Switched reluctance motor (SRM) drawback in high level of torque ripple (and hence acoustic noise), and speed fluctuation especially in low speed regions have limited the possibility of direct-drive applications in industry. Switched reluctance (SR) motors have the ability to generate high torques at very low speed and ideally suited for direct-drive application having robustness, high efficiency, low cost, simple structure, easy to maintain, high torque in low speed, smaller dimension of the motor in comparison to other motors.
Contrary to the conventional motors, the SRM is intended to operate in deep magnetic saturation to increase the output power density. Thus, due to the saturation effect and the variation of magnetic reluctance, all pertinent characteristics of the machine model (i.e. flux-linkage, inductance, phase torque etc.) are highly nonlinear functions of both rotor position and phase current. The ultimate outcome of all these nonlinearities is that the generated torque contains significant ripples when the motor is excited by the conventional rectangular pulse excitation scheme.
SRM is expanding it's application area, such as high speed application, oil pressure pumps, home electric appliances, etc, because of it's adequate mechanical strength, simple structure, maintenance free and low cost. Clearly, it is desirable to minimize torque ripple to improve performance characteristics and reduce acoustic noise and vibration
Several control methods and schemes have been proposed
to overcome these problems.
The techniques suggested in these works are mainly based on exhaustive measurement of magnetic characteristics of the SRM. All these methods address one fundamental issue (i.e. the determination of a suitable current flux waveform so that the
torque ripples are minimised).
This paper, analyzes the different torque ripple minimization techniques employed by researchers to overcome these problems.
Introduction
THE switched reluctance motor (SRM) drive is gaining recognition in the electric drive market due to its simple and rugged construction, low expected manufacturing cost, fault tolerance capability, easy to maintain, robustness ,high efficiency and high torque to inertia ratio.(J. Castro, P. Andrada, B. Blanqué 978-1-4673-0142-8/12/$26.00 ©2012 IEEE ) (Senior Member, IEEE, and Iqbal Husain, Member, IE IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 49, NO. 1, FEBRUARY 2002 )
Despte the advantages and variety of applications ,Torque ripple is a major problem of switched reluctance motor drive system, which causes undesirable vibration, acoustic noise and speed fluctuation especially in low speed regions have limited the possibility of direct-drive applications in industry.
Torque ripple is defined as the difference between the maximum and minimum instantaneous torque expressed as a percentage of the average torque during steady state operation
Torque Ripple (PLEASE SEE EQUATION)
(Jebarani Evangeline. S *, Suresh Kumar. S)
cause of torque ripple in SRM are mainly due to
the switching of phase currents into its windings (Jebarani Evangeline. S *, Suresh Kumar. S
This paper, analyzes the different "torque ripple minimization techniques" in switched reluctance motor drives employed by researchers to overcome these problems. Clearly, it is desirable to minimize torque ripple to improve performance characteristics and reduce acoustic noise and vibration.
It is generally suggested and agreed by the researchers that "There are primarily two approaches for reducing the torque ripple. One method is to improve the magnetic design of the motor, while the other is to use sophisticated electronic control techniques (Iqbal Husain, Senior Member, IEEE 0278-0046/02$17.00 © 2002 IEEE), (J. Castro, P. Andrada, B. Blanqué), (S.ayeed Mir, Member, IEEE, Malik E. Elbuluk, Senior Member, IEEE, and Iqbal Husain, Member, IE), (Jebarani Evangeline. S *, Suresh Kumar. S). These arguments clearly suggests that a good machine design and electronic control technique is a sure way to remove ripple torque in SWM drives. Despite this general agreement by researchers, a particular approach would be more effective. S. ayeded et al(IEEE TRANSACTIONS 1999) reported that Fuzzy control is one of the appropriate control schemes for torque control of SRM drives.
