第1页 / 共180页
第2页 / 共180页
第3页 / 共180页
第4页 / 共180页
第5页 / 共180页
第6页 / 共180页
第7页 / 共180页
第8页 / 共180页
自动驾驶系统测试文档.pdf
Chapter 1. Introduction and background
Project Background and Purpose
Federal Automated Vehicles Policy
Stakeholder Engagement
Chapter 2. Automated Driving System Features
Overview
Approach
Framework for Discussing ADS Features
Levels of Driving Automation
Design Specific Functionality
ADS Tactical and Operational Maneuvers
Identification of Concept aDS Features
Category 1, L3 Conditional Automated Traffic Jam Drive Feature
Category 2, L3 Conditional Automated Highway Drive Feature
Category 3, L4 Highly Automated Low-Speed Shuttle Feature
Category 4, L4 Highly Automated Valet Parking Feature
Category 5, L4 Highly Automated Emergency Takeover
Category 6, L4 Highly Automated Highway Drive Feature
Category 7, L4 Highly Automated Vehicle/Transportation Network Company (TNC) Feature
Summary of Generic ADS Features
Summary
Chapter 3. Operational Design Domain
Overview
Approach
Influences for Defining the ODD Framework
Automated Driving Systems 2.0 – A Vision for Safety
2016 SAE J3016
California Policy
PEGASUS Project
Others Referenced
Guiding Principles
Defining an ODD Taxonomy
ODD Category Descriptions
Physical Infrastructure
Roadway Types
Roadway Surfaces
Roadway Edges
Roadway Geometry
Operational Constraints
Speed Limit
Traffic Conditions
Objects
Signage
Roadway Users
Non-roadway User Obstacles/Objects
Environmental Conditions
Weather
Weather-induced Roadway Conditions
Particulate Matter
Illumination
Connectivity
Vehicles
Traffic Density Information
Remote Fleet Management System
Infrastructure Sensors and communications
Zones
Geo-fencing (Crosbie, 2017)
Traffic Management Zones
School/Construction Zones
Regions/States
Interference Zones
ODD Identification for ADS Features
Summary
Chapter 4. Object and Event Detection and Response Capabilities
Overview
approach
Findings
Baseline ODDs
L3 Conditional Automated Traffic Jam Drive Feature
L3 Conditional Automated Highway Drive Feature
L4 Highly Automated Vehicle/TNC Feature
Baseline OEDR Behaviors
L3 Conditional Automated Traffic Jam Drive Feature
L3 Conditional Automated Highway Drive Feature
L4 Highly Automated Vehicle/TNC Feature
L3 Conditional Automated Traffic Jam Drive Feature
L3 Conditional Automated Highway Drive Feature
L4 Highly Automated Vehicle/TNC Feature
Summary
Chapter 5. Preliminary Tests and Evaluation Methods
Overview
Approach
Findings
Testing Architecture
Modeling and Simulation
Closed-Track Testing
Open-Road Testing
Test Scenarios
Testing Challenges
International ADS Testing Programs
AdaptIVe
PEGASUS
Summary
Chapter 6. fail-operational and fail-safe mechanisms
Overview
Approach
Findings
Failure Modes and Effects
Sensing and Communication
Perception
Navigation and Control
Human-Machine Interface
Summary
ADS Behavior Mapping
Failure Mitigation Strategies
Fail-Safe Mechanisms
Fail-Operational Mechanisms
Summary
Chapter 7. summary and conclusions
Appendix A. Operational Design domain samples
L3 Conditional Traffic Jam Drive
L3 Conditional highway drive
L4 Highly automated TNC
Appendix B. Modeling and Simulation for SCENARIO TESTING
Parameter Characterization
Subsystem Testing
Fault Detection
Appendix C. sample test procedures
perform lane change/low-speed merge
ODD Characteristics
OEDR Characteristics
Failure Behaviors
Test Protocol
Vehicle Platforms
Vehicle Roles
Test Scenarios
Test Scenario Sample Visualizations
General Procedures
Ambient Conditions
Personnel
Test Data and Equipment
Test Facility
Scenario Test: PLC_Comp_15 – Straight Road, Complex, 15 mph
Scenario Description
Test Subject and Purpose
Initial Conditions
Test Velocities
Metrics
