Driving performance verification device and driving performance verification method of autonomous vehicle

Abstract

To provide a driving performance verification device and driving performance verification method of an autonomous vehicle.SOLUTION: The driving performance verification device according to the invention includes: a reception unit 110 for receiving sensing data provided from each sensor mounted on an autonomous vehicle; a branch unit 120 for branching the sensing data into first sensing data and second sensing data, which are identical to each other, and providing the branched data; a sensing data regeneration unit 130 for storing in real time the second sensing data provided from the branching unit in a driving mode, and regenerating the stored second sensing data in a verification mode; and a verification control unit 140 for providing the first sensing data to an electronic control unit (ECU) of the autonomous vehicle and obtaining first driving data in the driving mode, providing the generated second sensing data to the ECU and obtaining second driving data in the verification mode, and verifying driving performance depending on whether or not the first driving data and the second driving data are identical to each other in comparison therebetween.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to an apparatus and method for verifying driving performance of an autonomous vehicle, and more particularly to an apparatus and method for verifying driving performance of an autonomous vehicle, capable of detecting driving performance by performing an actual vehicle simulation based on sensing data sensed on an actual road. [Background technology]

[0002] An autonomous vehicle is a vehicle that moves to a destination by controlling its own driving, rather than being operated by a driver. In order to operate an autonomous vehicle, various types of sensors are required to monitor the surrounding environment on behalf of the driver, and the autonomous vehicle accurately recognizes objects such as humans, animals, other vehicles, and traffic signals (hereinafter referred to as "object information") using the sensing data obtained by the sensors, and performs driving control accordingly after recognition.

[0003] Recently, much research has been conducted into the control of autonomous driving, as autonomous vehicles have the advantage of being able to prevent accidents caused by careless driver errors and aggressive driving, as well as allowing unlicensed people, blind people, and even minors to use the vehicles freely. However, the road environment in which real vehicles drive has many unpredictable aspects and many possible variables, so in order to provide safe and accurate autonomous driving, it is necessary to verify the driving performance of autonomous vehicles.

[0004] Conventionally, verification of driving performance has been carried out using a simulator with various accident environments set up. However, because this is carried out using the accident environments generated within such simulators and the corresponding algorithms, there is a problem in that verification using actual vehicle inspections or road data sensed on actual roads is not carried out. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Korean Patent Publication No. 10-2022-0097646 (July 8, 2022) Summary of the Invention [Problem to be solved by the invention]

[0006] Therefore, an object of the present invention is to provide an apparatus and method for verifying running performance of an autonomous vehicle, which can overcome the above-mentioned problems of the conventional technology. Another object of the present invention is to provide an autonomous vehicle driving performance verification device and a driving performance verification method that can perform an actual vehicle simulation based on sensing data sensed on actual roads and accurately and effectively verify driving performance. [Means for solving the problem]

[0007] According to an embodiment of the present invention for achieving some of the above technical objectives, the driving performance verification device for an autonomous vehicle according to the present invention includes a receiving unit that operates in a driving mode and receives in real time sensing data sensed and provided in real time from each sensor attached to the autonomous vehicle while the autonomous vehicle is driving on an actual road; a branching unit that operates in the driving mode and branches the sensing data into identical first sensing data and second sensing data and provides them; a sensing data reproducing unit that stores the second sensing data provided from the branching unit in real time in the driving mode and reproduces the stored second sensing data in a verification mode; and a sensing data reproducing unit that reproduces the first sensing data provided in real time in the driving mode performed on an actual road by transmitting the first sensing data provided in real time to an electronic control unit (ECU) of the autonomous vehicle. and a verification control unit that controls the second sensing data to be provided to an electronic control unit (ECU) in a verification mode performed on a separately set test road, controls the second sensing data to be played back via the playback unit to be provided to the electronic control unit (ECU), causes the electronic control unit (ECU) to perform autonomous driving on the test road in response to the second sensing data to obtain second driving data, and compares the first driving data recording the driving conditions in the driving mode with the second driving data recording the driving conditions in the verification mode, and verifies driving performance based on whether they match.

[0008] In the verification mode, the receiver and the branching units may not operate.

[0009] The sensors may include at least one of a camera, a lidar, a radar, an ultrasonic sensor, a steering sensor, a temperature / humidity sensor, and a GPS.

[0010] The electronic control unit (ECU) is equipped with a trained AI autonomous driving algorithm and can perform autonomous driving in response to the first sensing data or the second sensing data.

