulation module that emulates sensors attached to the virtual autonomous vehicle in the virtual road environment rendered by the rendering module and generates virtual sensor data, a calculation module that converts the actual position and attitude of the autonomous vehicle into the position and attitude of the virtual autonomous vehicle in the virtual road environment and transmits them to the simulation engine, and an interlocking module that transmits the virtual sensor data generated by the emulation module to the autonomous vehicle, receives the actual position and attitude of the autonomous vehicle from the autonomous vehicle, and transmits them to the calculation module.
[0006] The actual position and attitude of the autonomous vehicle may differ from the virtual position and attitude of the virtual autonomous vehicle. The road on which the autonomous vehicle travels may be different from the road whose information is stored in the road environment DB. The road whose information is stored in the road environment DB may be changed while the road on which the autonomous vehicle travels remains unchanged. The roads whose information is stored in the road environment DB may include virtual roads and real roads. The real roads may include PGs, general roads, and expressways as roads that the autonomous vehicle can actually travel on in order to verify its functions. The calculation module may convert the actual position and orientation of the autonomous vehicle into a virtual position and orientation of the virtual autonomous vehicle in the virtual road environment. The calculation module may refer to information about the road shape stored in the road environment DB and calculate a virtual attitude of the virtual autonomous vehicle based on the virtual position and orientation.
[0007] The computation module may initialize and begin transforming the actual position and attitude of the autonomous vehicle and the position and attitude of the virtual autonomous vehicle if an initialization command is present. An autonomous driving simulation method according to another embodiment of the present invention includes the steps of: an autonomous driving simulator generating a virtual road environment, identifying a position and attitude of a virtual autonomous vehicle, generating a virtual NPC object, and generating a simulation scenario for verifying autonomous driving performance; a step of the autonomous driving simulator rendering the virtual road environment according to the generated scenario; a step of the autonomous driving simulator emulating sensors attached to the virtual autonomous vehicle in the rendered virtual road environment and generating virtual sensor data; a step of the autonomous driving simulator transmitting the generated virtual sensor data to the autonomous vehicle and receiving an actual position and attitude of the autonomous vehicle from the autonomous vehicle; and a step of the autonomous driving simulator converting the received actual position and attitude of the autonomous vehicle into a position and attitude of the virtual autonomous vehicle in the virtual road environment.
[0008] An autonomous driving simulator according to yet another embodiment of the present invention includes a simulation engine that generates a virtual road environment, identifies a position and attitude of a virtual autonomous vehicle, and generates a simulation scenario for verifying autonomous driving performance, a calculation module that converts the actual position and attitude of the autonomous vehicle into a position and attitude of the virtual autonomous vehicle in the virtual road environment and transmits it to the simulation engine, and an interworking module that receives the actual position and attitude of the autonomous vehicle from the autonomous vehicle and transmits it to the calculation module. An autonomous driving simulation method according to yet another embodiment of the present invention includes a step in which an autonomous driving simulator transmits virtual sensor data to an autonomous vehicle and receives an actual position and attitude of the autonomous vehicle from the autonomous vehicle, a step in which the autonomous driving simulator converts the actual position and attitude of the autonomous vehicle into a position and attitude of a virtual autonomous vehicle in a virtual road environment, and a step in which the autonomous driving simulator generates a simulation scenario for verifying autonomous driving performance in the virtual road environment based on the converted position and attitude of the virtual autonomous vehicle. Effect of the Invention
[0009] As described above, according to an embodiment of the present invention, it becomes possible to verify an autonomous driving function by linking an autonomous vehicle to a virtual road environment in a separate location without the autonomous vehicle having to go directly to an actual test road that replicates the virtual road environment. Furthermore, according to the embodiment of the present invention, it becomes possible to quickly verify autonomous driving functions based on the VIL system for various virtual road environments without changing the driving location, which will speed up the development of autonomous driving technology and improve its convenience. [Brief description of the drawings]
[0010] [Figure 1] FIG. 1 is a diagram showing a configuration of a VIL system according to an embodiment of the present invention. [Diagram 2] FIG. 1 is a diagram for explaining a method for converting an actual position / direction into a virtual position / direction. [Diagram 3]13 is a diagram illustrating a method for synchronizing position / posture information in the VIL mode. FIG. [Figure 4] FIG. 13 is a diagram for explaining a method for synchronizing position / posture information when not in VIL mode. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0011] In the following, the invention will be explained in more detail with reference to the drawings. In the embodiment of the present invention, a method for verifying autonomous driving functions based on a Vehicle-In-the-Loop (VIL) system is presented, which enables the verification of autonomous driving functions based on the VIL system even when the autonomous vehicle is on a different road or location than the actual road replicated by the virtual road environment. According to an embodiment of the present invention, there is no restriction on the actual roads on which the autonomous vehicle drives, i.e., it becomes possible to verify the autonomous driving function based on the VIL system regardless of where the autonomous vehicle drives. To this end, in an embodiment of the present invention, a virtual road that replicates the roads on which an autonomous vehicle can travel is constructed using a digital twin, the actual position / posture of the autonomous vehicle is converted into a virtual position / posture in the virtual road environment, and the position / posture of the autonomous vehicle on the actual road and the virtual autonomous vehicle on the virtual road are synchronized. 1 is a diagram showing the configuration of a VIL system according to an embodiment of the present invention. As shown in the figure, the VIL system according to the embodiment of the present invention includes an autonomous driving simulator 100 and an autonomous driving vehicle 200.
