A lane departure warning system in-the-loop simulation test and evaluation method

Abstract

The present invention provides a lane departure warning system in-loop simulation test and evaluation method, which belongs to the field of commercial vehicle intelligent driving simulation test, and solves the technical problem that the lane departure warning system hardware in-loop simulation test lacks methodological support. The present invention includes four aspects: scenario construction technology, test evaluation technology, test system process, and test evaluation process. Scenario construction technology includes regulatory basis scenario construction, regulatory extension scenario construction, experience scenario construction, and random derivative scenario construction; test evaluation technology includes deviation vehicle speed conversion technology, lane line indicator quantification technology, and test classification evaluation technology.

Description

Technical Field

[0001] The present invention relates to the technical field of advanced driver assistance simulation testing, and in particular to an in-the-loop simulation testing and evaluation method for a lane departure warning system. Background Art

[0002] The Advanced Driving Assistance System (ADAS) uses a variety of sensors installed in the car, such as millimeter-wave radar, lidar, monocular / binocular cameras, and satellite navigation, to sense the surrounding environment at any time during driving, collect data, identify, detect, and track static and dynamic objects, and combine navigation map data to perform systematic calculations and analysis, so as to allow the driver to be aware of possible dangers in advance, effectively increasing the comfort and safety of car driving.

[0003] In the field of advanced driver assistance simulation test technology, there is a lack of in-the-loop simulation testing and evaluation methods for lane departure warning systems that are commonly used in the industry. At present, domestic laws and regulations related to advanced driver assistance focus on the field of real vehicle testing and cannot directly guide virtual simulation testing. However, advanced driver assistance hardware in-the-loop simulation testing lacks a recognized method system in the industry and also lacks regulatory guidance in this regard. Taking the lane departure warning system as an example, the present invention innovatively proposes an in-the-loop simulation testing and evaluation method for a lane departure warning system. Summary of the invention

[0004] In view of the current situation in the field of advanced driver assistance simulation test technology that there is a lack of a common advanced driver assistance hardware in-the-loop simulation test method in the industry, the present invention takes the lane departure warning system as an example and proposes a lane departure warning system in-the-loop simulation test and evaluation method.

[0005] In this invention, a complete set of hardware-in-the-loop simulation test methods is proposed for the lane departure warning system, relying on the intelligent driving hardware-in-the-loop test bench to quickly and deeply evaluate the lane departure warning system. The invention includes four aspects: scenario construction technology, test evaluation technology, test system process, and test evaluation process.

[0006] Scenario construction technology includes regulatory-based scenario construction, regulatory-extended scenario construction, empirical scenario construction, and random-derived scenario construction; test evaluation technology includes deviation speed conversion technology, lane line indicator quantification technology, and test classification evaluation technology; the test process system is specifically manifested as follows: Figure 1 The test evaluation system is specifically manifested as follows: Figure 2 Content shown.

[0007] The construction of regulatory-based scenarios refers to the regulations guiding the field testing of lane departure warning systems such as GBT26773-2011, JTT 883 2014, ECE R130, ISO17361-2007, etc., and manually constructs virtual test scenarios in the scenario simulation software.

[0008] Regulatory extended scenario construction includes but is not limited to expanding road width, lane line width, curve curvature radius, lane line color, lane line type, weather and lighting, etc. for manually constructed virtual scenarios in scenario simulation software.

[0009] Experience scenarios are used to reproduce the problem scenarios encountered in field tests and road tests, as well as the problem scenarios recognized by the industry. The basis for the reproduction of experience scenarios is usually video scene slices, and the way to reproduce is to reconstruct the virtual test scene in the scene simulation software.

