bode plots - openstudyassets.openstudy.com/updates/attachments/533af8d4e4b0b...hamid roozbahani a...
TRANSCRIPT
Hamid Roozbahani
A Bode plot is a graph of the transfer function of a linear,
time-invariant system versus frequency, plotted with a log-
frequency axis, to show the system's frequency response. It
is usually a combination of a Bode magnitude plot,
expressing the magnitude of the frequency response gain,
and a Bode phase plot, expressing the frequency response
phase shift. [1]
Bode Plots
Bode diagram
Let us assume that transfer function can be divided into parts
Thus the frequency response of the whole G can be calculated as a sum of the
frequency responses of G1…Gn
Magnitude
Phase
The idea in Bode’s method is to plot magnitude curves using a logarithmic scale and
phase curves using a linear scale. This strategy allows us to plot a high order G (jw) by
simply adding the separate terms graphically
Advantages of Bode plots
1. Dynamic compensator design can be based entirely on bode plots
2. Bode plots can be determined experimentally
3. Bode plots of system in series simply add which is quite convenient
4. The use of a log scale permits a much wider range of frequencies to be displayed
on the single plot than is possible with linear scales
Open-loop bode diagram
Km=gain margin
Φm=phase margin
If the both are positive the system is stable
Recommended values
Closed loop Bode-diagram
Mp=resonance peak < 2.3…3.5
dB (recommended for servo
systems)
The bandwidth of the system is a measure of speed of response. For control system it is
defined as the frequency corresponding to 0.707 (3db) in closed-loop magnitude Bode
plot.
Relation between resonance peak and overshoot
Calculation of responses by Matlab
The open-loop transfer function of a servo is
The bode diagram is calculated by typing
NUM= [30] (nominator polynomial coefficients)
DEN= [2.3e-4 1.2e-2 1 0] (denominator coefficients)
Sys=tf (NUM, DEN) (calculates transfer function)
Margin (sys) (calculates and plots Bode-diagram with gain and phase margins)
With direct feedback H(s)=1, the closed loop transfer function is
This in Matlab is
NUM=[30]
DEN=[2.3e-4 1.2e-2 1 30]
bode(NUM,DEN)
sys=tf(NUM,DEN)
(calculates Bode-diagram without margins)
Closed-loop frequency response
-If in the open-loop Bode-diagram the gain margin Km is bigger than 6dB and phase
margin Φm bigger than 45o increase the gain until one reaches the limit
-If Km or Φm or both are smaller than required the stability of servo is insufficient =>
increase the cylinder and valve or apply better control method
-If –3dB bandwidth in closed-loop Bode-diagram is smaller than required ω-3dB the gain
must be increased. If stability becomes insufficient => increase the cylinder and valve or
apply better control method [3]
Hand Sketching: Step-by-step approach
1- Put transfer function in ZPK form (factored zeros and poles, with a constant
multiplier K out front).
2- Identify breakpoints: distance of poles and zeros from the origin. Mark those on
the frequency axis of the plot. Remember to convert from rad/sec to Hz.
3- Determine low or high frequency constant asymptote of gain by taking the limit
of H(s) as s → 0 or infinity, respectively. Convert to dB.
4- Start at one of the asymptotes that is constant. (see below if neither is constant).
Move along in frequency until you get to a breakpoint. Each breakpoint is
associated with a change in slope of +/-20 dB/decade (+/-6 dB/octave). From left
to right, a zero produces an increase in slope (The increase could be from
negative to less negative, or from positive to more positive, etc.) Each pole
produces a decrease in slope. Work through all the breakpoints, and check that
the final asymptote is correct.
5- Sketch in a smoother curve, 3 dB below or above each breakpoint (unless it is a
double pole or zero in which case it is 6 dB below or above).
6- Now fill in the phase by the same procedure: Find the phase on the asymptotes
by looking at the limit of H(jω) as ω → 0 or ω → infinity. If the limit is real, the
phase is 180 degrees or zero. If the limit is imaginary, the phase is +/-90 degress
(-90 for a negative imaginary limit). If the limit is zero, you have to look at what
direction it came in to zero from. If it came from a positive imaginary number, the
phase is 90 degree; a negative imaginary number -90 degrees. If it came from a
positive real number, the phase is 0, and if it came from a negative real number,
the phase is -90° Then from left to right, each pole causes a –90° transition in
phase, and every zero causes a +90° transition in phase. You can use the rules
for the slope of the transition, but it’s usually not worth the trouble to get that
exact.
7- OK, what if neither asymptote is flat? For example H(s) = s/[(s+1)(s+100)]:
HF limit: s/100 -> 0
LF limit: s/(s2) -> 0
Solution: Consider the shape—it will look like:
With breakpoints at 1/(2π) and 100/(2π)
Instead of starting at HF or LF asymptotes, you can start with the flat center part.
Consider the approximate value there, where s>>1 and s<<100. Then we can write the
transfer function, putting the negligible stuff in a small font: H(s) = s/[(s+1)(s+100)] Then.
Thus, we can approximate H(s) as s/[100s] = 1/100 or –40 dB in the flat middle part.
Then we can start there and sketch the rest. (It slopes down with +/- 20 dB/decade
slope on either side of the plateau.) [4]
Also you can find helpful information regarding Bode plots in below web address too:
http://www.swarthmore.edu/NatSci/echeeve1/Ref/LPSA/Bode/BodeHow.html
References:
[1] Wikipedia- Bode plot-http://en.wikipedia.org/wiki/Bode_plot
[2] Original EE 105 Discussion Notes from Meghdad Hajimorad (“Amin”)- Last Modified by: Bill Hung- Date: 5 August 2006
[3] Course material by Prof. Huapeng Wu
[4] Engineering Sciences 22 — Systems Summer 2004 -Bode Plots Page 1-BODE PLOTS