[0012] The autonomous driving simulator 100 is a computing system that generates a virtual road environment that replicates an actual road environment using a digital twin to provide a virtual road environment for verifying the functions of an autonomous driving vehicle 200. The autonomous vehicle 200 drives while performing vehicle control according to the virtual road environment generated by the autonomous driving simulator 100, but the location where the autonomous vehicle 200 actually drives may not be an actual road that replicates the virtual road environment, but may be a different road or location, such as a road or square where no cars pass. As a result, there is a difference between the actual position / attitude of the autonomous vehicle 200 and the virtual position / attitude of a virtual autonomous vehicle that replicates the autonomous vehicle 200 in the virtual road environment, so the autonomous driving simulator 100 converts the actual position / attitude of the autonomous vehicle 200 into the virtual position / attitude of the virtual autonomous vehicle.
[0013] The following is a detailed description of the configuration of the autonomous driving simulator 100. The autonomous driving simulator 100 includes a road environment DB 110, a simulation engine 120, a content DB 130, a road environment rendering module 140, a sensor emulation module 150, a position / attitude calculation module 160, and an interlocking module 170, as shown in FIG. The road environment DB 110 stores information on the environment of real roads, such as roads (road shape, road width, number of lanes), traffic lights, and static objects (road trees, street lights, buildings, etc.). The roads whose information is stored in the road environment DB110 may be PG (Proving Ground) roads on which autonomous vehicle 200 can actually travel and verify functions, or they may be actual roads. In the case of actual roads, the type of road is not important, including not only general roads but also expressways. However, it should be noted that the roads whose information is stored in road environment DB 110 are roads on which autonomous vehicle 200 can actually travel, and in an embodiment of the present invention, autonomous vehicle 200 does not travel on those roads, but rather travels on other roads or locations. The roads whose information is stored in the road environment DB 110 are determined according to need and purpose, and can be modified, edited, or otherwise altered.
[0014] The simulation engine 120 uses the road environment DB 110 to generate a virtual road environment, identify the position / posture of the virtual autonomous vehicle, generate virtual NPC (Non Player Character) objects, and generate a simulation scenario for verifying the autonomous driving performance. An NPC object is a dynamic object in a virtual road environment, such as a vehicle, a person, or an animal other than the virtual autonomous vehicle. In a simulation scenario, the NPC object moves, and may cause an accident due to abnormal movement, as well as normal movement. The content DB 130 stores content necessary to configure the road environment stored in the road environment DB 110 as a virtual environment, i.e., a digital twin of an actual road, and to add NPCs to the virtual environment. The road environment rendering module 140 uses the content stored in the content DB 130 to render a virtual road environment according to a scenario generated by the simulation engine 120 .