[0011] According to another embodiment of the present invention for achieving some of the above technical objectives, a method for verifying driving performance of an autonomous vehicle according to the present invention includes a first step of branching sensing data received in real time from each sensor attached to the autonomous vehicle while the autonomous vehicle is traveling on the real road, into first sensing data and second sensing data, the first sensing data being identical to each other; a second step of causing the electronic control unit (ECU) to perform autonomous driving on an actual road in response to the first sensing data, thereby acquiring first driving data; a fourth step of starting a verification mode performed on a separately set test road, reproducing the second sensing data, thereby causing the second sensing data to be provided to the electronic control unit (ECU); a fifth step of causing the electronic control unit (ECU) to perform autonomous driving on the test road in response to the second sensing data, thereby acquiring second driving data; and a sixth step of comparing the first driving data recording the driving conditions in the driving mode with the second driving data recording the driving conditions in the verification mode, and verifying driving performance based on whether or not they match.

[0012] The sensors may include at least one of a camera, a lidar, a radar, an ultrasonic sensor, a steering sensor, a temperature / humidity sensor, and a GPS.

[0013] Each of the first driving data and the second driving data may include information regarding the operating state of each part required for driving of the autonomous vehicle. Effect of the Invention

[0014] According to the present invention, the driving performance of an autonomous vehicle can be verified accurately and effectively by performing an actual vehicle simulation based on sensing data sensed on an actual road. [Brief description of the drawings]

[0015] [Figure 1] 1 is a schematic diagram of an autonomous vehicle driving performance verification device according to the present invention; [Diagram 2] 2 is a block diagram showing the internal configuration of the running performance verification device shown in FIG. 1. [Diagram 3] FIG. 4 is an operational sequence diagram of a driving mode. [Figure 4] FIG. 13 is an operational flowchart of a verification mode. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings, with the sole intention of providing a thorough understanding of the present invention to those skilled in the art to which the present invention pertains.

[0017] FIG. 1 is a schematic diagram of an autonomous vehicle running performance verification device according to the present invention, and FIG. 2 is a block diagram of the internal configuration of the running performance verification device of FIG.

[0018] As shown in Figures 1 and 2, the autonomous vehicle driving performance verification device 100 according to the present invention is for verifying the performance of the driving control of the autonomous vehicle, and is installed between a sensor 10 attached to the autonomous vehicle and an electronic control unit (ECU) 200, and includes a receiving unit 110, a branching unit 120, a reproducing unit 130, and a verification control unit 140.

[0019] The autonomous vehicle driving performance verification device 100 according to the present invention is for verifying driving performance based on sensing data acquired while driving on an actual road, and performs verification by dividing into a driving mode and a verification mode. In the main driving mode, sensing data is acquired while driving on an actual road, and in the verification mode, a test drive is performed based on the sensing data acquired in the driving mode, and the driving data in the driving mode and the driving data in the verification mode are compared to verify the driving performance of the autonomous vehicle. Here, the receiving unit 110 and the branching unit 120 operate only in the driving mode and do not operate in the verification mode, and the reproducing unit 130 and the verification control unit 140 operate in both the driving mode and the verification mode.

[0020] Here, the sensor 10 may include at least one of a camera, a lidar, a radar, an ultrasonic sensor, a steering sensor, a temperature / humidity sensor, and a GPS, and may further include various other types of sensors that are well known to ordinary technicians as being necessary for autonomous driving.

[0021] The electronic control unit (ECU) 200 is built into an autonomous vehicle to control the overall operation of the autonomous vehicle, and is configured to incorporate a learned AI autonomous driving algorithm and perform autonomous driving in response to sensing data sensed by the sensor 10. The configuration and operation of the electronic control unit (ECU) are well known to ordinary engineers.

[0022] The receiver 110 operates in a driving mode and receives sensing data provided in real time from each sensor 10 attached to the autonomous vehicle while driving on a real road, and provides the sensing data to the branch unit 120.

[0023] The branching unit 120 branches the sensing data received in real time through the receiving unit 110 into the same first sensing data and second sensing data. The branching unit 120 may include a module for branching data.

[0024] The branching section 120 may be provided independently according to the number of sensors, i.e., a branching section for a camera, a branching section for a lidar, a branching section for a radar, etc. may be provided for each sensor.

[0025] Here, the first sensing data is provided to the electronic control unit (ECU) under the control of the verification control unit 140, and the second sensing data is provided to the reproducing unit .

[0026] When the first sensing data is transmitted to the electronic control unit (ECU), the electronic control unit (ECU) analyzes the first sensing data using a learned AI autonomous driving algorithm to recognize and determine object information such as obstacles, signals, and lanes, and performs autonomous driving on an actual road. In this case, while the autonomous vehicle is driving on an actual road, it performs autonomous driving on the actual road in response to the first sensing data sensed in real time from each sensor 10.

