Sie sind auf Seite 1von 6

Proceedings of the 2nd IFAC

Workshop on Fractional Differentiation and its Applications


Porto, Portugal, July 19-21, 2006

ANALOG FRACTIONAL ORDER


CONTROLLER IN A TEMPERATURE
CONTROL APPLICATION

Gary W. Bohannan ,1


Wavelength Electronics, Bozeman, Montana, USA

Abstract: An analog fractional order PI controller using a fractional order


impedance device, a FractorTM (patent pending), is demonstrated in a temperature
control application. The performance improvement over a standard PI controller
was notable in reduction of overshoot and decreased time to stable temperature
while retaining the long term stability and set point accuracy of the standard
controller. The modi cation of the standard controller to a fractional order
controller was as simple as replacing the integrator capacitor with a Fractor TM .

Keywords: Fractional Order Control, fractance

1. INTRODUCTION er to implement the PI controller described by


Podlubny and others. (Podlubny, 1999) The idea
There has been signi cant work focused on the of using a non-integer order impedance to create a
fact that dynamic properties of materials, par- fractional order mathematical operator dates back
ticularly electrical impedance, can often be de- to at least 1961. (Carlson and Halijak, 1961)
scribed with high accuracy using non-integer
Since thermal loads involve thermal di usion, a
order impedance descriptions.(Macdonald, 1987)
half-order controller represents a better match to
The Warburg impedance, with the impedance
the physics of the plant to be controlled. Earlier
varying as the square-root of frequency, was
work (Petras and Vinagre, 2002) showed that a
originally described at the end of the 19th
digital fractional order controller was e ective for
century.(Warburg, 1899) The non-integer order
temperature control and suggested that the test
impedance spectroscopic description has been tied
should be conducted with an analog circuit. In
to the fractional order calculus time domain dy-
this paper, an analog fractional order controller,
namics by a number of researchers, see, e.g.
with  0.5, is demonstrated in a temperature
(Westerlund and Ekstam, 1994).
control application.
By reversing the point of view from the fractional
calculus describing the dynamic behavior of a
material to viewing the materials dynamics as 2. THE FRACTORTM CIRCUIT ELEMENT
carrying out the fractional calculus operation, we
can create broadband analog fractional order op-
A FractorTM is a two lead passive electronic
erators with a high degree of accuracy. The output
circuit element similar to a resistor or capaci-
of a fractional order integrator can be summed
tor, but exhibiting a non-integer order power-
with the output of a standard proportional ampli-
law impedance versus frequency. The term frac-
tance has been used to describe fractional order
1 The author also holds an appointment as Affiliated Pro- impedance, consistent with the terms resistance
fessor of Physics at Montana State University, Bozeman. and capacitance. A fractance has a nearly con-
100,000

Log |Z| [Ohms] 10,000

1,000

100
10 100 1,000 10,000 100,000 1,000,000
Frequency [Hz]

(a)

0
-10
-20
Phase Angle

-30
-40
-50
-60
-70
-80
-90
10 100 1,000 10,000 100,000 1,000,000
Frequency [Hz]

(b)

Fig. 1. Spectral response of the FractorTM used in this demonstration project; (a) the impedance
magnitude and (b) impedance phase. The multiple lines show the variation over 26 impedance
measurement scans.
stant loss ratio (tan) over a large e ective measurements and the ultimate device perfor-
bandwidth. mance without trying to determine the meaning of
such descriptions as Farads to a fractional power.
The impedance behavior is accurately modelled
(Westerlund and Ekstam, 1994)
with the power-law form
Figure 1 shows the frequency response for the
K FractorTM used in the temperature control demon-
ZF rac () = , (1)
(j ) stration. Note that the nearly constant phase cov-
where K is the impedance magnitude, in Ohms, ers over seven decades of frequency. Additional
at a calibration frequency C = 1/ , and is the measurements have shown that the fractance be-
non-integer exponent, 0 < < 1. The phase shift havior with  0.5 adheres to the milli-Hertz
is related to the exponent by = 90o . The regime. Samples have exhibited stable fractance
fractance description in equation 1 also includes properties for over a year. FractorTM devices with
ideal resistance and capacitance in the limits as phase shifts other than 45o have also been created.
0 and 1, respectively. No digital implementation has yet achieved this
Note that the impedance de nition of equation e ective bandwidth. The phase ripple over fre-
(1) uses only real, integer order units of Ohms, quency is no worse than that achieved through
seconds, and Hertz. For actual circuit performance approximations. The prototype FractorTM ele-
speci cation and design this provides for a di- ments are currently made by hand, but are still
rect link between the impedance spectroscopic not much larger than typical through-hole ca-
pacitors, 2.5cm 2.5cm 0.6cm, it is feasible Rewriting equation (3) in terms of the Laplace
to implement fractional order control without variable, s = j,
the excessive space requirements of a network of K 1
integer-order analog elements, or the computing G(s) = . (4)
RI (s )
power requirements of a digital approximation.
See (Tenriero Machado, 1997) for a description It is apparent that equation (4) has the form of
of some of these approximation techniques. the Laplace transform of a fractional order inte-
Haba, et.al., demonstrated that it is possible to grator of order . (Oldham and Spanier, 1974)
create fractional order impedances in the range When summed with the output of a proportional
of 100 kHz to 10 GHz by fabrication of a fractal ampli er, the result is a PI controller. Since the
structure on silicon. (Haba et al., 2005) Fractance summing ampli er is also typically inverting, pos-
behavior has now been demonstrated over the itive polarity is restored without further circuitry.
frequency range < 10 2 to > 109 Hz. While the impedance spectroscopy data is most
As with any circuit element, the FractorTM has useful in characterizing a FractorTM , the more
voltage and power dissipation limitations as well convincing demonstration of the power-law nature
as some other restrictions on operating envi- of the fractional order integrator is shown in the
ronment. These are being actively investigated. time domain. Figure 3 shows the response of an
FractorsTM developed to-date do not contain lead analog fraction order integrator with  0.5
or other hazardous substances currently covered to a square wave input. The circuit of gure 2
under European RoHS standards. was followed by a unity gain inverter to restore
positive polarity. The t shape is clearly evident
in the output.

