Truck suspensions

1
Truck suspensions
2
Conventional passive suspension
3
Active suspension
4
Fully-active suspensions
Actuator provides totalsuspension force
5
Slow-active suspension
6
Slow-active suspension
7
Semi-active suspension- dissipative forces only
8
(No Transcript)
9
Hardware-in-the-Loop simulation
10
Vs
Fd
Ar
P1 Pst DP
V1
P1
PV
ApAr
P2 Pst
P2
Vr
Pst
V2
Pst
Vu
Vrel Vu - Vs
Fd
11
Fd
Vs
Ar
P1 Pst DP
V1
P1
PV
ApAr
Vr
V2
Pst
P2P1
P2
Pst
Vu
P1 P2 Pst DP
Vrel Vu - Vs
Fd
12
Proportional control valve
13
Mechanical design

• Determine the leading dimensions of the damper
• rod length, diameter and wall thickness
• inner tube bore and wall thickness
• outer tube bore
• Remember the important specification that the
  bump and rebound force-velocity characteristics
  are to be symmetrical.

14
Damper design

• Convert the pressure?flow envelope of figure 7 to
  a damping force?relative velocity envelope for
  your design.
• Make plots on this chart of the damper force Fd
  versus relative velocity Vrel for values of Xv
  0.1, 0.2, 0.3, ?1.0.
• Make a separate plot of Fd versus Xv for
  different values of Vrel.

15
Force controller
16
Feedforward Feedback
17
Force controller design

• Given the linearised plant model, design a PI or
  PID controller for a chosen nominal operating
  condition, and check its robustness against
  changes in operating point.
• A suggested nominal operating condition is Fd0
  2500 N, Vrel0 0.15 m/s.
• Recall the specification that the desired
  bandwidth for the force controller is 20 Hz.

18
rltool
19
Alternative controller design

• Use the Ziegler-Nichols ultimate sensitivity
  method to design a PI or PID controller.
• That is, initially set the integral and
  derivative gains to zero, and increase the
  proportional gain until the system oscillates on
  the point of instability.
• Then measure the ultimate gain Ku and the
  ultimate period Pu, and apply the tuning rules
  to obtain a first-cut set of values for the
  controller gains.

20
fctrl.mdl
21
MSD controller design

• Design a real-time program for the M68HC11
  microcontroller to perform the semi-active damper
  control task.
• The MSD control law is defined in equations (4)
  and (5). Suitable initial parameter values are
  Cm 45 kN/(m/s) and ? 0.2.

22
Implementation

• Then implement your program in a
  hardware-in-the-loop simulation, using the
  SIMULINK model HiL_sys provided.
• The roadway roughness input can be selected to be
  deterministic (e.g., sinusoidal corrugations) or
  random (corresponding to a road profile that
  could be encountered on a main road at 70 km/h).
  
• Time histories of simulation variables will be
  written into the MATLAB workspace, so that the
  performance of the controller can be assessed.

23
Design tools provided

• SIMULINK model, SIM_sys
• This is identical with HiL_sys, except that a
  subsystem block M68HC11 is included as a
  representation of the microcontroller.
• You can modify this block to create your own
  SIMULINK representation of your controller code,
  to test its operation before attempting the HiL
  simulation.
• Ziegler-Nichols tuning tool fctrl
• invoke with fctrl_start

24
SIM_sys.mdl
25
Schedule
26
PID controllers

• PID Proportional Integral Derivative
• Also known as "three-term controller"
• About 90 of all control loops are closed with
  some form of PID controller
• In this group of lectures we will find out
• why PID controllers are used so often
• ways of "tuning" a PID controller
• how to deal with actuator saturation

27
Functions of control system

• Track reference input, or maintain set point,
  despite
• load disturbances (usually low frequency)
• sensor noise (usually high frequency)
• Achieve specified bandwidth, and transient
  response characteristics

28
Performance of control system

• Sensor noise reproduced just like reference input
• use low noise sensors!
• seek to make

• To reject disturbances, make

29
PID controller functions

• Output feedback
• from Proportional action
• Eliminate steady-state offset
• from Integral action
• Anticipation
• from Derivative action

compare output withset-point
apply constant control even when error is zero
react to rapid rate of change before error grows
too big
30
Transfer function of PID controller

• If no derivative action, we have PI controller

proportional gain
integral gain
31
Effects on open-loop transfer function

• s-plane

32
Effects on open-loop transfer function

• Frequency response

33
Application of PID control

• PID regulators provide reasonable control of most
  industrial processes, provided performance
  demands not too high
• PI control generally adequate when plant/process
  dynamics are essentially 1st-order
• plant operators often switch D-action off
  "dificult to tune"
• PID control generally OK if dominant plant
  dynamics are 2nd-order
• More elaborate control strategies needed if
  process has long time delays, or lightly-damped
  vibrational modes

34
Simulink PID models
35
Simulink PID models