Active Driver Assistance for Vehicle Lanekeeping

Similar documents
An Autonomous Lanekeeping System for Vehicle Path Tracking and Stability at the Limits of Handling

A Practical Solution to the String Stability Problem in Autonomous Vehicle Following

Environmental Envelope Control

University Of California, Berkeley Department of Mechanical Engineering. ME 131 Vehicle Dynamics & Control (4 units)

Vehicle Dynamics and Control

MOTOR VEHICLE HANDLING AND STABILITY PREDICTION

Preliminary Study on Quantitative Analysis of Steering System Using Hardware-in-the-Loop (HIL) Simulator

A dream? Dr. Jürgen Bredenbeck Tire Technology Expo, February 2012 Cologne

Steer-by-Wire for Vehicle State Estimation and Control

Keywords: driver support and platooning, yaw stability, closed loop performance

Handwheel Force Feedback for Lanekeeping Assistance: Combined Dynamics and Stability

EMERGING TRENDS IN AUTOMOTIVE ACTIVE-SAFETY APPLICATIONS

Estimation and Control of Vehicle Dynamics for Active Safety

A Presentation on. Human Computer Interaction (HMI) in autonomous vehicles for alerting driver during overtaking and lane changing

TSFS02 Vehicle Dynamics and Control. Computer Exercise 2: Lateral Dynamics

AEM 4321 / EE4231 Automatic Control Systems

Chapter01 - Control system types - Examples

Passenger Vehicle Steady-State Directional Stability Analysis Utilizing EDVSM and SIMON

Developing a Platoon-Wide Eco-Cooperative Adaptive Cruise Control (CACC) System

Research on Skid Control of Small Electric Vehicle (Effect of Velocity Prediction by Observer System)

Új technológiák a közlekedésbiztonság jövőjéért

MEMS Sensors for automotive safety. Marc OSAJDA, NXP Semiconductors

Influence of Kink Protection Systems on a Tram Passing Through Curve

The Influence of Roll and Weight Transfer on Vehicle Handling

FRONTAL OFF SET COLLISION

Modeling and Optimization of a Linear Electromagnetic Piston Pump

Fig. 1.1 Concept cars equipped with a steer-wire-system

Vehicle Steering Control with Human-in-the-Loop

a Challenge for Lift-Based, Rigid Wing AWE Systems

Keeping Cars on The Road!

The Heart of Passive Safety (Electronics) is Still Beating. Tom Borninski Autoliv Electronics America

1) The locomotives are distributed, but the power is not distributed independently.

Using Adams as master model for ECU system simulation

Development of Rattle Noise Analysis Technology for Column Type Electric Power Steering Systems

Review on Handling Characteristics of Road Vehicles

Steering Actuator for Autonomous Driving and Platooning *1

Integrated Control Strategy for Torque Vectoring and Electronic Stability Control for in wheel motor EV

Towards a maximally-robust self-balancing robotic bicycle without reaction-moment gyroscopes nor reaction wheels

Tech Tip: Trackside Tire Data

ANALYZING THE DYNAMICS OF HIGH SPEED RAIL

VEHICLE AUTOMATION. CHALLENGES AND POTENTIAL FOR FUTURE MOBILITY.

EECS 461 Final Project: Adaptive Cruise Control

A Novel Chassis Structure for Advanced EV Motion Control Using Caster Wheels with Disturbance Observer and Independent Driving Motors

Multi-body Dynamical Modeling and Co-simulation of Active front Steering Vehicle

Modeling and Optimization of a Linear Electromagnetic Piston Pump

MECA0492 : Vehicle dynamics

DEVELOPMENT ENVIRONMENT FOR HAPTIC FEEDBACK DEVICE ON MOBILE AGRICULTURAL EQUIPMENT

NADY BOULES Director, Electrical & Controls Integration Lab

MODELS FOR THE DYNAMIC ANALYSIS OF THE SUSPENSION SYSTEM OF THE VEHICLES REAR AXLE

Modeling and Simulation of Linear Two - DOF Vehicle Handling Stability

BASIC MECHATRONICS ENGINEERING

Transient Responses of Alternative Vehicle Configurations: A Theoretical and Experimental Study on the Effects of Atypical Moments of Inertia

