23-03-2010, 11:16 PM
[attachment=2779]
Four-Quadrant Control of Switched Reluctance Motors
Presented By:
Dr. Iqbal Husain
Department of Electrical Engineering
The University of Akron
Akron, OH 44325-3904
Introduction Switched reluctance motor (SRM) drives are simpler in construction compared to induction and synchronous types of machine. Their combination with power electronic controllers may yield an economical solution. The switched reluctance motor with passive rotor has a simple construction, but the solution of its mathematical model is relatively difficult due to its dominant non-linear behavior.
OVERVIEW
MAIN FEATURES OF SRM
OPERATION AND CONTROL OF SRM
4-QUADRANT CONTROL OF SRM
RESULTS FROM OPTIMIZED 4-Q CONTROL
MODELING FOR SIMULATION AND
CONTROLLER DEVELOPMENT
4-QUADRANT SENSORLESS CONTROL
SIMULATION RESULTS
EXPERIMENTAL RESULTS
CONCLUSIONS
REFERENCES
BASIC CONSTRUCTION OF AN SRM
The SRM is a doubly-salient, singly-excited machine with independent windings of the stator.
Hts stator structure is same as PM motor, but the rotor is simpler having no permanent magnet on it.
Stator windings on diametrically opposite poles are connected in series or parallel to form one phase of the motor.
Several combinations of stator and rotor poles are possible, such as 6/4 (6 stator poles and 4 rotor poles), 8/4, 10/6, 12/6 etc.
4/2, 2/2 configurations are also possible, but with these it is almost impossible to develop a starting torque when the stator and rotor poles are exactly aligned.
The configurations with higher number of stator/rotor pole combinations have less torque ripple.
The design objectives are to minimize the core losses, to have a good starting capability and to eliminate mutual coupling.
Energy partitioning during one complete working stroke. (a) Linear case. (b) Typical practical case. W = energy converted into mechanical work. R = energy returned to the dc supply.
The nonlinear saturating characteristics of real magnetic steel has a marked influence on the energy conversion process in an SRM.
Only for very low values of saturation, the characteristics approximate the ideal linear case.
The flux-current characteristics in the unaligned position is approximately linear, because the magnetic path is dominated by large airgap and flux densities are small.
In the aligned position the airgap reluctance is small and flux density is high, which causes high saturation at higher currents.
The SRM is always driven into deep saturation to maximize the energy transfer in each stroke.
SRM
CHARACTERISTICS
Rotor Position in degrees
Torque-current-angle characteristics of an SRM
The phases in an SRM produce torque independent of each other.
The total torque is the sum of the individual phase torques.
The static torque-angle-current characteristics shows the overlap angles, which is useful is determining the commutation angle.
The torque dip in the characteristics is an indirect measure of expected torque ripple.
TORQUE-SPEED CHARACTERISTICS
Rotor Speed (Per Unit)
Region #1: Constant Torque
Current, and hence torque, kept constant by PWM or chopping.
At low speeds current rises instantaneously
due to small back-emf.
At medium speeds, phase advancing is necessary. Phase turn-off is also advanced so that current decays to zero before rotor passes alignment. PWM or chopping is still possible.
TORQUE-SPEED CHARACTERISTICS (Cont.)
Region #2: Constant Power
High back-emf forces current to decrease once pole overlap begins.
PWM or chopping no longer possible.
Conduction angle is increased in proportion to speed, primarily through phase advancing.
Maximum current can still be injected into the motor to sustain high enough torque.
Core and windage losses increase rapidly.
Constant power can generally be maintained upto 2-3 times the base speed.
Region #3: Natural characteristics
Upper limit of conduction angle is reached when equals half the rotor pole-pitch., i.e., half the electrical cycle at the onset of region #3.
Conduction angle is fixed, but pulse position ca be advanced.
Maintaining torque production is no longer
rtrtcciktlo anrl it fa lie rtff inx/orcolw with cnooH2
POSITION FEEDBACK
Conventional Method
Inverter
0 H
r
Sensorless Method
Control Inputs
Inverter
Control Strategies
Appropriate positioning of the phase excitation pulses is the key in obtaining effective performance
Control parameters: 0on, 0dwell and Iph
Control parameters determine torque, efficiency and other performance parameters.
Different Control Methods
Voltage controlled drives.
Current controlled drives.
Advanced controllers:
T/A or efficiency Maximization.
Torque ripple minimization.
Acoustic Noise Minimization.
Sensorless controllers.
CQref
Outer Loop Controller
PWM Controller
Duty Cycle
Electronic Commutator
Gate Signal
Converter
Vph
Voltage Controlled Drive
In low performance drives, a fixed frequency PWM voltage control with variable duty cycle provides the simplest form of control.
The angle controller generates the turn-on and turn-off angles depending on the rotor speed.
The duty ratio is changed according to the voltage command signal.
A speed feedback loop can be added on the outside, if speed control is desired.
The drive typically also incorporates a current sensor, placed in the lower leg of the dc bus, for over-current protection.
Outer Loop Controller
Torque Controller
Current Controller!
Gate Signal
Converter
+
Sign(.) Angle eoff Electronic
Calculator
Commutator
e
Current Controlled Drive
Used in torque controlled drives, where current is controlled in the inner loop.
The controller needs current feedback information from each phase.
The reference current is set by the torque command and the torque-angle-current characteristics of the motor.
The method allows rapid resetting of the current level and has applications where fast motor response is required.
POSITION HOLDING
Effect of Position Holding
The motor operation at a constant position will excite one phase for a prolonged period. This will lead to local overheating of any phases.
