OSU Research Programs In Mechatronics · PDF fileOSU Mechatronics Lab. 1 OSU Research Programs In Mechatronics Systems ... OSU Mechatronics Lab. 2 ... Power Ratings of DGS unit 1 and
Post on 10-Mar-2018
230 Views
Preview:
Transcript
OSU Mechatronics Lab. 1
OSU Research Programs In Mechatronics Systems
Prof. Ali Keyhani
October 22, 2004 (IAB’04)
Mechatronics Systems LaboratoryDepartment of Electrical and Computer Engineering
The Ohio State University
OSU Mechatronics Lab. 2
Publications of the Year ’041. "Control of Distributed Generation Systems Part I: Voltage and Current
Control," IEEE Transactions on Power Electronics, Vol. 19, No. 6, November 2004, (M. N. Marwali and A. Keyhani)
2. "Control of Distributed Generation Systems Part II: Load Sharing," IEEE Transactions on Power Electronics, Vol. 19, No. 6, November 2004, (M. N. Marwali, J. W. Jung, and A. Keyhani)
3. "Composite Neural Network Load Models for Power System Stability Analysis," IEEE Power Engineering Society 2004 Power Systems Conference & Exposition, October 10-13, 2004, New York city, NY (Ali Keyhani; Wenzhe Lu; Gerald T. Heydt)
4. "Power Floe Control of a Single Distributed Generation Unit with Nonlinear Load," IEEE Power Engineering Society 2004 Power Systems Conference & Exposition, October 10-13, 2004, New York city, NY (Min Dai; Mohammad N. Marwali; Jin-Woo Jung; Ali Keyhani)
5. "Neural Network Modeling of Power System Loads," submitted to International Journal of Neural Systems (IJNS), July 2004, (Wenzhe Lu; Ali Keyhani; Gerald T. Heydt)
6. "Torque Ripple Analysis of a PM Brushless DC Motor Using Finite Element Method," IEEE Transactions on Energy Conversion, Vol. 19, No. 1 , Page 40-45, March 2004, (Dai, Min; Keyhani, A.; Sebastian, Tomy)
7. "A PWM rectifier control technique for three phase double-conversion UPS under unbalanced load," accepted for oral presentation on IEEE APEC'05, (Min Dai; Mohammad N. Marwali; Jin-Woo Jung; Ali Keyhani)
OSU Mechatronics Lab. 3
Current Graduate Students
Nanda Marwali, PhD student, graduating December, 2004Min Dai, PhD student, graduating June, 2005Wenzhe Lu, PhD student, graduating June, 2005Jin-woo Jung, PhD student, graduating June, 2005
OSU Mechatronics Lab. 4
Outline
Control of Z-Source Power Converters for Fuel Cell SystemsControl of Power Converters for Distributed GenerationBy-Wire CarsModeling and Control of SRM in Electromechanic Brake Systems
OSU Mechatronics Lab. 5
Control of Z-Source Power Converter for Fuel Cell
SystemsPhD Student: Jin-Woo JungAdvisor: Prof. Ali Keyhani
OSU Mechatronics Lab. 6
I. Introduction
Future Trends of Electric Utility Industry
OSU Mechatronics Lab. 7
I. Introduction
Operating System For DGS
OSU Mechatronics Lab. 8
II. Distributed Genration Systems
Two Applications of Low Voltage DGS
(a) Grid-interconnection (b) Standalone AC power supply
Fig. 1 Configurations for two applications of DGS.
OSU Mechatronics Lab. 9
III. Single DGS UnitA. System Modeling
Circuit Model of Single DGS unit
Linv
Cinv Cinv
Delta-Wye Transformer
LOAD
U
V
W
x
z
n
y
aIinv
bIinv
cIinv
aIload
bIload
cIload
Cgrass
ca
bc
ab
VinvVinvVinv
cn
bn
an
VloadVloadVload
ca
bcVpwmVpwmVpwm
Vdc
gating signals DSP system voltages and currentsmeasurement
ab
Fig. 3 Circuit model for a single DGS unit.