[0010] Random derivative scenarios are mainly for road scenarios imported into scenario simulation software in OpenDrive format. It is necessary to generalize 2000 coordinate points in advance. The y coordinate of each coordinate point is in the center of a lane, and the x coordinate falls on the straight road and is at least 200m away from the straight line of the curve. Match a fixed serial number (1-2000) to each coordinate point, and then randomly select 100 random numbers from small to large in these 2000 serial numbers. Divide these 100 random numbers into 25 groups from small to large, with 4 random numbers in each group. For each group of random numbers, randomly select a time from 60s to 300s as the stable running time of the group of random number use cases, randomly select a speed from 60km / h to 80km / h as the stable vehicle speed of the group of random number use cases, and randomly select an angle from -30° to 30° for each group of random numbers as the steering wheel angle of the group of random number use cases.

[0011] The 100 random numbers drawn are used as 100 test cases, which are divided into 25 groups of random number cases from small to large. The corresponding initial coordinates are matched for each group of random number cases, and the corresponding stable operation time, stable vehicle speed, and steering wheel angle are extracted. The virtual vehicle dynamics model (this vehicle) accelerates from a standstill to a stable vehicle speed and maintains a stable vehicle speed during the stable operation time. The curve following module is turned on during the whole process. After reaching the stable operation time, the curve following module is automatically turned off and the lane departure test is immediately performed with the extracted steering wheel angle.

[0012] The test evaluation technologies include deviation speed conversion technology, lane line indicator quantification technology, and test classification evaluation technology.

[0013] Deviation speed conversion technology is divided into straight road deviation speed conversion and curve deviation speed conversion.

[0014] Straight road deviation speed conversion refers to the process of reading the normal steering wheel angle (which should normally fluctuate around 0) of the virtual vehicle dynamics model (this vehicle) running in the straight road virtual scene of the scene simulation software on the intelligent driving hardware-in-the-loop simulation test bench, and then giving it a left steering wheel angle value, reading the deviation speed at this time, and using the reverse deduction method to try to find the steering wheel angles that meet the regulations for slow deviation (0.1m / s-0.3m / s) and fast deviation (0.6m / s-0.8m / s).

[0015] The speed conversion of deviation from a curve refers to the following module function in the driver model when the virtual vehicle dynamics model (the vehicle) runs in the virtual scene of the curve in the scene simulation software on the intelligent driving hardware-in-the-loop simulation test bench, and the steering wheel angle of the normal following curve is read from the host computer software (for multi-lane roads with the same radius, the steering wheel angle values ​​of the left and right curves are different). Then the speed conversion test of deviation from a straight road is carried out, and the legal slow deviation speed range is adjusted to 0-0.4m / s, and the legal fast deviation speed range is adjusted to 0.4-0.8m / s. When conducting the lane deviation test on a curve, the corresponding steering wheel angle can be obtained by adding and subtracting the "corresponding slow deviation and fast deviation steering wheel angles in the speed conversion of deviation from a straight road" just calculated on the basis of the normal following curve steering wheel angle. The speed conversion of deviation from a curve focuses on testing multi-lane left curves with a radius of 500m, multi-lane right curves with a radius of 500m, multi-lane left curves with a radius of 250m, and multi-lane right curves with a radius of 250m.

[0016] Lane line index quantification technology refers to the camera controller that can read the camera's private CAN signal, constantly reading the left and right lane line fitting coefficients C0 left and C0 right sent to the camera's private CAN, and reading the left and right lane line qualities Q left and Q right. Usually, if the lane line quality is less than 1, it can be considered that the lane line cannot be recognized; C0 left and C0 right represent the distance between the vehicle and the left and right lane lines, respectively, which need to be within a reasonable range.

[0017] Test classification evaluation techniques, such as flow charts Figure 2 As shown. When Qleft ≥ 1 and Qright ≥ 1, both the left and right lane lines should be identifiable; when |C0left| < 127 and |C0right| < 127, neither the left nor right lane lines are lost; when |C0left| + |C0right| > 2.5 and |C0left| + |C0right| < 4, the lane width is normal. When lane line related information can be read from the camera's private CAN message and the above three categories are normal, first determine whether it is a false alarm, and then further analyze it. The above test classification and evaluation technology can help to quickly locate problems and quickly analyze and solve them.