[0027] The reproducing unit 130 stores the second sensing data provided in real time from the branch unit 120 in a driving mode, and reproduces the stored second sensing data and provides it to the electronic control unit (ECU) in a verification mode. In the verification mode, sensing data sensed in real time through the sensor 10 is not provided to the electronic control unit (ECU), and the second sensing data reproduced by the reproducing unit 130 is provided to the electronic control unit (ECU), and the electronic control unit (ECU) analyzes the second sensing data using a learned AI autonomous driving algorithm to determine obstacles, signals, lanes, etc., and performs autonomous driving.

[0028] The verification control unit 140 controls the operation of a driving mode in which autonomous driving is performed on an actual road using the first sensing data sensed in real time through the sensor 10, and a verification mode in which autonomous driving is performed on a test road using the second sensing data played back by the playback unit 130.

[0029] Specifically, in the driving mode, the verification control unit 140 provides the first sensing data provided in real time to the electronic control unit (ECU) via the branch unit 120, and controls the driving of the autonomous vehicle on an actual road in response to the first sensing data from the electronic control unit (ECU). Also, in the verification mode, the driving is performed on a separately set test road, and the second sensing data reproduced and provided by the reproduction unit 130 is provided to the electronic control unit (ECU), and controls the test driving of the autonomous vehicle on the test road in response to the second sensing data from the electronic control unit (ECU).

[0030] Thereafter, the first driving data recording the driving conditions in the driving mode provided by the electronic control unit (ECU) is compared with the second driving data recording the driving conditions in the verification mode, and the driving performance of the autonomous vehicle is verified based on whether or not they match.

[0031] In the past, in order to verify the driving performance of an autonomous vehicle, a virtual scenario was created, virtual signals were injected, and corresponding driving operation data was collected to verify vehicle control. However, in this case, it was difficult to simulate conditions due to weather changes such as snow, rain, and fog, and since the data was not related to the actual environment, it was difficult to evaluate the driving control operation on the actual road, and there were problems in that it was costly and time-consuming to create a virtual scenario.

[0032] In the present invention, driving performance is evaluated using sensing data obtained while driving on actual roads, and since sensing data obtained directly from the autonomous vehicle is used, there is an advantage that a more accurate evaluation is possible.

[0033] The first driving data and the second driving data are generated and stored in the electronic control unit (ECU), respectively, and provided to the verification control unit 140.

[0034] Each of the first driving data and the second driving data may include information on the operating state of each part required for driving of the autonomous vehicle, including the speed information, steering information, pedal information, etc. of the autonomous vehicle. In particular, the first driving data and the second driving data may include information on the driving speed, the brake operating state, the accelerator pedal operating state, the wheel operation amount, etc.

[0035] The verification control unit 140 can verify the driving performance of the autonomous vehicle by comparing the first driving data with the second driving data, and can compare differences in mechanical operation differences and steering angles in the same sensor environment by comparing the movements of the autonomous vehicle. In addition, it is possible to compare differences in response speed to the same sensing data, steering, acceleration, deceleration, and the timing and extent to which driving devices such as steering, acceleration, and deceleration are operated. This makes it possible to determine whether there is a problem with the recognition of sensing data by the autonomous vehicle, a problem with the software, or a problem with the hardware (mechanical device).

[0036] Hereinafter, a method for verifying driving performance using the autonomous vehicle driving performance verification device 100 will be described with reference to Figures 3 and 4. Figure 3 is an operation sequence diagram of the driving mode, and Figure 4 is an operation sequence diagram of the verification mode.

[0037] Prior to the start of the driving mode and the verification mode, the above-mentioned driving performance verification device 100 is mounted on the autonomous vehicle. The driving performance verification device 100 is mounted between a sensor 10 and an electronic control unit (ECU) mounted on the autonomous vehicle, and configured so that sensing data sensed by the sensor 10 is not directly transmitted to the electronic control unit (ECU) but is transmitted to the electronic control unit (ECU) via the driving performance verification device 100, specifically, via the receiving unit 110 and the branching unit 120.

[0038] As shown in FIG. 3, when a driving mode performed on an actual road is started (S110), the autonomous vehicle to be verified drives on an actual road and receives sensing data sensed in real time from each sensor attached to the autonomous vehicle, which is received by the receiver 110, and the branching unit 120 branches the data into the first sensing data and the second sensing data, which are identical to each other (S120).

[0039] Thereafter, the branch unit 120 provides the first sensing data to an electronic control unit (ECU) of the autonomous vehicle (S130), and the electronic control unit (ECU) controls the autonomous vehicle to travel on the actual road in response to the first sensing data (S140). At this time, the second sensing data is stored in the playback unit 130.

[0040] The first driving data recorded during the autonomous driving in the driving mode is transmitted to the verification control unit 140, where it is acquired and stored (S150).

[0041] Thereafter, the verification mode starts (S210) as shown in Fig. 4. The verification mode is carried out not on an actual road but on a separately set test road.