OSU Mechatronics Lab. 10
A. System ModelingIII. Single DGS Unit
Equivalent Model of ∆/Y Transformer
DELTA-WYE TRANSFORMER
+
-
+
-
+
-
Ltrans
Ltrans
Ltrans
n
Rtrans
Rtrans
Rtrans
aIsnd
bIsnd
cIsnd
bIsndtr ⋅−
cIsndtr ⋅− aIsndtr ⋅−
u
v
w
x
y
zabVinvtr ⋅−
bcVinvtr ⋅−
caVinvtr ⋅−
Fig. 4 ∆/Y transformer model.
OSU Mechatronics Lab. 11
A. System ModelingIII. Single DGS Unit
State equations in abc reference frame
abciinv
abcinv
abc sndITrC
invICdt
invVd rrr
⋅⋅
−⋅
=3
13
1
abcinv
abcinv
abc invVL
pwmVLdt
invId rrr
11−=
abcload
abcload
abc loadIC
sndICdt
loadVd rrr
11−=
abctran
abcvtran
sndtran
transabc loadVL
invVTrL
ILR
dtsndId rrrr
11−⋅+−=
[ ]Tcabcababc VinvVinvVinvinvV =r [ ]
[ ]Taccbba
Tcabcababc
IinvIinvIinvIinvIinvIinv
IinvIinvIinvinvI
−−−=
=r
where,
[ ]Tcbaabc IsndIsndIsndsndI =r[ ]Tcbaabc VloadVloadVloadloadV =
r
−−
−⋅=
010001100
trTrv[ ]Tcbaabc IloadIloadIloadloadI =r
−−
−⋅=
112211
121trTri
OSU Mechatronics Lab. 12
A. System ModelingIII. Single DGS Unit
State equations in the stationary qd reference frame
0031
31
qdqdinv
qdinv
qd sndITriC
invICdt
invVd rrr
⋅⋅
−⋅
=
qdinv
qdinv
qd invVL
pwmVLdt
invId rrr
11−=
000 11
qdload
qdload
qd loadIC
sndICdt
loadVd rrr
−=
000 11
qdtran
qdqdtran
qdtran
tranqd loadVL
invVTrvL
sndILR
dtsndId
−⋅+−=r
r
[ ]
−⋅=⋅⋅= −
013031
23
2,11
0 trKTrKTri rowSisqd
−
−−=
5.05.05.0232305.05.01
32
SKwhere,
[ ]
−⋅=⋅⋅= −
0013
31
21
2,11 trKTrKTrv colSvsqd
OSU Mechatronics Lab. 13
B. Control System DesignIII. Single DGS Unit
Control Block Diagram
Outer Loop: Robust ServoMechanism VoltageController (RSC)
Inner Loop: Discrete-time Sliding Mode Current Controller (DSMC)
Robust ServoMechanismController Limiter Discrete Sliding Mode
Controller
( )krefV qd
r
+-
+
-
( )kloadV qd
r
Vqder
*qdcmdI
rqdcmdI
r
Iqder
( )kinvI qd
r
Line-to-LineVoltageSpaceVectorPWM
( )kpwmV qd
r
PWM timing
states states
Fig. 5 Overall control system.
OSU Mechatronics Lab. 14
C. Experimental ResultsIII. Single DGS Unit
System Parameters
Cgrass = 90 µFOutput Filter
Ts = 1/(3.2 kHz) = 320 µsSwitching/Sampling Period
Turn Ratio: 245 : 208Ltrans = 48 µH (≈ 0.03 p.u.), Rtrans = 0.02 Ω
∆/Y Transformer
Cinv = 540 µF, Linv = 300 µHInverter Filters
Vload = 208 V (LL-RMS), 120 V (LN)f = 60 Hz
AC Output Voltage
Vdc = 540 V (nom.) 390 V (min)DC Bus Voltages
OSU Mechatronics Lab. 15
C. Experimental ResultsIII. Single DGS Unit
Steady-state performance comparisonsbetween PI controllers and proposed controllers
(b) Proposed control methods(a) PI controllers
Fig. 8 Steady-state performance comparisons under nonliner load.(Top: three phase load currents. Bottom: Three-phase load voltages)
OSU Mechatronics Lab. 16
C. Experimental ResultsIII. Single DGS Unit
Steady-state linear load (1)
(a) 100% resistive balanced (b) 100% 0.8 p.f. load
Fig. 9 Steady-state linear load.(Top: load currents. Middle: Bottom: Three-phase load voltages)
OSU Mechatronics Lab. 17
C. Experimental ResultsIII. Single DGS Unit
Steady-state linear load (2)
(c) 100% resistive unbalanced (phase A unloaded) (d) 100% resistive unbalanced (phase A&B unloaded)
Fig. 10 Steady-state linear load.(Top: load currents. Middle: load voltages. Bottom: inverter voltages)
OSU Mechatronics Lab. 18
C. Experimental ResultsIII. Single DGS Unit