[0018] It can be seen from the above technical solutions that the present invention has the following advantages:

[0019] The present invention solves the technical problems of the in-loop simulation test and evaluation method of the lane departure warning system, and innovatively proposes a complete set of test method theory systems, including four aspects: scenario construction technology, test evaluation technology, test step flow, and test evaluation flow. It is an in-depth exploration of the in-loop simulation test and evaluation method of the lane departure warning system, and there is currently no similar patent disclosure. The present invention is of great significance to the establishment of intelligent driving simulation test methods and specifications in the industry, and to the development and testing of intelligent driving systems in the industry, and is an important breakthrough in the direction of simulation testing. The present invention solves the blindness of the current lane departure warning system hardware in-loop simulation test method and test purpose, and also solves the problem that the lane departure warning system real vehicle field test regulations have limited guiding role in the field of simulation testing, making the intelligent driving hardware in-loop simulation test more systematic, more accurate, and more practical, and can discover software defects more timely and comprehensively, shorten the development cycle, and save a lot of risk costs brought by real vehicle testing. In addition, the theoretical systematicity and method feasibility of the present invention also play an important role in promoting the industry progress of intelligent driving virtual simulation testing, and the beneficial effects of its implementation are obvious. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0021] Figure 1 It is a flow chart of the test step process of one embodiment of the present invention.

[0022] Figure 2 It is a flow chart of a test evaluation process of an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0024] Embodiment 1

[0025] like Figure 1As shown, an embodiment of the present invention provides a flowchart of a lane departure warning system hardware-in-the-loop simulation test system, and the flowchart details the process of using the method to guide the lane departure warning system hardware-in-the-loop simulation test steps.

[0026] The process is mainly divided into four parts: basic version test, extended version test, experience scenario test, and random derivative test. If any part of the test fails, the second frame diagram - test evaluation process will be entered. If all four parts of the above tests pass, "Use case passed" will be displayed, and the test report will be printed, and the process ends.

[0027] Embodiment 2

[0028] An embodiment of the present invention provides a test evaluation process flow chart, and the block diagram introduces in detail the process of using the method to guide the hardware-in-the-loop simulation test evaluation of the lane departure warning system.

[0029] This process is divided into four steps, using C0 left, C0 right, Q left, and Q right parameters to quantify lane line indicators and classify and judge use cases that fail the test.

[0030] When Qleft ≥ 1 and Qright ≥ 1, both the left and right lane lines should be identifiable; when |C0left| < 127 and |C0right| < 127, both the left and right lane lines are not lost; when |C0left| + |C0right| > 2.5 and |C0left| + |C0right| < 4, the lane width is normal. When lane line related information can be read from the camera's private CAN message and the above three categories are normal, first determine whether it is a false alarm, and then further analyze it. The above test classification evaluation technology can help to quickly locate the problem and quickly analyze and solve the problem.

[0031] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A lane departure warning system in-loop simulation test and evaluation method, characterized in that: The methods include: scenario construction technology, test evaluation technology, test step process and test evaluation process; Scenario construction techniques include regulatory-based scenario construction, regulatory-expanded scenario construction, empirical scenario construction, and random-derived scenario construction; The test evaluation technologies include deviation speed conversion technology, lane line index quantification technology and test classification evaluation technology; Randomly derived scenarios are imported into the road scenarios of scenario simulation software in OpenDrive format; Generalize 2000 coordinate points in advance. The y coordinate of each coordinate point is in the center of a lane, and the x coordinate falls on the straight road and is at least 200m away from the straight line of the curve. Match a fixed serial number to each coordinate point, and then randomly select 100 random numbers from small to large within the fixed serial number; Divide the 100 random numbers into 25 groups from small to large, with 4 random numbers in each group. For each group of random numbers, randomly select a time from 60s to 300s as the stable running time of the use case of this group of random numbers. For each group of random numbers, randomly select a speed from 60km / h to 80km / h as the stable vehicle speed of the use case of this group of random numbers. For each group of random numbers, randomly select an angle from -30° to 30° as the steering wheel angle of the use case of this group of random numbers. The 100 random numbers drawn are used as 100 test cases, which are equally divided into 25 groups of random number cases from small to large. The corresponding initial coordinates are matched for each group of random number cases, and the corresponding stable running time, stable vehicle speed, and steering wheel angle are extracted. The virtual vehicle dynamics model accelerates from a standstill to a stable speed and then maintains a constant speed within the stable running time. During this period, the curve following module is turned on throughout the process; after reaching the stable running time, the curve following module is automatically turned off and a lane departure test is performed with the extracted steering wheel angle.