ZFract For the temperature control demonstration, a


RI standard 5 Amp analog temperature controller
VIN (Wavelength Electronics MPT-5000) was modi-
VOUT ed by removing the 1F integrator capacitor and
installing the FractorTM characterized by gure 1.
No other changes were required.

4. THE THERMAL LOAD


Fig. 2. Schematic for a fractional order integrator.
ZF represents the FractorTM element. The The thermal load (the test plant) for this
schematic symbol for the FractorTM was de- demonstration was a thermoelectric cooler (TEC)
signed to give the impression of a generalized mounted on a 14cm 7.5cm 0.5cm aluminum
Warburg impedance; a mixture of resistive block. The system was driven to current limit
and capacitive characteristics. to demonstrate the nonlinear e ect of actuator
saturation.
A 10 k negative temperature coe cient (NTC)
3. THE FRACTIONAL ORDER thermistor was used as the temperature sensor.
INTEGRATOR The MPT Series controllers provide a constant
100A excitation current for probing the ther-
mistor resistance. Since the impedance of the sen-
Analog fractional order operators can be designed
sor varies nearly exponentially with temperature,
using standard rules for operational ampli er cir-
there is an inherent non-linearity in the feedback
cuits. The gain of an ampli er circuit, such as
gain. The Steinhart-Hart relation was used to
shown in gure 2, is represented in the frequency
convert resistance to temperature.
domain as the ratio of the feedback to input
impedances.
5. RESULTS
VOU T ZF B ()
G() = = . (2)
VIN ZIN () A comparison of the e ectiveness of a conventional
PI controller versus a PI controller is shown in
In this instance the feedback impedance is given
gures 4a and 4b. Note that the overshoot is vir-
by equation (1) and the input impedance is a
tually eliminated. Time to stable temperature is
resistance, RI , resulting in an overall gain of
reduced by a factor of almost three. The integrator
K 1 windup e ect due to the time spent in actuator
G() = . (3)
RI (j ) saturation was virtually eliminated without the
Fractor with Square Wave Input

0.6

0.4
VIN VOUT
0.2
Voltage

0.0

-0.2

-0.4

-0.6
0.00 0.10 0.20 0.30 0.40 0.50
Time [sec]

Fig. 3. Time domain response of a half-order integrator using a FractorTM with properties similar to
those shown in gure 1. Curve b
10/19/05 Input Output

30
Overshoot due to integrator windup
28
Temperatures [C]

26
24
22
20
Current limit in effect
18
16
12:10 12:20 12:30 12:40 12:50 13:00
Time

Set Point Temp Actual Temp

(a) Conventional PI.

30
28
Temperatures [C]

26
Overshoot nearly gone Time to settle at temperature reduced
24
22
20
Current limit in effect
18
16
10:10 10:20 10:30 10:40 10:50 11:00
Time
Set Point Temp Actual Temp

(b) PI control.

Fig. 4. Comparison of results with standard PI (a) versus PI control (b).