[0042] In the verification mode, the second sensing data stored in the playback unit 130 during the driving mode is provided to the electronic control unit ECU, instead of the sensing data sensed in real time by the sensor 10 attached to the autonomous vehicle (S220).

[0043] Thereafter, the electronic control unit (ECU) controls autonomous driving of the autonomous vehicle on the test road in response to the second sensing data (S230).

[0044] At this time, the second driving data recorded during the autonomous driving in the verification mode is transmitted to the verification control unit 140, where it is acquired and stored (S240).

[0045] Thereafter, the first driving data recording the driving conditions in the driving mode and the second driving data recording the driving conditions in the verification mode are compared, and the driving performance of the autonomous vehicle is verified based on whether they match (S250).

[0046] The driving performance of the autonomous vehicle can be verified by comparing the movements of the autonomous vehicle and comparing the differences in mechanical operation and steering angle in the same sensor environment. Also, it is possible to compare the differences in response speed to the same sensing data, steering, acceleration, deceleration, and the timing and extent to which driving devices such as steering, acceleration, and braking are operated. This makes it possible to determine whether there is a problem with the autonomous vehicle's recognition of sensing data, a problem with the software, or a problem with the hardware (mechanical device).

[0047] As described above, according to the present invention, the driving performance of an autonomous vehicle can be verified accurately and effectively by performing an actual vehicle simulation based on sensing data sensed on an actual road.

[0048] The above description of the embodiments is merely an example with reference to the drawings to allow a more thorough understanding of the present invention, and is not intended to limit the present invention. In addition, it is obvious to those skilled in the art that various changes and modifications can be made without departing from the basic principles of the present invention. [Explanation of symbols]

[0049] 110 Receiving unit 120 Branch 130 Playback Department 140 Verification control section

Claims

1. A receiving unit that operates in a driving mode and receives sensing data provided in real time by each sensor attached to the autonomous vehicle while driving on an actual road; a branching unit that operates in a driving mode and branches the sensing data into first sensing data and second sensing data that are identical to each other and provides the first sensing data and the second sensing data; a sensing data reproducing unit that stores the second sensing data provided by the branch unit in real time in a driving mode and reproduces the stored second sensing data in a verification mode; a verification control unit that, in a driving mode performed on an actual road, causes the first sensing data provided in real time to be provided to an electronic control unit (ECU) of the autonomous driving vehicle, and acquires first driving data by performing autonomous driving on the actual road in response to the first sensing data, and, in a verification mode performed on a separately set test road, controls the second sensing data reproduced via the reproduction unit to be provided to the electronic control unit (ECU), and acquires second driving data by performing autonomous driving on the test road in response to the second sensing data, and compares the first driving data recording the driving conditions in the driving mode with the second driving data recording the driving conditions in the verification mode, and verifies driving performance based on whether they match.

2. An autonomous vehicle driving performance verification device comprising:

2. In the verification mode, the receiver and the branching units are not operational. The autonomous vehicle driving performance verification device according to claim 1 .

3. The sensor includes at least one of a camera, a lidar, a radar, an ultrasonic sensor, a steering sensor, a temperature / humidity sensor, and a GPS. The autonomous vehicle driving performance verification device according to claim 1 .

4. The electronic control unit (ECU) has a learned AI autonomous driving algorithm built in and performs autonomous driving in response to the first sensing data or the second sensing data. The autonomous vehicle driving performance verification device according to claim 1 .

5. A method for verifying driving performance of an autonomous vehicle, comprising: A first step of starting a driving mode performed on a real road and branching sensing data sensed by each sensor attached to the autonomous vehicle and received in real time while driving on the real road into first sensing data and second sensing data that are identical to each other; a second step of providing the first sensing data to an electronic control unit (ECU) of the autonomous vehicle and storing the second sensing data; a third step of acquiring first driving data by having the electronic control unit (ECU) perform autonomous driving on an actual road in response to the first sensing data; A fourth step of starting a verification mode performed on a separately set test road and reproducing the second sensing data so that the second sensing data is provided to the electronic control unit (ECU); a fifth step of acquiring second driving data by having the electronic control unit (ECU) perform autonomous driving on the test road in response to the second sensing data; and and a sixth step of comparing first driving data in which a driving condition in the driving mode is recorded with second driving data in which a driving condition in the verification mode is recorded, and verifying driving performance based on whether or not they match. A method for verifying the driving performance of an autonomous vehicle, comprising:

6. The sensor includes at least one of a camera, a lidar, a radar, an ultrasonic sensor, a steering sensor, a temperature / humidity sensor, and a GPS. The method for verifying driving performance of an autonomous vehicle according to claim 5.

7. Each of the first driving data and the second driving data includes information on an operating state of each part required for driving of the autonomous vehicle. The method for verifying driving performance of an autonomous vehicle according to claim 5.

Images