From MATLAB® and Simulink® to
Real Time with TI DSPs
Vehicle Dynamics
Content developed in partnership with
Tel-Aviv University
© 2007 Texas Instruments Inc,
0-1
Objectives
• To implement a simplified differential
equation for the motion of a car.
• To build and test a Simulink Model.
• To run the model in real-time using the
ezDSP F2812 hardware.
© 2007 Texas Instruments Inc, Slide 2
Motion of a Vehicle
• Consider the case of a car driving in a
straight line along a flat road.
© 2007 Texas Instruments Inc, Slide 3
Engine Power
• The driving force is supplied by the engine.
Engine Power
© 2007 Texas Instruments Inc, Slide 4
Vehicle Weight
• The weight of the vehicle will need to be
overcome to move the vehicle.
Vehicle Weight
© 2007 Texas Instruments Inc, Slide 5
Wind Resistance
• As the car moves, there will be wind
resistance.
Wind
Resistance
© 2007 Texas Instruments Inc, Slide 6
Vehicle Speed
• The engine power, vehicle weight and wind
resistance determine the vehicle speed.
Vehicle Speed
© 2007 Texas Instruments Inc, Slide 7
Combined Factors
• These factors can be brought together into an
equation of motion.
v
F
b.v
© 2007 Texas Instruments Inc,
m
Slide 8
Differential Equation
F = m.dv/dt + b.v where:
– F = force provided by the engine
– m = mass of vehicle
– dv/dt = rate of change of velocity (acceleration)
– b = damping factor (wind resistance)
– v = velocity (vehicle speed)
© 2007 Texas Instruments Inc, Slide 9
Transformed Equation
• To implement the equation using Simulink, the
equation needs to be first transformed.
• F/m –v.b/m= dv/dt
• We will set up a subsystem with:
– Force F as the input.
– Speed v as the output.
© 2007 Texas Instruments Inc, Slide 10
Continuous Implementation
• Using Simulink,
the equation can
be implemented
as a continuous
system as shown
in the diagram.
• To generate v, we
need to integrate
the acceleration
dv/dt.
© 2007 Texas Instruments Inc, Slide 11
Simulink Model
© 2007 Texas Instruments Inc, Slide 12
The Simulink Model
• The model of vehicle motion is shown below:
© 2007 Texas Instruments Inc, Slide 13
Description of Model
• The input to the system is the gas pedal, under
control of the driver.
• The “Engine Management” sub-system converts gas
pedal to engine power.
• The “Vehicle Dynamics” sub-system converts engine
power to vehicle speed.
• The output is provided in horsepower.
© 2007 Texas Instruments Inc, Slide 14
Engine Management Subsystem
• This converts the gas pedal input (0-100%)
to engine output power (0 – 200 hp).
© 2007 Texas Instruments Inc, Slide 15
Lookup Tables
• The conversion from
rpm to power can be
implemented using a
lookup table.
© 2007 Texas Instruments Inc, Slide 16
Lookup Table Curve
• The table values
can be adjusted to
fit a smooth curve.
© 2007 Texas Instruments Inc, Slide 17
Vehicle Dynamics Subsystem
• To implement
the equation of
motion on the
C28x, a Discrete
Time Integrator
is required.
© 2007 Texas Instruments Inc, Slide 18
Running the Simulation
• The ramp
generator
gently changes
the Gas Pedal
from 0% to
100%.
• This simulates
smooth
acceleration.
© 2007 Texas Instruments Inc, Slide 19
Tuning the Model
• Alter the mass m of the vehicle between 1 ton (for a
small compact car) and 35 tons (for a truck).
• Increase the wind resistance by increasing variable
b.
• Using real data from a car manufacturers website for
the Lookup Table. You could also profile a diesel
engine.
• Replace the Ramp input with a Step input to simulate
stamping on the gas pedal!
© 2007 Texas Instruments Inc, Slide 20
Introduction to Laboratory
© 2007 Texas Instruments Inc, Slide 21
Overview of Laboratory
• The Simulink model will be modified to run on the
ezDSP F2812 hardware.
• A potentiometer will be used to simulate the gas
pedal.
• The output speed of the system will be monitored
using a multi-meter.
© 2007 Texas Instruments Inc, Slide 22
Modifications for C28x
• To run on the ezDSP F2812, additional blocks from
the Embedded Target for TI C2000 DSP are required.
© 2007 Texas Instruments Inc, Slide 23
ADC Scaling
• The ADC input 0-
4095 needs to
be scaled 0-
100%.
• Using fixed-point
math, this can
be implemented
as multiply by
800 then divide
by 32768.
© 2007 Texas Instruments Inc, Slide 24
DAC Scaling
• The input 0-200 kph needs to be scaled 0-62500
for the DAC.
© 2007 Texas Instruments Inc, Slide 25
References
• ezDSP F2812 Technical Reference.
© 2007 Texas Instruments Inc, Slide 26