Disengagements
Separation Distances
Signal Status
Execution of Procedure
Trial Validity
Evaluation Metrics
Perform Vehicle Following
ODD Characteristics
OEDR Characteristics
Failure Behaviors
Test Protocol
Vehicle Platforms
Vehicle Roles
Test Scenarios
Test Scenario Sample Visualizations
General Procedures
Ambient Conditions
Personnel
Test Data and Equipment
Test Facility
Scenario Tests: VF_S_25_Slow – Straight Road, POV Slower than SV
Scenario Description
Test Subject and Purpose
Initial Conditions
Test Velocities
Metrics
Disengagements
Following Distance
Deceleration Rate
Execution of Procedure
Trial Validity
Evaluation Metrics
Move Out of Travel Lane/Park
ODD Characteristics
OEDR Characteristics
Failure Behaviors
Test Protocol
Vehicle Platforms
Vehicle Roles
Test Scenarios
Test Scenario Sample Visualizations
General Procedures
Ambient Conditions
Personnel
Test Data and Equipment
Test Facility
Scenario Tests: MOTL_Comp_15 – Straight Road, Complex, 15 mph
Scenario Description
Test Subject and Purpose
Initial Conditions
Test Velocities
Metrics
Disengagements
Separation Distances
Deceleration Rate
Execution of Procedure
Trial Validity
Evaluation Metrics
Detect and Respond to School Buses
ODD Characteristics
OEDR Characteristics
Failure Behaviors
Test Protocol
Vehicle Platforms
Vehicle Roles
Test Scenarios
Test Scenario Sample Visualizations
General Procedures
Ambient Conditions
Personnel
Test Data and Equipment
Test Facility
Scenario Tests: SB_OD_25_Straight – Opposing Direction in Adjacent Lanes, Straight Road
Scenario Description
Test Subject and Purpose
Initial Conditions
Test Velocities
Metrics
Disengagements
Separation Distance at Stop
Execution of Procedure
Trial Validity
Evaluation Metrics (Performance Metrics – Pass/Fail Criteria)
Detect and Respond to Encroaching Oncoming Vehicles
Test Protocol
Vehicle Platforms
Vehicle Roles
Test Scenarios
Test Scenario Sample Visualizations
General Procedures
Ambient Conditions
Personnel
Test Data and Equipment
Test Facility
Scenario Tests: EOV_S_45_40 – Straight Road, 45 mph, 40 mph Opposing Vehicle
Scenario Description
Test Subject and Purpose
Initial Conditions
Test Velocities
Metrics
Disengagements
Avoidance Distance
Deceleration Rate
Yaw Rate
Execution of Procedure
Trial Validity
Evaluation Metrics
Detect and Respond to Pedestrians
Test Protocol
Vehicle Platforms
Vehicle Roles
Other Definitions
Test Scenarios
Test Scenario Sample Visualizations
General Procedures
Ambient Conditions
Personnel
Test Data and Equipment
Test Facility
Scenario Tests: Ped_Crosswalk_Sign_S_25 – Crosswalk Markings and Signs, Straight, 25 mph
Scenario Description
Test Subject and Purpose
Initial Conditions
Test Velocities
Metrics
Disengagements
Separation Distance
Deceleration Rate
Execution of Procedure
Trial Validity
Evaluation Metrics
Appendix D. Behavior competency Comparison
References
September 2018
DOT HS 812 623
A Framework for
Automated Driving System
Testable Cases and
Scenarios
DISCLAIMER
This publication is distributed by the U.S. Department of Transportation, National
Highway Traffic Safety Administration, in the interest of information exchange.
The opinions, findings and conclusions expressed in this publication are those of
the authors and not necessarily those of the Department of Transportation or the
National Highway Traffic Safety Administration. The United States Government
assumes no liability for its contents or use thereof. If trade or manufacturers’ names
are mentioned, it is only because they are considered essential to the object of the
publication and should not be construed as an endorsement. The United States
Government does not endorse products or manufacturers.