Many studies reported on the use of Torque -Sharing Function (TSF) Technique in minimizing torque ripples. Jebarani and Suresh (IEEE -------) found that to maintain the desired instantaneous torque, a high bandwidth current regulator is needed as shown in the figure below
Fig. 8. Torque Control using TSF
by tracking the torque produced in every individual phase by a torque sharing function (TSF) managed by the static characteristics of the motor which are functions of turn ON angle, overlap angle and motor speed. This proposed technique by Jebarani and Suresh is found wanting in the area of the spiky applied voltage as you must apply floquet transformation to obtain reference phase voltage. Instead of this Schramm et.al (paper [35],)proposed a linear TSF in which the torque varies linearly during the commutation and the value of current in both incoming and outgoing phases are equal at the central commutation point. Also, Iqbal Husain(------) agreed but added that magnetic linearity and constant-speed operation. Also, the "spiky" voltage requirement for this controller poses a constraint on
the converter power supply.
From these proposals it is clear that special designs of the motor magnetic structure will aid in the use of torque sharing function in minimizing torque ripples. These approaches for torque-ripple minimization subsequently generated many successful control techniques
Analysis has it that The torque distribution with respect to position in an SRM is arbitrary, which is at the root of the difficulty in developing the torque-ripple-minimizing controller.
A number of attempts in minimizing torque ripple were dedicated in designing torque-sharing functions, which would define the individual phase torque requirement. The methods control the rate of change of torque rather than the rate of change of current during commutation according to the predefined TSFs. (Iqbal Husain, Senior Member,IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 49, NO. 1, FEBRUARY 2002)
These approaches for torque-ripple minimization subsequently generated many successful control techniques as confirmed by Taylor et al ( 17th S-ymp Increment Motion Contr. Syst Do., Champaign, IL, June 1988, pp. 173-184)
From critical analysis of these approaches The choice of the TSF is not unique and many functions will satisfy the requirement and so there is no certainty of perfect minimization of torque ripples but result fluctuation.
For smooth production of torque in SRM, the phase currents are derived from the fundamental component of inductance model of the stator winding With higher harmonic content in it, it is difficult to solving the eq. (8) mathematically using conventional techniques.
For Torque Control Techniques the concept of rotating vector is extended to derive the spatially rotating inductance vector eventually produces a smooth torque [24]. Jebarani Evangeline. S *, Suresh Kumar. S () Holds that the major drawback of vector control is the need of pulse encoder and presence of a signal delay due to inclusion of PWM modulator limits the ability of rapid flux control.
Direct Torque Control (DTC) is indeed a best alternative to smooth and fast torque response which is based on the best voltage chosen based on the difference between the reference and actual value of torque and flux.
Jebarani Evangeline. S *, Suresh Kumar. S
Another widely proposed technique is the use of Linearization and Decoupling Techniques (LDTs) which is linearizes the motor and load charactersistics and decouples the control requirements of the SRM phases. The technique originated from (Taylor et al. [12], [13] )
The motor nonlinearities were taken into account by a reliable motor model. A nonlinear state feedback controller provided the desired decoupling assuming that the state variables (position,velocity, and current) were available.
An electronic commutator is used for tasks designation of the different phases. Torque and motion on one designation phase while other phases takes care of and stabilizing the stator current dynamics to guide against the production of negative torque.
Iqbal Husain found that the control law of the designated phase was utilized to cancel
the nonlinearities of the motor and load in order to linearize the system. The commutator, designed to be a fixed-time-interval type, forced the quickest ramp down of the turned-off phase. The fixed timing of commutation resulted in large peaks of phase
currents at low speeds. The balanced commutator identifies and delegates the responsibilities of all the phases, thereby decoupling the phases for control action. The currents in phases other than the strongest phase conducts as little as possible, assuring low copper losses. The balanced commutator approach achieves a secondary objective of maximizing drive efficiency (by minimizing ohmic losses in phases) in addition to the objective of torque-rippleminimization
From Iqbal Husain analysis it shows that the torque sharing among the phases using an accurate SRM model assures low torque ripple.
This technique determines the reference current accounting the motor non linearity. It is also buttressed by Jebarani and Suresh.