Transient Response
(a) 0% to 100% (b) 100% to 0%
Fig. 11 Resistive load transient.(Top: three-phase load currents. Bottom: three-phase load voltages)
OSU Mechatronics Lab. 19
C. Experimental ResultsIII. Single DGS Unit
Response of Three-Phase Short Circuits
Fig. 12 Three-phase short-circuits on output terminals.(Top: inverter currents. Middle: load voltages. Bottom: inverter voltages)
OSU Mechatronics Lab. 20
C. Experimental ResultsIII. Single DGS Unit
Output Voltage THD
2.7%Crest load (3:1)1.89%100% unbal. resistive (ph.A&B)1.70%100% unbal. resistive (ph.A)1.32%100% 0.8 pf load1.30%100% balanced resistive load0.90%No load
Output voltages THDTYPES OF LOAD
OSU Mechatronics Lab. 21
IV. Two DGS UnitsA. System Configuration
Circuit Model of Two DGS units
Fig. 13 Circuit model for two DGS units.
OSU Mechatronics Lab. 22
B. Control RequirementsIV. Two DGS Units
Complex Power at the load due to inverter i*iiii IVjQPS ⋅=+=
*sincos
−+=∴
i
iiiii Lj
VjEEVS
ωδδ*
* sincos
−+=
i
iiiii Lj
VjEEI
ωδδ
ii
ii LVE
P δω
sin=i
iii L
VVEQ
ωδ 2cos −
=
Active and Reactive Power flowing from the i-th inverter
11 δ∠E 22 δ∠E
1Ljω 2Ljω
0∠V
1I 2I
Fig. 14 Two inverters connected to a load.
OSU Mechatronics Lab. 23
B. Control RequirementsIV. Two DGS Units
Control variables for P and Q
P → power angle “δ”
Q → inverter output voltage “E”
Control Contraints
Locally measurable feedback signals (voltages/currents)
Data communications between each DGS about real power and reactive power
Wire impedance mismatches between inverter output and load bus
Voltage/current sensor measurement error mismatches
Tie-line impedance between loads
OSU Mechatronics Lab. 24
D. Simulation ResultsIV. Two DGS Units
Case 1Power Ratings of DGS unit 1 and 2: 600 kVA, respectivelyLoad 1: Pload1 = 480 kW/ Qload1 = 360 kVar (p.f. = 0.8)Load 2: Pload2 = 480 kW/ Qload2 = 360 kVar (p.f. = 0.8)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
100
200
300
400
500
600
Time [s ec]
P1, P
2 [kW
]
Two Active P owers
P 1P 2
(b) Real Powers (P1 and P2) 1.9 1.91 1.92 1.93 1.94 1.95 1.96 1.97 1.98 1.99 2
-1000
-800
-600
-400
-200
0
200
400
600
800
1000
Time [s ec]
I L1, I
L2 [A
]
Two Load Currents (Trans ient)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
100
200
300
400
500
600
Time [s ec]
Q1, Q
2 [kva
r]
Two Reactive P owers
Q1Q2
(a) Load currents (IL1 and IL2)
(c) Reactive Powers (Q1 and Q2)
Fig. 19 Simulation Results for Case 1.
OSU Mechatronics Lab. 25
D. Simulation ResultsIV. Two DGS Units
Case 2Power Ratings of DGS unit 1 and 2: 600 kVA, 500 kVA, respectively
1.9 1.91 1.92 1.93 1.94 1.95 1.96 1.97 1.98 1.99 2-1000
-800
-600
-400
-200
0
200
400
600
800
1000
Time [s ec]
I L1, I
L2 [A
]
Two Load Currents (Trans ient)
0 0.2 0.40
100
200
300
400
500
600
P1, P
2 [kW
]
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
100
200
300
400
500
600
Time [s ec]
Q1, Q
2 [kva
r]
Two Reactive Powers
Q1Q2
(a) Load currents (IL1 and IL2) (b) Real Pow
(c) Reactive Powers (Q1 and Q2)
Load 2: Pload2 = 400 kW/ Qload2 = 300 kVar (p.f. = 0.8)Load 1: Pload1 = 480 kW/ Qload1 = 360 kVar (p.f. = 0.8)
0.6 0.8 1 1.2 1.4 1.6 1.8 2Time [s ec]
Two Active P owers
P 1P 2
ers (P1 and P2)
Fig. 20 Simulation Results for Case 2.