Suggested APA Format Citation:
Staplin, L., Mastromatto, T., Lococo, K. H., Kenneth W. Gish, K. W., & Brooks, J. O. (2018,
September). The effects of medical conditions on driving performance (Report No. DOT
HS 812 623). Washington, DC: National Highway Traffic Safety Administration.
i
Technical Report Documentation Page
1. Report No.
DOT HS 812 623
4. Title and Subtitle
A Framework for Automated Driving System Testable Cases and
Scenarios
2. Government Accession No.
3. Recipient's Catalog No.
5. Report Date
6. Performing Organization Code
September 2018
7. Authors
Eric Thorn,** Shawn Kimmel,*** Michelle Chaka*
9. Performing Organization Name and Address
Virginia Tech Transportation Institute*
3500 Transportation Research Plaza (0536)
Blacksburg, VA 24061;
Southwest Research Institute**
6220 Culebra Rd.
San Antonio, TX 78238;
Booz Allen Hamilton***
20 M St. SE
Washington DC 20003
12. Sponsoring Agency Name and Address
National Highway Traffic Safety Administration
1200 New Jersey Avenue SE.
Washington, DC 20590
8. Performing Organization Report No.
10. Work Unit No. (TRAIS)
11. Contract or Grant No.
13. Type of Report and Period Covered
Final Report
14. Sponsoring Agency Code
15. Supplementary Notes
16. Abstract
This report describes a framework for establishing sample preliminary tests for Automated Driving Systems.
The focus is on light duty vehicles exhibiting higher levels of automation, where the system is required to
perform the full dynamic driving task, including lateral and longitudinal control, as well as object and event
detection and response.
17. Key Words
automated driving systems, fail-safe mechanisms, object
and event detection and response, tests, operational
design domain
19. Security Classif. (of this report)
-Unclassified
20. Security Classif. (of this page)
Unclassified
18. Distribution Statement
No restrictions. This document is available to
the public through the National Technical
Information Service, www.ntis.gov.
21. No. of Pages
180
22. Price
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
i
EXECUTIVE SUMMARY
Automated driving systems (ADS) are being developed to perform the primary functions of the
dynamic driving task (DDT). These technologies hold great promise to improve safety and
mobility for transportation. The goal of this research was to develop an example of a preliminary
test framework for ADS that are in development and may come to market in the near to mid
future. The following steps were conducted to support the development of the sample test
framework.
1. Identify concept ADS
2. Identify attributes that define the operational design domain (ODD)
3. Identify object and event detection and response (OEDR) capabilities
4. Identify and assess failure modes and failure mitigation strategies
Technologies of interest in this work included light-duty automated driving functions that fell
within Level 3 (L3) to Level 5 (L5) of the SAE1 levels of driving automation (SAE International,
2016). The functions were identified based on prototype vehicles and conceptual systems. A
literature review which included popular media, press releases, technical journals, and
conference proceedings was performed. This review identified potential concept ADS being
developed or proposed by original equipment manufacturers (OEMs), suppliers, technology
companies, and other organizations. The identified ADS were categorized into a set of generic
names. The terminology was modified to ADS features (as opposed to functions) to more closely
align with the standardization community’s language.
Twenty-four conceptual features were identified, and although a thorough search was conducted,
the list is not exhaustive. The identified features were grouped into seven generic categories.
• L4 Highly Automated Vehicle/Transportation Network Company (TNC)
• L4 Highly Automated Highway Drive
• L4 Highly Automated Low Speed Shuttle
• L4 Highly Automated Valet Parking
• L4 Highly Automated Emergency Takeover
• L3 Conditional Automated Highway Drive
• L3 Conditional Automated Traffic Jam Drive
The generic names were developed to align with terminology from the SAE levels of driving
automation (i.e., conditional driving automation [L3], high driving automation [L4], and full
driving automation [L5]). Three of these generic features were selected to further support the
development of an example of a testing framework for ADS (L3 Conditional Automated Traffic
Jam Drive, L3 Conditional Automated Highway Drive, and L4 Highly Automated
Vehicle/TNC).