[0015] The sensor emulation module 150 emulates sensors such as cameras, lidar, radar, GNSS / INS, V2X, etc. attached to the virtual autonomous vehicle in the virtual road environment rendered by the road environment rendering module 140, and generates virtual sensor data. The sensors installed in the virtual autonomous vehicle for which virtual sensor data is generated by sensor emulation module 150 are implemented in the same manner as the sensors actually installed in autonomous vehicle 200. Position / attitude calculation module 160 converts the actual position / attitude of autonomous vehicle 200 into a position / attitude in the virtual road environment. In the case of conventional VIL system-based verification, the actual position / attitude of autonomous vehicle 200 matches the virtual position / attitude of the virtual autonomous vehicle in the virtual road environment, but in the embodiment of the present invention, the two do not match, so position / attitude conversion by position / attitude calculation module 160 is necessary. Interlocking module 170 transmits the virtual sensor data generated by sensor emulation module 150 and the virtual position / attitude information calculated by position / attitude calculation module 160 to autonomous vehicle 200. The following provides a detailed description of the configuration of autonomous vehicle 200. As shown in FIG 1, autonomous vehicle 200 includes a perception / judgment / control module 210, a vehicle control module 220, a position / attitude generation module 230, a position monitoring module 240, and a position / attitude receiving module 250.
[0016] The position / attitude generation module 230 measures the actual position / attitude of the autonomous vehicle 200 using sensors such as a Global Navigation Satellite System (GNSS) sensor, an Inertial Measurement Unit (IMU), an Inertial Navigation System (INS), etc., or estimates it through a recognition algorithm. The actual position / attitude generation module 230 generates the actual position / attitude information by measurement / estimation, which is transmitted to the position / attitude calculation module 160 via the interlocking module 170 of the autonomous driving simulator 100, and converted into virtual position / attitude information of the virtual autonomous vehicle in the virtual road environment. The position / attitude receiving module 250 receives the virtual position / attitude information calculated by the position / attitude calculation module 160 of the autonomous driving simulator 100 and transmits it to the perception / judgment / control module 210. The recognition / judgment / control module 210 recognizes / judges the vehicle state and road conditions using an autonomous driving algorithm based on the virtual sensor data received via the interlocking module 170 and the virtual position / posture information transmitted via the position / posture receiving module 250, and generates control commands for the autonomous vehicle 200.
[0017] Vehicle control module 220 controls the actual driving of autonomous vehicle 200 according to the control commands generated by autonomous vehicle 200 . The position monitoring module 240 detects whether the actual position of the autonomous vehicle 200, as measured / estimated by the position / attitude generation module 230, leaves the drivable area. To do so, the position monitoring module 240 maintains / utilizes an electronic map. Vehicle travel control by perception / judgment / control module 210 is performed based on a virtual position in the virtual road environment. Thus, in the virtual road environment, a virtual autonomous vehicle may be able to travel at a position, but the corresponding actual position in the actual travel location may be a position where travel is impossible. In such a case, position monitoring module 240 stops autonomous vehicle 200 via vehicle control module 220. Furthermore, in this case, position initialization, which will be described later, is required. The following describes in detail how position / attitude calculation module 160 converts the actual position / attitude of autonomous vehicle 200 into a virtual position / attitude of the virtual autonomous vehicle in the virtual road environment.
[0018] First, position / attitude calculation module 160 initializes the position and orientation of autonomous vehicle 200 and the virtual position and orientation of the virtual autonomous vehicle. When there is an initialization command, such as via an initialization button, the position and orientation are initialized and conversion begins. Initialization sets the current position and orientation of autonomous vehicle 200 to the initial actual position and orientation at the time of initialization, and sets the current position and orientation of the virtual autonomous vehicle to the initial virtual position and orientation. When position monitoring module 240 causes autonomous vehicle 200 to leave a drivable location and initialization is performed, autonomous vehicle 200 moves to a drivable location and then initialization is performed. In this case, the moved position and direction of autonomous vehicle 200 are set as the initial actual position and direction, and the final position and direction of the virtual autonomous vehicle are set as the initial virtual position and direction. Because the virtual autonomous vehicle is not moved, the final position and direction of the virtual autonomous vehicle become the initial virtual position and direction. On the other hand, the direction to be initialized means the yaw of the vehicle's attitude (roll, pitch, yaw), which is the rotation angle based on the z-axis perpendicular to the road plane. Thereafter, position / attitude calculation module 160 refers to information related to road shape among the information related to the road environment stored in road environment DB 110, and calculates an initial virtual attitude based on the initial virtual position and direction of the virtual autonomous vehicle. Because there is no relationship between the shape of the actual road on which autonomous vehicle 200 runs and the shape of the virtual road on which the virtual autonomous vehicle runs, it is not possible to directly convert / calculate the virtual attitude from the actual attitude without referring to road environment DB 110. After initialization, when autonomous vehicle 200 starts traveling, position / attitude calculation module 160 calculates the virtual position and orientation of the virtual autonomous vehicle based on the changes in the actual position and orientation of autonomous vehicle 200.
