Reduction in iron losses In Indirect Vector-Controlled IM Drive Using FLC
C. Srisailam Mukesh Tiwari Dr. Anurag Trivedi
Electrical Engineering Department Electrical Engineering Department Electrical Engineering DepartmentJabalpur Engineering College, Jabalpur Engineering College Jabalpur Engineering CollegeJabalpur, Madhya Pradesh Jabalpur, Madhya Pradesh Jabalpur, Madhya PradeshEmail:chikondra007@gmail.com Email:mukesh_tiwari836@yahoo.co.in Email:dr.anuragtrivedi@yahoo.co.in
Abstract
-
This paper describes the use of fuzzy logiccontroller for efficiency optimization control of a drive whilekeeping good dynamic response. At steady-state light-loadcondition, the fuzzy controller adaptively adjusts theexcitation current with respect to the torque current to givethe minimum total copper and iron loss. The measured inputpower such that, for a given load torque and speed, the drivesettles down to the minimum input power, i.e., operates atmaximum efficiency. The low-frequency pulsating torquedue to decrementation of flux is compensated in a feedforward manner. If the load torque or speed commandschanges, the efficiency search algorithm is abandoned andthe rated flux is established to get the best dynamic response.The drive system with the proposed efficiency optimizationcontroller has been simulated with lossy models of converterand machine, and its performance has been thoroughlyinvestigated.
Key words: Fuzzy logic controller (FLC), Fuzzy logic motor control (FLMC), Adjustable speed drives (ASDs)
.
I.INTRODUCTION
Efficiency improvement in variable frequency driveshas been getting a lot of attention in recent years. Higherefficiency is important not only from the viewpoints of energy saving and cooling system operation, but also fromthe broad perspective of environmental pollution control.In fact, as the use of variable speed drives continues toincrease in areas traditionally dominated by constantspeed drives, the financial and environmental payoffsreach new importance.A drive system normally operating at rated flux givesthe best transient response. However, at light loads, ratedflux operation causes excessive core loss, thus impairingthe efficiency of the drive. Since drives operate at lightload most of the time, optimum efficiency can be obtainedby programming the flux. A number of methods forefficiency improvement through flux control have beenproposed in the literature. They can be classified intothree basic types. The simple precompiled flux program asa function of torque is widely used for light loadefficiency improvement. The second approach consists inthe real time computation of losses and correspondingselection of flux level that results in minimum losses. Asthe loss computation is based on a machine model,parameter variations caused by temperature and saturationeffects tend to yield suboptimal efficiency operation. Theon-line efficiency optimization control [ 1]-[3] on thebasis of search, where the flux is decremented in stepsuntil the measured input power settles down to the lowestvalue, is very attractive. The control does not require theknowledge of machine parameters, is completelyinsensitive to parameter changes, and the algorithm isapplicable universally to any arbitrary machine.II. CONTROL SYSTEM DESCRIFTION“Fig. 1”, show the block diagram of an indirectvector controlled induction motor drive incorporating theproposed efficiency optimization controller. The feedback speed control loop generates the active or torque currentcommand () as indicated. The vector rotator receivesthe torque and excitation current commands andrespectively, from the two positions of a switch: thetransient position (l), where the excitation current isestablished to the rated value () and the speed loopfeeds the torque current and the steady state position,where the excitation and torque currents are generated bythe fuzzy efficiency controller and feed forward torquecompensator which will be explained later.
Fig.1. Efficiency optimization controller. Indirect vector controlledinduction motor drive incorporating the FLC.
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 1, April 2010313http://sites.google.com/site/ijcsis/ISSN 1947-5500
The fuzzy controller becomes effective at steady-statecondition, i.e., when the speed loop error
∆
,approaches zero. Note that the dc link power, instead of input power, has been considered for the fuzzy controllersince both follow symmetrical profiles. The principleof efficiency optimization control with rotor fluxprogramming at a steady-state torque and speed conditionis explained in “Fig. 2”. The rotor flux is decreased byreducing the magnetizing current, which ultimately resultsin a corresponding increase in the torque current(normally by action of the speed controller), such that thedeveloped torque remains constant. As the flux isdecreased, the iron loss decreases with the attendantincrease of copper loss. However, the total system(converter and machine) loss decreases, resulting in adecrease of dc link power. The search is continued untilthe system settles down at the minimum input power pointA, as indicated. Any excursion beyond the point A willforce the controller to return to the minimum power point.