Jebarani and Suresh. Added that current tracking requires prior knowledge about the
static characteristics of motor. The electronic commutator used is commutated at fixed frequency with ramping down of the phase torque [11]. At low speeds, it results in large peaks in the phase currents. A balanced commutator algorithm with LDT approach reduces the peak current[29]. The computation of reference phase current is a function of rotor position e. In this approach, the strong phase which yields the desired torque with fewer phases current is identified. Critical rotor position occurs when the two phases produces the desired torque with the same phase current. A sinusoidal torque profile is used in this approach. The torque ripple index reduces with the maximum pole arcs
An issue as reported by researchers on LTDs is that the current in the designated torque producing phase increases just before commutation due to late commutation. A constant rate of change of current is used instead of controlling the torque rate employed
by the TSF techniques. very high flux derivatives are required to solve a single phase at a time in order to minimize torque ripple. This often leads to voltage saturation. This also limitss the upper speed limit of operation.
Iqbal Husain, (IEEE VOL. 49, NO. 1, FEBRUARY 2002) observed thatThe simulation of the LDTs demonstrated that they are capable of producing a ripple-free torque, but only when an accurate model of the machine is available. With less accurate models, the performance of the method degraded and, in some cases, even led to an unstable response due to high sensitivity to model uncertainties. The high computational requirement of the similarity transformation is another drawback, as this makes the digital real-time implementation possible only at lower speeds.
Torque-ripple minimization of switched reluctance drives can also be achieved using Fuzzy-Logic-Based Controller. The controller produces smooth torque up to the motor base speed.
The controller produces smooth torque up to the motor base speed. The torque is generated over the maximum positive torque-producing region of a phase. This increases the torque density and avoids high current peaks. The controller is robust toward errors in the rotor position information, which means an inexpensive crude position sensor can be used.( Iqbal Husain, 2002 IEEE)
The adaptive fuzzy control block shown in Fig. 2. The rotor position is used as the input and the outputs are the phase currents. The controller used an estimated torque obtained from a robust torque estimator in closed loop to determine the membership distribution of the adaptive fuzzy control block
Details of adaptive fuzzy control block shown in Fig. 12.
The phase advancing is achieved by continuous adaptation of the conduction segment of the fuzzy membership function. In order to maximize the conduction region, an optimum commutation angle is achieved by ending the conduction with a very small negative torque
Fuzzy control is one of the appropriate control schemes for torque control of SRM drives
S.ayeed et.al also proposed that the controller uses the rotor position as input and the outputs are the phase currents. The input is divided into membership functions designed to allow the SRM to conduct over the entire positive torque producing area. Max-product rule of inference scheme is used and the output is determined using the center of average for defuzzification. The membership sets of output variable are defined as singleton. The value (weight) of each singleton is initialized randomly and is optimized through adaptation
single phase or unipolar operation has disadvantages in SRM torque ripples. The magnetic nonlinearity and double salience , the mechanical torque produced is not constant at a one phase constant current excitation hence the total torque would be composed of torque pulses yield by each phase alone which leads to existence of quite large torque ripple.
Multiphase excitation of Switched reluctance motor with a special accent on three phase operation would certainly reduce the torque ripple [28]. Jebarani Evangeline. S *, Suresh Kumar. S
Chris et.al analyzed the distribution of magnetic forces (Radial and Axial force components) of the machine with Bipolar excitation. Bipolar Excitation has produced more number of Short Flux Path Excitation (SFPE) which improves the average torque produced and reduces the vibrations and acoustic noise in the motor [7].