Seeing Sound: A New Way To Reduce Exhaust System Noise

Analysis on Steering Gain and Vehicle Handling Performance with Variable Gear-ratio Steering System(VGS)

Parameter Design and Tuning Tool for Electric Power Steering System

Vehicle Dynamics and Drive Control for Adaptive Cruise Vehicles

AUTONOMOUS REVERSING OF HEAVY GOODS VEHICLES

Development of Integrated Vehicle Dynamics Control System S-AWC

SAE Mini BAJA: Suspension and Steering

Investigating two-wheeler balance using experimental bicycles and simulators

Development of a Multibody Systems Model for Investigation of the Effects of Hybrid Electric Vehicle Powertrains on Vehicle Dynamics.

ρ ref Virtual driver () s - s f s a T a s k Human driver Vehicle dynamics Steering system T s Self aligning torque

Automated Driving - Object Perception at 120 KPH Chris Mansley

Special edition paper

ABS. Prof. R.G. Longoria Spring v. 1. ME 379M/397 Vehicle System Dynamics and Control

MANY VEHICLE control systems, including stability

Linear Flexible Joint Cart Plus Single Inverted Pendulum (LFJC+SIP)

Full-Scale Experimental Study of Vehicle Lateral Control System

Friday, 27 June Realizing a small UAV for medical transport in developing countries Master thesis: Ferdinand Peters. Dr.One

Using HVE to Simulate a Nine Vehicle Accident Involving a Heavy Truck

CHAPTER 1. Introduction and Literature Review

Autonomous Ground Vehicle

Fuzzy based Adaptive Control of Antilock Braking System

Low cost active devices to estimate and prevent off-road vehicle from rollover

Introduction Projects Basic Design Perception Motion Planning Mission Planning Behaviour Conclusion. Autonomous Vehicles

Development Trend of Advanced Steering System

THE PENNSYLVANIA STATE UNIVERSITY DEPARTMENT OF ENGINEERING SCIENCE AND MECHANICS ACTIVE SUSPENSION FOR V2V AND V2I LEARNING OF ROAD COMFORT

Design Optimization of Active Trailer Differential Braking Systems for Car-Trailer Combinations

Headlight Test and Rating Protocol (Version I)

Assessment of driver fitness: An alcohol calibration study in a high-fidelity simulation 26 April 2013

Statcom Operation for Wind Power Generator with Improved Transient Stability

Vehicle Performance. Pierre Duysinx. Research Center in Sustainable Automotive Technologies of University of Liege Academic Year

Single Vehicle Loss of Control

DEVELOPMENT OF INFRASTRUCTURE-BASED AUTONOMOUS DRIVING SUPPORT SYSTEM USING DYNAMIC MAPS

Estimation of Vehicle Parameters using Kalman Filter: Review

Automobile Body, Chassis, Occupant and Pedestrian Safety, and Structures Track

Microsoft Robotics Studio

MARITIME AFTERNOON. Torben Ole Andersen. June 14, 2017 Aalborg University, Denmark

Identification of a driver s preview steering control behaviour using data from a driving simulator and a randomly curved road path

Reduction of Self Induced Vibration in Rotary Stirling Cycle Coolers

The potential impact of electric powertrains on vehicle dynamics, control systems and active safety

LIDAR. ABSTRACT This paper. complexity. of tire skid-pad tests. scans could view. (a) Figure 1. LIDAR

Formal Methods will not Prevent Self-Driving Cars from Having Accidents

Fault-tolerant Control System for EMB Equipped In-wheel Motor Vehicle

Reentry Demonstration Plan of Flare-type Membrane Aeroshell for Atmospheric Entry Vehicle using a Sounding Rocket

EMaSM. Analysis of system response

Fundamentals of Steering Systems ME5670

Functional Algorithm for Automated Pedestrian Collision Avoidance System

Deep Learning Will Make Truly Self-Driving Cars a Reality

Dynamic performance of flow control valve using different models of system identification

Transcription:

Active Driver Assistance for Vehicle Lanekeeping Eric J. Rossetter October 30, 2003 D D L ynamic esign aboratory