2. The lane departure warning system in-loop simulation test and evaluation method according to claim 1, characterized in that: The construction of regulatory-based scenarios refers to the regulations GBT 26773-2011, JTT 883 2014, ECE R130 and ISO17361-2007 that guide the actual vehicle field testing of lane departure warning systems, and manually constructs virtual test scenarios in the scenario simulation software.

3. The lane departure warning system in-loop simulation test and evaluation method according to claim 1, characterized in that: Regulatory extended scenario construction includes, but is not limited to, extending road width, lane line width, curve curvature radius, lane line color, lane line type, and weather and lighting for manually constructed virtual scenarios in scenario simulation software.

4. The lane departure warning system in-loop simulation test and evaluation method according to claim 1, characterized in that: Empirical scenarios are used to reproduce problem scenarios encountered in field tests and road tests, as well as problem scenarios recognized by the industry. The basis for reproducing empirical scenarios is usually video scene slicing, and the method of reproduction is to reconstruct virtual test scenes in scene simulation software.

5. The lane departure warning system in-loop simulation test and evaluation method according to claim 1, characterized in that: Deviation speed conversion technology is divided into straight road deviation speed conversion and curve deviation speed conversion; The straight road deviation speed conversion is carried out on the intelligent driving hardware-in-the-loop simulation test bench. When the virtual vehicle dynamics model runs in the straight road virtual scene of the scene simulation software, the steering wheel angle of normal driving is read from the host computer software, and then a left steering wheel angle value is given to it. The deviation speed at this time is read, and the reverse method is used to try to find the steering wheel angle corresponding to the slow deviation and fast deviation that meet the regulations; The speed conversion of deviation from curve is realized on the intelligent driving hardware-in-the-loop simulation test bench. When the virtual vehicle dynamics model runs in the curve virtual scene of the scene simulation software, the curve following module function built into the driver model is turned on, and the steering wheel angle of the normal curve is read from the host computer software. Then conduct a straight road deviation speed conversion test, and adjust the legal slow deviation speed range to 0-0.4m / s, and the legal fast deviation speed range to 0.4-0.8m / s; When conducting the curve lane departure test, the slow and fast deviation steering wheel angles corresponding to the straight road deviation speed conversion just calculated are added and subtracted to the normal following-curve steering wheel angle, thus obtaining the corresponding steering wheel angle; the curve deviation speed conversion is tested on a multi-lane left curve with a radius of 500m, a multi-lane right curve with a radius of 500m, a multi-lane left curve with a radius of 250m, and a multi-lane right curve with a radius of 250m.

6. The lane departure warning system in-loop simulation test and evaluation method according to claim 1, characterized in that: The lane line indicator quantification technology is aimed at the camera controller that can read the camera's private CAN signal. It always reads the left and right lane line fitting coefficients C0 left and C0 right sent to the camera's private CAN, and reads the left and right lane line qualities Q left and Q right. If the lane line quality is less than 1, it is considered that the lane line cannot be recognized; C0 left and C0 right represent the distance between the vehicle and the left and right lane lines, respectively, which are within a reasonable range.

7. The lane departure warning system in-loop simulation test and evaluation method according to claim 1, characterized in that: When Qleft ≥ 1 and Qright ≥ 1, both left and right lane lines should be recognizable; when |C0left| < 127 and |C0right| < 127, both left and right lane lines are not lost; when |C0left| + |C0right| > 2.5 and |C0left| + |C0right| < 4, the lane width is normal; When lane line related information can be read from the camera's private CAN message and the above three categories are normal, first determine whether it is a false alarm and then analyze it.

Images