1 In 2006 the Society of Automotive Engineers changed its name to SAE International. It’s standards are still called SAE standards.
ii
The ODD describes the specific operating domains in which the ADS is designed to function.
The ODD will likely vary for each ADS feature on a vehicle and specifies the condition in which
that feature is intended and able to operate with respect to roadway types, speed range, lighting
conditions, weather conditions, and other operational constraints. The ODD is specified by the
technology developer, and the ADS should be able to identify whether it is operating within or
outside of that ODD.
A literature review was performed for all the generic ADS features to identify the attributes that
define the ODD. The review included popular media, press releases, technical journals, videos,
and conference proceedings. An ODD taxonomy for this report was then defined. This taxonomy
is hierarchical and includes the following top-level categories.
• Physical Infrastructure
• Operational Constraints
• Objects
• Connectivity
• Environmental Conditions
• Zones
Some of the challenges associated with ODD elements include their variability (e.g., rain droplet
sizes can vary greatly: light rain; moderate rain; and heavy rain), as well as identifying or
defining their boundaries. The work performed to identify the ODD lays a foundational
framework from which the ODD can be further defined and delineated by the developer, and
from which industry standards for ODD definition can be established.
OEDR refers to the subtasks of the DDT that include monitoring the driving environment
(detecting, recognizing, and classifying objects and events and preparing to respond as needed)
and executing an appropriate response to such objects and events (i.e., as needed to complete the
DDT and/or DDT fallback; (SAE International, 2016).
A notional concept of operations (ConOps) was considered for the three selected ADS features.
This served as a basis to perform an evaluation of the normal driving scenarios each ADS feature
may encounter, including expected hazards (e.g., other vehicles, pedestrians) and
sporadic/fluctuating events (e.g., emergency vehicles, construction zones). Baseline ODDs were
identified for each of the features to frame this analysis. These baseline ODDs and scenario
analyses were used to identify important OEDR functional capabilities. This analysis, along with
the survey of ADS features, helped to identify two key sets of behaviors for the selected ADS
features.
• Tactical Maneuver Behaviors
• OEDR Behaviors
Tactical maneuver behaviors may be viewed as more control-related tasks (e.g., lane following,
turning). OEDR behaviors may be regarded as perception and decision-making related tasks
iii
(e.g., detecting and responding to pedestrians). This analysis generated a list of fundamental
objects that may be relevant to an ADS’s driving task, as well as important events, which can be
viewed as interactions with those objects. A list of potential responses the ADS could implement
was identified, and these responses were mapped to the objects and events.
To develop a preliminary testing framework, existing test methods and tools were identified and
evaluated to formulate an appropriate, comprehensive testing architecture. The evaluation
resulted in three main components of a testing architecture for ADS, as well as advantages and
disadvantages of each.
• Modeling and Simulation (M&S)
• Closed-Track Testing
• Open-Road Testing
A test scenario framework that fit flexibly within the test architecture was then identified and
developed. The framework can be viewed as a multidimensional test matrix, with the following
principal elements.
• Tactical Maneuver Behavior
• ODD Elements
• OEDR Behavior
• Failure Mode Behaviors
An ADS test scenario can be defined at a high level by these dimensions. Each of these
dimensions can be viewed as a checklist of sorts to identify the maneuvers, ODD, OEDR, and
failure mode behaviors that will outline the test setup and execution. Preliminary test procedures
for a sampling of defined scenarios were then developed and these included, among other things,
information on potential test personnel, test facilities, test execution, data collection,
performance metrics, and success criteria that are translated from collected data and results.
Key challenges related to testing and evaluating ADS were also identified. These challenges
were associated with the technology itself, as well as test setup and execution.