Motivation In 2001, 43% of all vehicle fatalities in the U.S. were caused by a collision with a fixed obstacle (NHTSA) This accounted for over 18,000 fatalities in 2001 Stanford University Active Driver Assistance for Vehicle Lanekeeping - 2 Dynamic Design Lab

Spectrum of Lanekeeping Assistance PASSIVE ACTIVE Stanford University Active Driver Assistance for Vehicle Lanekeeping - 3 Dynamic Design Lab

Spectrum of Lanekeeping Assistance PASSIVE ACTIVE Lane Departure Warning These systems warn the driver of an imminent lane departure (Suzuki and Jansson (2003), Feng et al. (2000), Lee et al. (1999)) Visual, Auditory, or Haptic Lanekeeping is still the drivers responsibility Stanford University Active Driver Assistance for Vehicle Lanekeeping - 4 Dynamic Design Lab

Spectrum of Lanekeeping Assistance PASSIVE ACTIVE Lane Departure Warning Fully Autonomous Completely autonomous systems remove the driver entirely from the lanekeeping task (Fenton and Mayhon (1991), Shladover et al. (1991)) Requires an extremely high level of robustness Stanford University Active Driver Assistance for Vehicle Lanekeeping - 5 Dynamic Design Lab

Spectrum of Lanekeeping Assistance PASSIVE ACTIVE Lane Departure Warning Fully Autonomous Active Assistance Active assistance combines driver and controller commands to aid in lanekeeping Stanford University Active Driver Assistance for Vehicle Lanekeeping - 6 Dynamic Design Lab

Design Challenge for Active Assistance Design objectives for a lanekeeping assistance system: Provide lanekeeping ability without driver inputs Must not hinder normal driving These are often competing objectives How do we design the lanekeeping controller to satisfy both these objectives? Stanford University Active Driver Assistance for Vehicle Lanekeeping - 7 Dynamic Design Lab

Talk Outline Enabling Technologies Spring Analogy for Lanekeeping Stability Conditions Spring attachment point Sensing location Sizing the Spring Experimental Results Conclusions Stanford University Active Driver Assistance for Vehicle Lanekeeping - 8 Dynamic Design Lab

Enabling Technologies The ability to determine lane position Vision Systems (Gehrig et al. (2002), Franke et al. (1997)) GPS combined with Precision Road Maps (Omae & Fujioka (1999), Farrell & Barth (2000), Alban (2002)) Control of steering, braking, and throttle Brake- and Throttle-by-Wire Steer-by-Wire Source: Delphi Automotive Stanford University Active Driver Assistance for Vehicle Lanekeeping - 9 Dynamic Design Lab

Steer-by-Wire Corvette 1997 Corvette was modified to a steer-by-wire format No mechanical connection between the steering wheel and front road wheels handwheel steering column intermediate shaft belt drive handwheel angle sensor handwheel feedback motor steering actuator universal joints power assist unit pinion angle sensor gear assembly pinion rack Conventional Steering System Steer-by-Wire System Stanford University Active Driver Assistance for Vehicle Lanekeeping - 10 Dynamic Design Lab

Global Positioning System The Global Positioning System (GPS) is a worldwide positioning system consisting of over 24 satellites that broadcast ranging information Stanford University Active Driver Assistance for Vehicle Lanekeeping - 11 Dynamic Design Lab

Position Using GPS Distance from each satellite is determined by the time of flight of the satellite signal ρ = c t Satellite 2 Satellite 1 Satellite 3 ρ 1 Stanford University Active Driver Assistance for Vehicle Lanekeeping - 12 Dynamic Design Lab

Position Using GPS Distance from each satellite is determined by the time of flight of the satellite signal ρ = c t Satellite 2 Satellite 1 ρ Satellite 3 2 ρ 1 Stanford University Active Driver Assistance for Vehicle Lanekeeping - 13 Dynamic Design Lab

Position Using GPS Distance from each satellite is determined by the time of flight of the satellite signal ρ = c t Satellite 2 Satellite 1 ρ Satellite 3 2 ρ 1 ρ 3 Stanford University Active Driver Assistance for Vehicle Lanekeeping - 14 Dynamic Design Lab