The Bipolar operation of SRM would improve the current profile and reduce the total harmonic distortion (THD) by 26.77% [1].
special accent on three phase operation would certainly reduce the torque ripple [28]. Chris et.al analyzed the distribution of magnetic forces (Radial and Axial force components) of the machine with Bipolar excitation. Bipolar Excitation has produced more number of Short Flux Path Excitation (SFPE) which improves the average torque produced and reduces the vibrations and acoustic noise in the motor [7]. The Bipolar operation of SRM would improve the current profile and reduce the total harmonic distortion (THD) by 26.77% [1]. In a two phase excitation model discussed in [3], the torque ripple minimization is carried out using Fibonacci and exhaustive search methods. It was observed that Fibonacci search yielded better ripple reduction. Themodulation scheme for phase current considered in the above is by addition, subtraction and multiplication of constants and variables.
The use of Direct Instantaneous torque control in the minimization of torque ripple Is also explored as one of the possibilities. DITC is based on phase overlap to compensate fall of torque due to the phase commutation. Firing angles are selected from a look up
table according to the operation conditions (torque and
speed).
J. Castro, P. Andrada, B. Blanqué put it that the outgoing phase begins its demagnetization state, reducing the torque value, and the incoming phase is not able to provide enough torque to maintain a regular value until reaching a certain current value, The intelligent core of DITC is the generator of switching logic signals, which combines the firing angles
switching signals and the torque hysteresis regulator signals to produce the appropriate control signals to the power converter switches. The power converter, in this case a four phases asymmetric converter with two power switches and two diodes per phase, has three states, which are denominated magnetization state (Vcc), freewheeling state (fw) and demagnetization state (-Vcc). Firing angles are selected from a look up
table according to the operation conditions (torque and speed).
Iqbal Husain again offered an interesting hybrid torque-ripple-minimizing controller that incorporates the attractive features of some of the techniques developed. Infusing the concept of torque sharing over an extended region with the balanced commutator approach to minimize torque ripple. The ripple minimization over a wide speed range is achieved by varying the central commutation angle between and as a function of speed.
A concept which can also be called Torque -Sharing Function (TSF) with a flux-based model with a
cascade controller structure consisting of a feedforward nonlinear
torque compensator and a flux controller as shown in Fillicori et al. [15]
Two optimization algorithms were run to determine two sets of flux profiles to
achieve a secondary objective in addition to the primary objective of ripple-free torque.
The wide-speed-range drive is adapted with a self-tuning controller to improve the overall robustness of the system in [19]. The controller uses a nonlinear model of the machine, which is updated online using a recursive identification algorithm.
The computational requirement for the controller is much less than the LDTs
Unfortunately the the model was originally developed for an SRM with a larger number of rotor poles and does not hold for all types of SRMs.
During commutation both the incoming and outgoing phases have constant current ramps. The shortcoming
The LDTs use a constant rate of change of current during
commutation instead of controlling the torque rate employed
by the TSF techniques described earlier. Both the incoming and
outgoing phases have constant current ramps during commutation.
The drawback is that the current in the designated torque producing
phase increases just before commutation due to late
commutation. The torque ripple is minimized by solving a single
phase at a time and, thus, requires very high flux derivatives,
which often leads to voltage saturation. This also restricts the
upper speed limit of operation. The simulation of the LDTs
demonstrated that they are capable of producing a ripple-free
torque, but only when an accurate model of the machine is available.
With less accurate models, the performance of the method
degraded and, in some cases, even led to an unstable response
due to high sensitivity to model uncertainties. The high computational
requirement of the similarity transformation is another
drawback, as this makes the digital real-time implementation
possible only at lower speeds.( Iqbal Husain, Senior Member, IEEE
0278-0046/02$17.00 © 2002 IEEE
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 49, NO. 1, FEBRUARY 2002)
(􀂥). In harmonic
injection technique, the estimation of combined value of low harmonics for the load torque and ripple in the produced torque can be performed by an algorithm and harmonics currents can be injected into SRM for effective cancellation
of ripple due to current harmonics [36].
A Linearizing and Decoupled Techniques (LDT) is a method
to determine the reference current accounting the motor non
linearity. Jebarani Evangeline. S *, Suresh Kumar. S
B. Linearization and Decoupling Techniques (LDTs)
An advanced control system that linearizes the motor and load
characteristics and decouples the control requirements of the
SRM phases originated from the work of Taylor et al. [12], [13].