A high-level system failure mode and effects analysis (FMEA) was performed for a
representative ADS. This representative ADS is described by a functional architecture under
development by SAE International (Underwood, 2016). This notionally identified potential ADS
failure modes, as well as their potential causes and effects. These failure modes were then
mapped back to the selected ADS features. The FMEA focused on subsystems and processes
related to the ADS, and the identified failure modes could largely be attributed to lack of
information (e.g., resulting from a hardware failure) or poor/inadequate information (e.g.,
resulting from system latency). These potential failures could have significant impacts,
ultimately resulting in collisions that could damage the vehicle or harm its occupants or other
roadway users.
iv
Potential failure mitigation strategies, including both fail-operational (FO) and fail-safe (FS)
techniques, were then identified and analyzed. FS techniques are used when the ADS cannot
continue to function, and may include options such as the following.
• Transitioning control to fallback-ready user
• Safely stopping in lane
• Safely moving out of travel lane/park
FO techniques can be used to allow the ADS to function at a reduced capacity, potentially for a
brief period of time or with reduced capabilities, and may include options such as the following.
• Adaptive compensation – weighting data from a complementary component or
subsystem more heavily (e.g., weighting camera data more heavily if lidar fails, etc.)
• Degraded modes of operation:
o Reduced speed operation
o Reduced level of automation operation
o Reduced ODD operation
o Reduced maneuver behavior operation
o Reduced OEDR behavior operation
The appropriate failure mitigation strategy is highly dependent on the nature of the failure and
the initial conditions under which the failure occurs. As such, implementing a hierarchy of
techniques, which may include the list above, may be appropriate. ADS internal health-
monitoring capabilities, such as measurement and indication of sensor and localization
subsystem performance, were also identified as being important.
v
4D/RCS
ACC
ABS
ADS
AEB
ALC
ASILS
BSW
CBD
ConOps
CV
DARPA
DDT
DOD
DSRC
ECU
ESC
FCW
FHWA
FMEA
FMECA
FMVSS
FO
FS
FTA
GPS
HAV
HazOP
HIL
HMI
HOV
HWD
IMU
INS
ISO
LDW
LKA
LTAP/OD
M&S
MRC
MUTCD
GLOSSARY OF TERMS AND ACRONYMS
4-dimensional real-time control system
adaptive cruise control
antilock braking system
automated driving system
automatic emergency braking
automated lane centering
ISO 26262 Automotive Safety Integrity Levels
blind spot warning
central business districts
concept of operations
connected vehicle
Defense Advanced Research Projects Agency
dynamic driving task
Department of Defense
dedicated short-range communication
electronic control unit
electronic stability control
forward collision warning
Federal Highway Administration
failure mode and effects analysis
failure modes, effects, and criticality analysis
Federal Motor Vehicle Safety Standard
fail-operational
fail-safe
fault tree analysis
global positioning system
Highly Automated Vehicle
Hazard and operability analysis
hardware-in-the-loop
human-machine interface
high-occupancy vehicle
highway drive
inertial measurement unit
inertial navigation system
International Organization for Standardization
lane departure warning
lane keeping assist
left turn across path/opposite direction
modeling and simulation
minimal risk condition
Manual on Uniform Traffic Control Devices
vi
相关推荐
2023年江西萍乡中考道德与法治真题及答案.doc
2012年重庆南川中考生物真题及答案.doc
2013年江西师范大学地理学综合及文艺理论基础考研真题.doc
2020年四川甘孜小升初语文真题及答案I卷.doc
2020年注册岩土工程师专业基础考试真题及答案.doc
2023-2024学年福建省厦门市九年级上学期数学月考试题及答案.doc
2021-2022学年辽宁省沈阳市大东区九年级上学期语文期末试题及答案.doc
2022-2023学年北京东城区初三第一学期物理期末试卷及答案.doc
2018上半年江西教师资格初中地理学科知识与教学能力真题及答案.doc
2012年河北国家公务员申论考试真题及答案-省级.doc
2020-2021学年江苏省扬州市江都区邵樊片九年级上学期数学第一次质量检测试题及答案.doc
2022下半年黑龙江教师资格证中学综合素质真题及答案.doc