[0019] Figure 2 shows the initial actual position ( TIFF0007596432000001.tif1424) and the actual position moved by the autonomous vehicle 200 during its travel ( TIFF0007596432000002.tif1420) and the initial virtual position ( TIFF0007596432000003.tif1527) and the virtual position moved by the autonomous vehicle 200 ( TIFF0007596432000004.tif1523). The position change of the autonomous vehicle 200 in the coordinate system {R} ( TIFF0007596432000005.tif1722) is as follows: TIFF0007596432000006.tif18114On the other hand, in the coordinate system {V}, the position change of the virtual autonomous vehicle ( TIFF0007596432000007.tif1623) shows the position change of the autonomous vehicle 200 ( TIFF0007596432000008.tif1722). TIFF0007596432000009.tif1658 And the virtual position ( TIFF0007596432000010.tif1523) may be calculated as follows: TIFF0007596432000011.tif45142
[0020] 2 shows the result of calculating the virtual direction of the virtual autonomous vehicle based on the amount of change in direction of autonomous vehicle 200. The virtual direction of the virtual autonomous vehicle is calculated by reflecting the amount of change in actual direction of autonomous vehicle 200 in the initial virtual direction of the virtual autonomous vehicle. Next, the position / attitude calculation module 160 refers to information on the road shape among the information on the road environment stored in the road environment DB 110, and calculates a virtual attitude based on the virtual position and direction of the virtual autonomous vehicle. Hereinafter, a method for synchronizing position / attitude information between the autonomous driving simulator 100 and the autonomous driving vehicle 200 in the VIL system shown in FIG. 1 will be described in detail with reference to FIGS. 3 is a diagram for explaining a method for synchronizing position / attitude information in the VIL mode. The VIL mode is a mode for verifying the autonomous driving function based on the VIL system, and the above content is based on the VIL mode.
[0021] As shown in the figure, first, the position / attitude generation module 230 of the autonomous vehicle 200 measures / estimates the actual position / attitude of the autonomous vehicle 200 (S210), and transmits the measured / estimated actual position / attitude to the position / attitude calculation module 160 via the linkage module 170 of the autonomous driving simulator 100 (S220). Then, position / attitude calculation module 160 converts the actual position / attitude of autonomous vehicle 200 received in step S220 into a virtual position / attitude of the virtual autonomous vehicle in the virtual road environment (S230). In the VIL mode, the actual position / attitude is converted into a virtual position / attitude, so the dynamic characteristics of autonomous vehicle 200 are not reflected. On the other hand, if the position / attitude needs to be initialized (S240-Y), the position / attitude calculation module 160 sets the actual position / attitude of the autonomous vehicle 200 received in step S220 as the initial position / attitude of the autonomous driving simulator 100, and sets the virtual position / attitude converted in step S230 as the initial position / attitude of the virtual autonomous vehicle (S250). The position / attitude must be initialized when the VIL mode is started, when the vehicle leaves a drivable area, etc. If the position / attitude does not need to be initialized (S240-N), step S250 is not executed.
[0022] Next, the position / attitude calculation module 160 transmits the virtual position / attitude information converted in step S230 to the position / attitude receiving module 250 of the autonomous vehicle 200 via the linking module 170 (S260). 4 is a diagram for explaining a method of synchronizing position / attitude information when not in the VIL mode. As shown in the figure, the perception / judgment / control module 210 of the autonomous vehicle 200 transmits a vehicle control command to the position / attitude calculation module 160 via the interlocking module 170 of the autonomous driving simulator 100 (S310). Then, the position / attitude calculation module 160 of the autonomous driving simulator 100 analyzes the dynamic characteristics of the autonomous driving vehicle based on the vehicle control command received in step S310, estimates the vehicle's behavior, and calculates the position / attitude (S320). Next, position / attitude calculation module 160 transmits the position / attitude information calculated in step S320 to position / attitude receiving module 250 of autonomous vehicle 200 via linking module 170 (S330).