extra complexity or expense of anti-windup cir- ditionally, a number of projects for acoustic and
cuitry. image signal conditioning are been considered.
In addition to the simplicity in the arrangement, Much work remains to be done to bound thermal,
the system exhibited much less sensitivity to aging and other e ects in the FractorTM devices.
changes in other parameters. Additional, assembly and packaging options for
shrinking the size of the devices are being in-
vestigated. Undoubtedly, di erent circumstances
6. ONGOING RESEARCH AND FUTURE will dictate the impedance magnitude, phase, and
PROSPECTS frequency band requirements, inevitably leading
to a catalog of fractance devices with a range
By allowing for digital control of device selection, of sizes and shapes. The requirements for addi-
it is possible to consider an auto-tuning mecha- tional FractorTM properties will evolve, guiding
nism for selection of gain magnitude and phase. additional work into the electrochemistry of the
See, e.g. (Chen et al., 2004). Additionally, hy- devices.
brid analog/digital computational processors are
demonstrating the potential to increase compu-
tational throughput over digital only systems by
7. CONCLUSION
many orders of magnitude while dramatically re-
ducing power requirements. (Cowen et al., 2005)
The FractorTM will soon add a very simple de-
Offloading fractional order operators to an analog
vice to the catalog of electronic elements avail-
co-processor o ers even greater savings in com-
able to the system designer, allowing for intu-
putational overhead while guaranteeing stability
itive and straightforward design of fractional or-
and gaining immunity from aliasing and round-
der controllers. Circuit con gurations that would
o errors.
be impractical or impossible to implement with
Other circuit con gurations besides that shown conventional devices should start to become com-
in gure 2 are possible, such as a fractional or- monplace.
der derivative operator, D , by rearrangement
of the input and feedback impedances. Expo-
nent arithmetic is also possible by incorporating
FractorsTM of di erent exponent orders. Numer- ACKNOWLEDGEMENT
ical modelling to predict the behavior of these
circuits, particularly when non-linear switched The author is indebted to Dr. S. Hurst of Montana
parameter conditions are included, stretches the State University, Bozeman, Montana, for develop-
current state-of-the-art in solving fractional or- ment of the FractorTM devices.
der di erential equations. See, e.g. (El-Mesiry et
al., 2004).
Given the very deep memory of the FractorTM , REFERENCES
it becomes all the more critical to bound the
Carlson, G. E. and C. A. Halijak (1961). Sim-
memory disturbance e ect outlined by Hartley
ulation
of the fractional derivative operator

and Lorenzo. (Hartley and Lorenzo, 2002) No-
s and the fractional integral operator 1/ s.
tably, the memory of the rst step transition
In: Central States Simulation Council Meet-
did not negatively a ect the performance at the
ing on extrapolation of analog computation
second temperature step. With devices exhibiting
methods, Kansas State University. Vol. 45.
such nearly exact single exponent behavior it will
pp. 122.
now be possible to arrange experiments to test
Chen, Y-Q., C. Hu, B. M. Vingre and C. A. Monje
the physical meaning of the initial conditions in
(2004). Robust PI controller tuning rule
the fractional calculus. While many mathemati-
with iso-damping property. In: Proc. Amer-
cal de nitions have been o ered, nature should
ican Control Conference (ACCS2004).
determine which of these should be accepted for
Cowen, G. E. R., R. C. Melville and Y. P. Tsividis
physical system modelling. (Podlubny, 1999)
(2005). A VLSI analog computer/math co-
More complex control demonstration projects are processor for a digital computer. In: Proc.
being considered, including control of flexible 2005 IEEE Solid-State Circuits Conf., San
robotic arms to give them more natural or Francisco, CA, Feb 2005.
human-like motion such as would be required in El-Mesiry, A. E. M., A. M. A. El-Sayed and
prosthetic devices or other human augmentation H. A. A. El-Saka (2004). Numerical so-
systems. These projects have given impetus to lution for multi-term fractional (arbitrary)
better understand the process of characterizing orders di erential equations. Computational
plants for e ective fractional order control. Ad- and Applied Mathematics 23(1), 3353.
Haba, T. C., G. Ablart, T. Camps and F. Olivie
(2005). Influence of the electrical parameters
on the input impedance of a fractal structure
realised on silicon. Chaos, Solitons and Frac-
tals 24, 479490.
Hartley, T. T. and C. F. Lorenzo (2002). Dynam-
ics and control of initialized fractional-order
systems. Nonlinear Dynamics 29, 201233.
Macdonald, J. Ross (1987). Impedance Spec-
troscopy. John Wiley & Sons. New York.
Oldham, K. and J. Spanier (1974). The Fractional
Calculus. Academic Press. New York.
Petras, I. and B. M. Vinagre (2002). Practical ap-
plication of dgital fractional-order controller
to temperature control. Acta Montanistica
Slovaca 7, 131137.
Podlubny, I. (1999). Fractional Di erential Equa-
tions: An Introduction to Fractional Deriva-
tives, Fractional Di erential Equations, to
Methods of their Solution and some of their
Applications. Vol. 198 of Mathematics in Sci-
ence and Engineering. Academic Press. San
Diego, CA.
Tenriero Machado, J. A. (1997). Analysis and de-
sign of fractional-order digital control sys-
tems. SAMS 27, 107122.
Warburg, E. (1899). Uber das Verhalten soge-
nannter unpolarisierbarer Electroden gegen
Wechselstrom. Ann. Phys. Chem. 67, 493
499.
Westerlund, S. and L. Ekstam (1994). Capacitor
theory. IEEE Trans. Dielectrics and Electri-
cal Insulation 1(5), 826839.

Das könnte Ihnen auch gefallen