Differential GPS A base station and vehicle see the same atmospheric errors Centimeter level accuracy is obtained with differential corrections Ionosphere ~20,000 km Troposphere ~1 km Stanford University Active Driver Assistance for Vehicle Lanekeeping - 15 Dynamic Design Lab

Control Concept Make vehicle behave as though it is attached to the road with a mechanical spring Road Centerline Force = k e Spring does not have to be extremely stiff Control force smoothly assists the driver Energy interpretation is useful for analysis Lateral Error: e Stanford University Active Driver Assistance for Vehicle Lanekeeping - 16 Dynamic Design Lab

Realizing this Control Concept With by-wire technology the steering, braking, and throttle are controlled to create the equivalent spring force F yl (δ) F yr (δ) F x F xlf F xrf F x x cf F y M F y F xlr F xrr Physical System Equivalent Control Force Stanford University Active Driver Assistance for Vehicle Lanekeeping - 17 Dynamic Design Lab

Realizing this Control Concept With control of only the front steering the control force is constrained at the front axle F yl (δ) F yr (δ) F x F y F x a M F y Physical System Equivalent Control Force Stanford University Active Driver Assistance for Vehicle Lanekeeping - 18 Dynamic Design Lab

Vehicle Behavior How does the vehicle behave under this control concept? Perform a linear analysis of the system dynamics: Model the vehicle dynamics with the lanekeeping controller Linearize the vehicle dynamics Look at eigenvalues (tell us about the system behavior) Lanekeeping behavior is related to the spring attachment point Stanford University Active Driver Assistance for Vehicle Lanekeeping - 19 Dynamic Design Lab

Attachment Point Condition Intuitively, creating the effect of a spring attached near the front of the vehicle will rotate the vehicle in the direction of the applied force Desired Location Stanford University Active Driver Assistance for Vehicle Lanekeeping - 20 Dynamic Design Lab

Attachment Point Condition Intuitively, creating the effect of a spring attached near the front of the vehicle will rotate the vehicle in the direction of the applied force Desired Location Stanford University Active Driver Assistance for Vehicle Lanekeeping - 21 Dynamic Design Lab

Attachment Point Condition Intuitively, creating the effect of a spring attached near the front of the vehicle will rotate the vehicle in the direction of the applied force Desired Location Stanford University Active Driver Assistance for Vehicle Lanekeeping - 22 Dynamic Design Lab

Attachment Point Condition What if the attachment point moves too far back? Desired Location Stanford University Active Driver Assistance for Vehicle Lanekeeping - 23 Dynamic Design Lab

Attachment Point Condition What if the attachment point moves too far back? Desired Location Stanford University Active Driver Assistance for Vehicle Lanekeeping - 24 Dynamic Design Lab

Attachment Point Condition What if the attachment point moves too far back? Desired Location Vehicle moves away from desired location! Stanford University Active Driver Assistance for Vehicle Lanekeeping - 25 Dynamic Design Lab

Neutral Steer Point Critical point is called the neutral steer point Neutral steer point is the location on a vehicle where an external force creates no rotation (yaw rate) Neutral Steer Point UNSTABLE STABLE Location of the neutral steer point depends on vehicle properties Tires and weight distribution Stanford University Active Driver Assistance for Vehicle Lanekeeping - 26 Dynamic Design Lab

Neutral Steer Point This point is used in vehicle design to ensure appropriate responses to physical disturbances Side-wind or gravity force on banked roads What is the implication for the design of a lanekeeping control system? Must coordinate the steering, braking, and throttle commands to ensure that the resultant force is in front of the neutral steer point With only front steering the control force is always in front of the neutral steer point Stanford University Active Driver Assistance for Vehicle Lanekeeping - 27 Dynamic Design Lab

Simulation 2 Lateral Error of Vehicle Lateral Position (m) 1.5 1 0.5 Force behind neutral steer point Force in front of neutral steer point 0-0.5 0 1 2 3 4 5 6 7 8 9 10 Time (s) Stanford University Active Driver Assistance for Vehicle Lanekeeping - 28 Dynamic Design Lab