Solution
¢ When the position is held at a constant level, the oute loop must allow to dither the rotor around zero speed.
¢ This effect results in some rotor movement on the order of one step angle. Therefore, the motor goes through more than one phase during position hold.
20
16
12
8
Phase Currents During Position Hold
OPTIMIZATION RESULT COMPARISON
Rise Time Comparison
The response time (rise time) is considered as the time required for translational movement of 20% to 80% of the position command.
Operating parameters Rise time (msec)
Optimal turn on and turn off angles 48
Optimal turn on plus 10 and optimal turn off 58
Optimal turn on minus 10 and optimal turn off 51
Optimal turn on and optimal turn off plus 10 55
Optimal turn on and optimal turn off minus 10 52
The test results prove that the optimal turn-on and turn-off angles give the fastest response.
OBSERVER BASED 4Q SENSORLESS DRIVE
SIMULATION AND EXPERIMENTS
¢ The dynamics of the machine in state-space format is run in parallel with the real machine
SOFTWARE
| The model has the same inputs as the physic. machine. The difference between model outputs and measured outputs are used to force the estimated values to converge to the actual values.
The function ef provides an indirect way of evaluating the sign of position error.
CURRENT ESTIMATION
The dynamics of the machine in state-space format is run in parallel with the real machine.
The model has the same inputs as the physica machine. The difference between model outputs and measured outputs are used to force the estimated values to converge to the actual values.
Estimated Current:
The phase flux goes to zero in each, which helps impose the zero initial condition for the integrator repeatedly.
ERROR DYNAMICS OF SMO
Error Dynamics:
ISSUES WITH
4Q SENSORLESS DRIVE
> Effect of motor losses on model based flux calculation.
> Integrator problem of flux estimator at low and zero speed operation.
> Voltage measurement. >Zero speed sensorless operation.
> Continuous existence of information.
> Sensorless starting.
EFFECT OF MOTOR LOSSES
ON FLUX CALCULATION
Rm is considered across the back-emf to consider the effect of core losses. Therefore
and
X = XS (1 - e - lff (6))
where f= 7 lph-> and 7 is a core loss dependent parameter.
MODIFIED FLUX ESTIMATOR
Problems near zero speed
At low speeds, the flux estimator output may exceed the saturation flux due to measurement noise. This makes the observer unstable.
In the modified estimator, flux is reset to a lower value when saturation flux is reached.
Results on Modified Flux Estimator:
Time (sec)
¢The capacitor voltage rise due to regeneration must be considered for accurate flux estimation.
ZERO SPEED SENSORLESS
OPERATION
Measured and estimated position and speed during speed inversion and zero speed operation:
A high frequency bipolar speed is commanded to dither the motor at a constant position. This allows extracting information from the response of the system.
OTHER SENSORLESS ISSUE;
> Reference Current Lower Limit:
¢ SMO becomes unobservable due to nonexistence of information.
¢ To receive the continuous information fo the SMO, the lower limit of the iref can be set to a small current level, say 0.5A.
> Sensorless Starting:
¢ The error may be large at start-up, leadin to improper phase excitation, hesitation and possibly rotation reversal.
Solution:
¢ Drive the motor towards the desired direction using some preset phase voltages.
¢ This allows starting the motor as an open loop fashion. This time is sufficient to reach at the sliding surface.
Measured and estimated speed:
4Q Sensorless Control
^ SMO provides accurate position information in all quadrants at higher speeds.
> Toggling through zero speed has been achieved.
>The SMO may fail to estimate at zero speed operation (with extended stay at zero speed) due to dominating integrator problem.
>SMO needs a finite convergence time.
CONCLUSIONS
SRM is suitable for servo-type actuator
applications.
Four-quadrant control is necessary for
dynamic actuator type loads.
Appropriate turn-on and turn-off angles
based on certain optimization criterion, such as torque maximization,efficiency maximization, response time minimization etc. delivers high performance.
Development of four-quadrant sensorless
controller must consider practical limitations.
4Q sensorless operation demonstrated in
laboratory prototype experiments.
REFERENCES
: S.A. Hossain, I. Husain, H. Klode, B.P. Lequesne and A.M. Omekanda, "Four Quadrant Control of a Switched Reluctance Motor for a Highly Dynamic Actuator Load," to appear in IEEE-APEC, Mar. 2002.
: S.A. Hossain and I. Husain, Modeling of Switched Reluctance Motors for Practical Digital Implementation, submitted to IEEE Transactions on Power Electronics.
A.V. Radun, "Design Considerations for the Switche Reluctance Motor," IEEE Trans. on Industry Applications, Vo 31, No. 5, pp. 1079-1087, Sept./Oct. 1995.
: M.S. Islam, I. Husain, R.J. Veillette and C. Batur, "Design and Performance Analysis of Sliding-Mode Observers for Sensorles: Operation of Switched Reluctance Motors," accepted for publiation in IEEE Transactions on Circuits and Systems.
: R. McCann, M.S. Islam and I. Husain, "Application of Sliding Mode Observer for Switched Reluctance Motor Drives," To appear in Jan./Feb. 01 issue of IEEE Trans. on Industry Applications.
: S.A. Hossain, I. Husain, H. Klode, A.M. Omekanda and S.
Gopalakrishnan, "Four Quadrant and Zero Speed Sensorless Control of a Switched Reluctance Motor," to be presented in IEEE-IAS Annual Conference in Oct. 2002, Pittsburgh, PA.
please read http://esat.kuleuven.be/electa/research/...b_page.pdf for more of Four-Quadrant Control of Switched Reluctance Motors