[0023] So far, the method for verifying autonomous driving functions based on a VIL system has been described in detail with reference to a preferred embodiment. In an embodiment of the present invention, a method is presented that allows an autonomous vehicle to verify the VIL system-based autonomous driving performance at a different location without having to go directly to a real test road such as a PG that replicates a virtual road environment. This will enable rapid testing of autonomous driving functions based on the VIL system in a variety of virtual road environments without having to change the driving location, accelerating the development of autonomous driving technology and improving convenience. In particular, according to an embodiment of the present invention, the roads whose information is stored in the road environment DB 110 can be changed while the roads on which the autonomous vehicle drives remain unchanged, making it possible to easily and quickly verify the autonomous vehicle's performance on various virtual roads. In the above embodiment, the roads stored in the road environment DB 110 are assumed to be real roads, but it is also possible to store information about virtual roads in the road environment DB 110 and verify the autonomous driving performance.
[0024] Meanwhile, the technical idea of the present invention may also be applied to a computer readable recording medium incorporating a computer program for performing the functions of the apparatus and method according to the present embodiment. The technical idea of various embodiments of the present invention may be realized in the form of computer readable code recorded on a computer readable recording medium. The computer readable recording medium may be any data storage device that can be read by a computer and can store data. For example, the computer readable recording medium may be a ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical disk, hard disk drive, etc. The computer readable code or program stored on the computer readable recording medium may be transmitted via a network connected between computers.
[0025] Although the preferred embodiment of the present invention has been described in detail above with reference to the accompanying drawings, the present invention is not limited to the above embodiment. It is clear that a person having ordinary knowledge in the technical field to which the present invention pertains can conceive of various modified or altered examples within the scope of the technical intent described in the claims, and it is understood that these also naturally belong to the technical scope of the present invention.
Claims
1. a road environment DB in which information on the road environment of a first actually existing road is stored; a simulation engine that generates a virtual road environment for a first real road using information stored in the road environment DB, identifies a position and posture of a virtual autonomous vehicle in the virtual road environment, generates a virtual NPC (Non Player Character) object, and generates a simulation scenario for verifying autonomous driving performance; a rendering module for rendering a virtual road environment in response to a scenario generated by the simulation engine; an emulation module that emulates sensors attached to the virtual autonomous vehicle in the virtual road environment rendered by the rendering module to generate virtual sensor data; a computation module that converts an actual position and an attitude of the autonomous vehicle that actually travels on a second real road, but not on the first real road, into a virtual position and a virtual attitude of the virtual autonomous vehicle in the virtual road environment, and transmits the virtual position and the virtual attitude to a simulation engine; an interlocking module that transmits virtual sensor data generated by the emulation module to the autonomous vehicle, receives the actual position and attitude of the autonomous vehicle from the autonomous vehicle, and transmits the data to the calculation module; Including, The computation module is When it is detected that the actual position of the autonomous vehicle has left a location where the autonomous vehicle can travel, after the autonomous vehicle moves to a location where the autonomous vehicle can travel, a first initialization is performed with respect to the actual position and attitude of the autonomous vehicle and the virtual position and virtual attitude of the virtual autonomous vehicle; In the first initialization, the position and direction of the autonomous vehicle after it has moved are set to the initial position and direction of the autonomous vehicle, the final position and yaw direction of the virtual autonomous vehicle when the actual position of the autonomous vehicle moves away from a drivable location are set to the initial virtual position and virtual yaw direction of the virtual autonomous vehicle, and an initial virtual attitude of the virtual autonomous vehicle is calculated based on the virtual position, the virtual yaw direction, and information about the road shape stored in the road environment DB. An autonomous driving simulator characterized by:
2. The actual position and orientation of the autonomous vehicle actually traveling on the second real road are 2. The autonomous driving simulator according to claim 1, wherein a virtual position and a virtual attitude of the virtual autonomous vehicle traveling in the virtual road environment are different from those of the first real road.
3. The second real road where autonomous vehicles actually drive is 3. The autonomous driving simulator according to claim 2, wherein the first actual road, which is a copy of a road whose information is stored in the road environment DB, is a different road.
4. An autonomous driving simulator as described in claim 3, characterized in that, while a second real road on which an autonomous vehicle actually drives remains unchanged, a road whose information is stored in the road environment DB may be changed from a road that replicates a first real road to a road that replicates a third real road.
5. The roads whose information is stored in the road environment database are:
2. The autonomous driving simulator according to claim 1, comprising a virtual road and an actual road.
6. The actual road is 6. The autonomous driving simulator according to claim 5, wherein roads on which the autonomous vehicle can actually travel to verify its functions include PG (Proving Ground), general roads, and expressways.