A. Efficiency Optimization Control
“Fig.2”explains the fuzzy efficiency controlleroperation. The input dc power is sampled and comparedwith the previous value to determine the increment
∆
.In addition, the last excitation current decrement (L
∆
)is reviewed. On these bases, the decrement step of
∆
isgenerated from fuzzy rules through fuzzy inference anddefuzzification [4], as indicated. The adjustable gainsand, generated by scaling factors computation block,convert the input variable and control variable,respectively, to per unit values so that a single fuzzy rulebase can be used for any torque and speed condition. Theinput gain as a function of machine speed can begiven as(2)Where (3)
where the coefficients was derived fromsimulation studies. The output gain is computed fromthe machine speed and an approximate estimate of machine torque ().Efficiency Optimization Control Fig.3 explains thefuzzy efficiency controller operation. The input dc poweris sampled and compared with the previous value todetermine the increment
∆
. In addition, the lastexcitation current decrement (L
∆
) is reviewed. On
Fig.2. Principle of efficiency optimization control with rotor fluxprogramming
these bases, the decrement step of
∆
is generated fromfuzzy rules through fuzzy inference and defuzzification[4], as indicated. The adjustable gains and,generated by scaling factors computation block, convertthe input variable and control variable, respectively, to perunit values so that a single fuzzy rule base can be used forany torque and speed condition. The input gain as afunction of machine speed can be given as where thecoefficients was derived from simulation studies.The output gain is computed from the machine speedand an approximate estimate of machine torque ().
Theappropriate coefficients and were derived fromsimulation studies. A few words on the importance of theinput and output gains are
appropriate here. In the absenceof input and output gains, the efficiency optimizationcontroller would react equally to a specific v
alue of Δ
,
resulting from a past action, irrespective of operating speed. Since the optimal efficiency point A (seeFig. 2) is speed dependant, the control action could easilybe too conservative, resulting in slow convergence, orexcessive, yielding an overshoot in the search processwith possible adverse impact on system stability. As bothinput and output gains are function of speed, this problemdoes not arise. Equation (2) also incorporates that a prioriknowledge that the optimum value of , is a function of torque as well as machine speed. In this way, for different
speed and torque conditions, the same Δ
(pu) will
result in different Δ
, ensuring a fast convergence. Oneadditional advantage of per unit basis operation is that thesame fuzzy controller can be applied to any arbitrary
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 1, April 2010314http://sites.google.com/site/ijcsis/ISSN 1947-5500
machine, by simply changing the coefficients of input andoutput gains.The membership functions for the fuzzy efficiencycontroller are shown in Fig. 4. Due to the use of input andoutput gains, the universe of discourse for all variables arenormalized in the [-1, 1] interval. It was verified that,while the co
ntrol variable Δ
, required seven fuzzy setsto provide good control sensitivity, the past control action
LΔ
(i.e.,
∆
(k - 1)) needed only two fuzzy sets, sincethe main information conveyed by them is the sign. Thesmall overlap of the positive (P) and negative (N)membership functions is required to ensure properoperation of the height defuzzification method [4], i.e., toprevent indeterminate result in case L
∆
approacheszero.An example of a fuzzy rule can be given as IF the
power increment (Δ
) is negative medium (NM) and the
last Δ
(L
∆
) is negative (N), THEN the newexcitation increment (
∆
) is negative medium (NM).The basic idea is that if the last control action indicated adecrease of dc link power, proceed searching in the samedirection, and the control magnitude should be somewhatproportional to the measured dc link power change. Incase the last control action resulted in an increase of
(
Δ
> 0), the search direction is reversed, and the
∆
,step size is reduced to attenuate oscillations in the searchprocess.
Fig.4. Membership functions for efficiency controller. (a) Change of dc link power
∆
(pu)). (b) Last change in excitation current(L
∆
(pu)). (c) Excitation current control increment (
∆
(pu)).
B. Feed forward Pulsating Torque Compensation
Fig. 5. Feed forward pulsating torque compensator block diagram
As the excitation current is decremented with adaptivestep size by the fuzzy controller, the rotor flux willdecrease exponentially [5], which is given by:= (4)The decrease of flux causes loss of torque, whichnormally is compensated slowly by the speed controlloop. Such pulsating torque at low frequency is veryundesirable because it causes speed ripple and may createmechanical resonance. To prevent these problems, a feedforward pulsating torque compensator has been proposed.Under correct field orientation control, the developedtorque is given byFor an invariant torque, the torque current, shouldbe controlled to vary inversely with the rotor flux. Thiscan be accomplished by adding a compensating signal
∆
(t) to the original to counteract the decrease influx
∆(t), where t ε [O,T] and T is the sampling
period for efficiency optimization control. Let (0) and(0) be the initial values for, and, respectively,for the k-th step change of. For a perfect compensation,the developed torque must remain constant,
and thefollowing equality holds:Solving for
∆
yields (6)(7)Eq (7) is adapted to produceCompensated torque current step is computed by(8)
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 1, April 2010315http://sites.google.com/site/ijcsis/ISSN 1947-5500
Search History:
Result 00 of 00
00 results for result for
p.
Reduction in iron losses In Indirect Vector-Controlled IM Drive Using FLC
This paper describes the use of fuzzy logic controller for efficiency optimization control of a drive while keeping good dynamic response. At steady-state light-load condition, the fuzzy controller adaptively adjusts the excitation current with respect to…
(More)
167
Reads
1
Readcasts
1
Embed Views
More From This User
Notes
Load more