High Speed Oscillations Satisfying the neutral steer point condition is not sufficient for high speed lanekeeping 8 Lateral Error Also need to look at sensing location Drivers look forward while driving Lateral Position (m) 6 4 2 0-2 -4-6 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Time (s) Stanford University Active Driver Assistance for Vehicle Lanekeeping - 29 Dynamic Design Lab

Sensing Location Controller must use a certain amount of lookahead for high speed stability Base spring compression/extension on the lateral error in front of the vehicle Sensing at C.G. 8 6 No Lookahead 15 m Lookahead 4 Using Lookahead lateral position (m) 2 2 0 4 6 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 time (s) Stanford University Active Driver Assistance for Vehicle Lanekeeping - 30 Dynamic Design Lab

Sizing the Spring Control force must be large enough to keep the vehicle in the lane without driver commands Most important factor for sizing the control force (or spring) is the road curvature Spring force must provide the necessary centripetal force for the turn Stanford University Active Driver Assistance for Vehicle Lanekeeping - 31 Dynamic Design Lab

Sizing the Spring: Basic Concept Centripetal force needed to go around a turn of radius R at a velocity V is: F c = m V 2 R Recall that the spring force is: F s = ke Equate the two forces and solve for the spring stiffness k = m V 2 Re Road Edge R Stanford University Active Driver Assistance for Vehicle Lanekeeping - 32 Dynamic Design Lab

Sizing the Spring In reality it is a little more complicated Curvature is not constant Vehicle dynamics influence motion Transient effects Basic spring analogy and concept of energy storage is used to scale the gain of the controller (or spring) Lyapunov theory Stanford University Active Driver Assistance for Vehicle Lanekeeping - 33 Dynamic Design Lab

Lyapunov Approach Create an energy-like function of the states with the properties: L(x) > 0 L(x) < 0 x(0),x(0) x max Energy Bowl 10 8 Sublevel Sets 100 6 80 4 Energy 60 40 20 Velocity 2 2 0 Initial Condition Max Position 0-10 -5 0 Velocity 5 10 10 5 0 Position -5-10 4 6 8 10 10 5 0 5 10 Position Stanford University Active Driver Assistance for Vehicle Lanekeeping - 34 Dynamic Design Lab

Experimental System Test vehicle is a 1997 Corvette Modified to include steer-by-wire DC-DC Converters GPS Receivers Inertial Sensors Control Computer Motor Amplifier Stanford University Active Driver Assistance for Vehicle Lanekeeping - 35 Dynamic Design Lab

Video Stanford University Active Driver Assistance for Vehicle Lanekeeping - 36 Dynamic Design Lab

Experiment vs. Simulation 100 Map 2 Lateral Error Simulation Experimental 50 1.5 0 1 North (m) 50 100 150 200 Lateral Error (m) 0.5 0-0.5 250-1 300-1.5 350 200 100 0 100 200 300 East (m) -2 20 30 40 50 60 70 80 90 100 110 Time (s) Experimental results show one loop at 11m/s (25mph) without driver steering inputs Guarantee the lanekeeping ability of the system Stanford University Active Driver Assistance for Vehicle Lanekeeping - 37 Dynamic Design Lab

Lanekeeping Bounds 100 Map 2 Lateral Error Simulation Experimental 50 1.5 0 1 North (m) 50 100 150 200 Lateral Error (m) 0.5 0-0.5 Reachable Set 250-1 300-1.5 350 200 100 0 100 200 300 East (m) -2 20 30 40 50 60 70 80 90 100 110 Time (s) Experimental results show one loop at 11m/s (25mph) without driver steering inputs Guarantee the lanekeeping ability of the system Stanford University Active Driver Assistance for Vehicle Lanekeeping - 38 Dynamic Design Lab

Conclusions Lanekeeping system works remarkably well Two important conditions for a well-behaved lanekeeping system Neutral steer point condition Sensing location (lookahead) Quantitative guarantee on the lanekeeping performance Stanford University Active Driver Assistance for Vehicle Lanekeeping - 39 Dynamic Design Lab

The End Questions? Stanford University Active Driver Assistance for Vehicle Lanekeeping - 40 Dynamic Design Lab