The motor nonlinearities were taken into account by a reliable
motor model. A nonlinear state feedback controller provided
the desired decoupling assuming that the state variables (position,
velocity, and current) were available. An electronic commutator
designated the appropriate tasks of the different phases.
The torque and motion were controlled by one designated phase,
while the other phases were controlled to stabilize the stator current
dynamics so that they do not produce any negative torque.
The control law of the designated phase was utilized to cancel
the nonlinearities of the motor and load in order to linearize the
system. The commutator, designed to be a fixed-time-interval
type, forced the quickest ramp down of the turned-off phase. The
fixed timing of commutation resulted in large peaks of phase
currents at low speeds.
The balanced commutator identifies
and delegates the responsibilities of all the phases, thereby
decoupling the phases for control action. The torque sharing
among the phases using an accurate SRM model assures low
torque ripple. The currents in phases other than the strongest
phase conducts as little as possible, assuring low copper losses.
The balanced commutator approach achieves a secondary
objective of maximizing drive efficiency (by minimizing ohmic
IMPORTANTlosses in phases) in addition to the objective of torque-ripple
minimization.
(􀂥). In harmonic
injection technique, the estimation of combined value of low
harmonics for the load torque and ripple in the produced
torque can be performed by an algorithm and harmonics
currents can be injected into SRM for effective cancellation
of ripple due to current harmonics [36].
Nihat et.al scheme presented a modulation based on the
square proportionality of current on Torque, in which the iref
is squared value of sum of phase current is implemented [26]
"." = rje1 'II
where q is the no of stator phases. A considerable decrease in
torque dip is seen. Due to hysteresis current control, the
system experiences subsonic noise which could be reduced by
reducing the controller bandwidth and increasing the
switching frequency. The above modulation strategy is
implemented with Sliding Mode Control (SMC) for the speed
control of the drive [25]. The SMC used, adjust the iref and
torque dips are eliminated with the modulation strategy. The
low frequency vibration in torque has been eliminated by
SMC. As a part of online approach proposed by Zhengyu et.al
[38], the current profiling are carried out from flux linkage
profiling or magnetization characteristics of motor using two
dimensional B-spline neural network (BSNN). The system
proposed with BSNN does not necessitate the use of high
bandwidth current controller and a torque sensor for initial
training of the network. A model based online estimation of
current is carried and compensates the reference current
using neuro-fuzzy system which also yielded less ripple in the
torque [21]. Jebarani Evangeline. S *, Suresh Kumar. S
5. 4 Torque Control Techniques Jebarani Evangeline. S *, Suresh Kumar. S
modulator limits the ability of rapid flux control. Direct
Torque Control (DTC) is a best alternative to smooth and fast
torque response. In DTC, the best voltage vector is chosen
based on the difference between the reference and actual
value of torque and flux. The commutation strategy is
determined by the hysteresis controller for flux and
torque[1O]. For control of instantaneous torque, Direct
Instantaneous Torque Control (DITC) is a closed loop DTC,
which comprises a digital torque hysteresis controller, without
torque profile function and auxiliary commutating strategy to
control the air gap torque online.
Torque ripple can be minimized over a wider operating
range by using electronic torque control techniques(J. Castro, P. Andrada, B. Blanqué)
Fuzzy control is one of the appropriate control
schemes for torque control of SRM drives
Fig. 2 shows a
block diagram of a proposed control scheme. The adaptive
fuzzy control block is shown in Fig. 3. The controller uses the
rotor position as input and the outputs are the phase currents.
The input is divided into membership functions designed to
allow the SRM to conduct over the entire positive torque producing
area. Max-product rule of inference scheme is used
and the output is determined using the center of average for
defuzzification
(S.ayeed Mir, Member, IEEE, Malik E. Elbuluk, Senior Member, IEEE, and Iqbal Husain, Member, IE).