7. The computation module is 2. The autonomous driving simulator according to claim 1, further comprising: a virtual autonomous vehicle in a virtual road environment; a virtual position and a virtual orientation of the virtual autonomous vehicle in the virtual road environment;
8. The computation module is 8. The autonomous driving simulator according to claim 7, further comprising: a virtual attitude of the virtual autonomous driving vehicle calculated based on the virtual position and the virtual direction by referring to information about road shapes stored in a road environment DB.
9. The computation module is The autonomous driving simulator according to claim 1, characterized in that, if an initialization command is received, a second initialization is performed regarding the actual position and attitude of the autonomous vehicle and the virtual position and virtual attitude of the virtual autonomous vehicle, and conversion is started.
10. An autonomous driving simulator generates a virtual road environment for a first real road, identifies a position and an attitude of a virtual autonomous vehicle in the virtual road environment, generates a virtual NPC (Non Player Character) object, and generates a simulation scenario for verifying autonomous driving performance; - rendering, by the autonomous driving simulator, a virtual road environment in response to the generated scenario; an autonomous driving simulator emulating sensors attached to the virtual autonomous vehicle in the rendered virtual road environment to generate virtual sensor data; A step in which an autonomous driving simulator transmits the generated virtual sensor data to an autonomous driving vehicle that actually drives on a second real road instead of the first real road, and receives an actual position and attitude of the autonomous driving vehicle from the autonomous driving vehicle; an autonomous driving simulator converting the received actual position and attitude of the autonomous vehicle into a virtual position and attitude of the virtual autonomous vehicle in the virtual road environment; Including, The conversion step is When it is detected that the actual position of the autonomous vehicle has left a location where the autonomous vehicle can travel, after the autonomous vehicle moves to a location where the autonomous vehicle can travel, a first initialization is performed with respect to the actual position and attitude of the autonomous vehicle and the virtual position and virtual attitude of the virtual autonomous vehicle; In the first initialization, the position and direction of the autonomous vehicle after it has moved are set to the initial position and direction of the autonomous vehicle, the final position and yaw direction of the virtual autonomous vehicle when the actual position of the autonomous vehicle moves away from a drivable location are set to the initial virtual position and virtual yaw direction of the virtual autonomous vehicle, and an initial virtual attitude of the virtual autonomous vehicle is calculated based on the virtual position, the virtual yaw direction, and information about the road shape stored in the road environment DB. A method for simulating autonomous driving, comprising:
11. A simulation engine that generates a virtual road environment for a first real road, identifies a position and an attitude of a virtual autonomous vehicle in the virtual road environment, and generates a simulation scenario for verifying autonomous driving performance; a computation module that converts an actual position and an attitude of the autonomous vehicle that actually travels on a second real road, but not on the first real road, into a virtual position and a virtual attitude of the virtual autonomous vehicle in the virtual road environment, and transmits the virtual position and the virtual attitude to a simulation engine; An interlocking module that receives the actual position and attitude of the autonomous vehicle from the autonomous vehicle and transmits it to the calculation module; Including, The computation module is When it is detected that the actual position of the autonomous vehicle has left a location where the autonomous vehicle can travel, after the autonomous vehicle moves to a location where the autonomous vehicle can travel, a first initialization is performed with respect to the actual position and attitude of the autonomous vehicle and the virtual position and virtual attitude of the virtual autonomous vehicle; In the first initialization, the position and direction of the autonomous vehicle after it has moved are set to the initial position and direction of the autonomous vehicle, the final position and yaw direction of the virtual autonomous vehicle when the actual position of the autonomous vehicle moves away from a drivable location are set to the initial virtual position and virtual yaw direction of the virtual autonomous vehicle, and an initial virtual attitude of the virtual autonomous vehicle is calculated based on the virtual position, the virtual yaw direction, and information about the road shape stored in the road environment DB. An autonomous driving simulator characterized by:
Citation Information
Patent Citations
Image generating system, program and method, and simulation system, program and method
JP2018060512A
Automatic operation program evaluation system, and automatic operation program evaluation method
JP2020123259A
Autonomous vehicle simulator using a network platform
JP2022506797A
Automatic driving simulator and map generation method for automatic driving simulator
WO2019065409A1
Vehicle position detection device, and parameter set creation device for vehicle position detection
WO2020226071A1
