Automated Vehicle Longitudinal Control Development: Passenger and Heavy-Duty Vehicles

1
Automated Vehicle Longitudinal Control
Development Passenger and Heavy-Duty Vehicles

• X. Y. Lu
• J. K. Hedrick
• Program Manager D. Empey
• (Demo-2003)
• PATH, U. C. Berkeley

2
Part 1. Passenger Cars

• Merging Simulation for High Speed Implementation
• Transition Control - From Manual to Automatic
• Real-time Tire Slip Ratio Estimation

3
Merging Simulation for High Speed Implementation

• General merging maneuver
• Minimizing acceleration demanding of merging
  vehicle
• Dynamic real-time simulation PATH home page
• http//www.path.berkeley.edu/PA
  TH/General/WhatsNew/

4
Transition control - From Manual to Automatic

• Longitudinal Transition Control
• For operation of single vehicle
• Multiple transitions between manual and automatic
• To be able to automatically stop in the end of
  test track
• Control problem Dynamic trajectory planning

5
Transition when v lt max_spd
6
Transition when v gt max_spd
7
Real-time Tire Slip Ratio Estimation

• Friction Models (static or dynamic) are Tire Slip
  Dependent
• Tire Slip Definition Based on Short Distance
• Easy to implement
• Can be used for static and dynamic friction
  models
• Application to Vehicle Longitudinal moving
  Distance Measurements
• Vehicle in AHS installed with magnets
• Vehicles with GPS in normal highway systems
  (similar and omitted)

8
Tire Slip Definition Based on Vehicle Speed
9
Tire Slip Definition Based on Short Distance
10
Distance Measurement Vehicle in AHS installed
with magnets

• Tire slip in moving distance measurement based on
  magnets

11
Distance Measurement Vehicle in AHS installed
with magnets
12
Part 2. Heavy Duty Trucks

• Modeling and Control System Structure
• Main Characteristics From Control Point of View
• Modeling
• Upper Level Control
• Lower Level Control
• Communication Systems
• Most Difficult Problems in Control

13
Modeling Control System Structure
Air
Turbocharger
Lockup
Transmission Retarder
Torque Converter
Engine
Automatic Transmission
EGR
Engine Brake
Propeller shaft
Exhaust
Tire
Pneumatic Brake
Final Driving Gear
Drive shaft
Wheel speed
Tire slip effect
Control
Sensors Comm. Syst.
Vehicle speed
14
Main Characteristics From Control Point of View

• Mass dominant road inclination
  plays a major roll
• Low (power / inertia) ratio
  Throttle saturation
• Large time delay in actuators, mainly air brake
• Large load variation
• Prominent disturbance during gear shifting

15
Modeling

• Engine
• Other Power Consumption
• Drive-line
• Engine Braking Effect
• Rolling Resistance

16
Engine
17
Engine (Eng_brk 0)
Air
Turbocharger
Pm
Engine_fric torque
-
Engine
Net Output Torque
EGR
Engine_ind torque

Exhaust
18
Engine (Eng_brk 2,4)
Engine_ind_torque gt Engine_fric torque
engine_comp torque
Air
Turbocharger
Pm
Engine_fric torque
-
Engine

Net Output Torque
Engine_ind torque
EGR
Engine_comp torque
-
Exhaust
19
Engine (Eng_brk 2,4)
Engine_ind_torque lt Engine_fric torque
engine_comp torque
Air
Turbocharger
Pm
Engine_fric torque
-
Engine

Total Eng_brk Torque
Engine_ind torque
EGR
Engine_comp torque
-
Exhaust
20
Engine (Eng_brk 6)
Air
Turbocharger
Pm
-
Engine
Engine_fric torque
Total Eng-brk Torque
EGR
Engine_comp torque
-
Exhaust
21
Other Power Consumption

• New Freightliner Total power 435hp
• Fan (clutch engaged)
• Air compressor
• Alternator
• Air conditioner Compressor

22
Drive-line

• Torque Converter
• Drive line Torsional Effect and Its Compensation
  in Platoon Control Design
• At low speed
• During gear shifting
• Maneuver with large torque demanding (large
  acceleration, deceleration grading)

23
Engine Braking Effect

• Total resistance in longitudinal direction can be
  divided into two parts

24
Rolling Resistance

• Rolling Resistance - function of vehicle speed
  and mass

25
(No Transcript)
26
(No Transcript)
27
Upper Level Control

• General Strategy
• Sensors and Actuators (Do as it is)
• Close-loop Open-loop vehicle mass estimation
• Following Strategy, Data Fusion and String
  Stability

28
General Strategy Model based

• System Uncertainties
• Model mismatch
• External disturbances
• Sensor noises
• To reduce model mismatch as much as possible
• To choose proper filters
• To design controller with less overshoot (to deal
  with saturation)
• Developing prediction technique (to deal with
  time delay)

29
Sensors and Actuators (Do as it is)

• Two Radar Systems Eaton Vorad Lidar
• Magnetometer
• Inclinometer
• Built-in sensors
• Speedometer
• Accelerometer
• brake pressure
• Other J-BUS readings
• No extra actuator to be added

30
Close-loop open-loop vehicle mass estimation

• For string stability
• To be established during a short time after
  vehicle start moving

31
Following Strategy, Data Fusion and String
Stability

• Two-way (multi-way) following
• Corporative Platooning
• Optimal dynamic trajectory planning for leader
  vehicle
• Following and data fusion strategy in normal
  mode and sensor fault mode

32
Following Strategy, radar and/or magnetometer
is normal
33
Following Strategy, radar and/or magnetometer
is normal
34
Following Strategy, radar and magnetometer fault
35
Following Strategy, communication fault ?
36
Control Design Methods

• Multiple surface sliding mode Nonlinear
  H-infinity
• Newly developed continuous finite time sliding
  mode
• New filtering and observer techniques
• Other nonlinear control techniques

37
Lower Level Control

• Internal (built-in) Throttle Control System
• Torque control command
• Speed control command
• Fuel rate limit to guarantee optimal performance

Turbocharger
Air
Pm
Eng-speed
Engine
Throttle Dynamics
Fuel Rate
Des-Eng-speed
PI Controller
Des-Output Tq
Eng-Net-Torque
Exhaust
38
Lower Level Control (cont.)

• Brake Control System
• Combination of
• Engine brake
• Air brake
• Transmission retarder
• Taking into consideration of Engine Braking
  Effect
• Proper coordination

39
Communication System

• 2-way communication 2.4 GHz WaveLAN Radio
• Range 300500m

40
Most Difficult Problems

• Proper data fusion and optimal trajectory
  planning for string stability
• To deal with road inclination (grade) with
  reasonably good performance
• To overcome time delay in actuators