OSU Mechatronics Lab. 26
D. Simulation ResultsIV. Two DGS Units
Case 3Power Ratings of DGS unit 1 and 2: 600 kVA, respectively
Load 2: Pload2 = 240 kW → 480 kW (after 1.6s), Qload2 = 180 kVar → 360 kVarLoad 1: Pload1 = 480 kW/Qload1 = 360 kVar (p.f. = 0.8)
1.4 1.5 1.6 1.7 1.8 1.9 2-1000
-800
-600
-400
-200
0
200
400
600
800
1000
Time [s ec]
I L1, I
L2 [A
]
Two Load Currents (Trans ient)
1.4 1.5 1.6 1.7 1.8 1.9 20
100
200
300
400
500
600
Time [s ec]
Q1, Q
2 [kva
r]
Two Reactive P owers
Q1Q2
(a) Load currents (IL1 and IL2)
(c) Reactive Powers (Q1 and Q2)
1.4 1.5 1.6 1.7 1.8 1.9 20
100
200
300
400
500
600
Time [s ec]
P1, P
2 [kW
]
Two Active P owers
P 1P 2
(b) Real Powers (P1 and P2)
Fig. 21 Simulation Results for Case 3.
OSU Mechatronics Lab. 27
D. Simulation ResultsIV. Two DGS Units
Case 4: Nonlinear load
1.5 1.51 1.52 1.53 1.54 1.55 1.56 1.57 1.58 1.59 1.6-800
-600
-400
-200
0
200
400
600
800
Time [s ec]
I L1, I
L2 [A
]
Two Load Currents (Trans ient)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.60
50
100
150
200
250
300
350
400
450
500
Time [s ec]
P1, P
2 [kW
]
Two Active P owers
P 1P 2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.60
50
100
150
200
250
300
350
400
450
500
Time [s ec]
Q1, Q
2 [kva
r]
Two Reactive P owers
Q1Q2
(a) Load currents (IL1 and IL2) (b) Real Powers (P1 and P2)
(c) Reactive Powers (Q1 and Q2)
Power Ratings of UPS unit 1 and 2: 600 kVA, respectivelyLoad 1: Three-Phase Bridge Diode (CDC1 = 10000 µF and RL1 = 3.25 Ω) Load 2: Three-Phase Bridge Diode (CDC2 = 10000 µF and RL2 = 3.25 Ω)
Fig. 22 Simulation Results for Case 4.
OSU Mechatronics Lab. 28
E. Experimental ResultsIV. Two DGS Units
System Parameters
Cgrass = 90 µFOutput Filter
Ts = 1/(3.2 kHz) = 320 µsSwitching/Sampling Period
Turn Ratio: 245 : 208Ltrans = 48 µH (≈ 0.03 p.u.), Rtrans = 0.02 Ω
∆/Y Transformer
Cinv = 540 µF, Linv = 300 µHInverter Filters
Vload = 208 V (LL-RMS), 120 V (LN)f = 60 Hz
AC Output Voltage
Vdc = 540 V (nom.) 390 V (min)DC Bus Voltages
OSU Mechatronics Lab. 29
E. Experimental ResultsIV. Two DGS Units
Steady-state linear load sharing performance
(b) Unbalanced resistive load(a) Balanced resistive load
Fig. 23 Steady-state linear load sharing.(Each trace box: per-phase load voltage and two units’s output currents)
OSU Mechatronics Lab. 30
E. Experimental ResultsIV. Two DGS Units
Steady-state nonlinear load sharing performance
Fig. 24 Steady-state nonlinear load sharing.(two units’s output currents for phase a, b, c)
OSU Mechatronics Lab. 31
E. Experimental ResultsIV. Two DGS Units
Dynamic load sharing performance
(a) 0% to 100% (b) 100% to 0%
Fig. 25 Resistive load transient.(two units’s output currents for phase a, b, c)
OSU Mechatronics Lab. 32
E. Experimental ResultsIV. Two DGS Units
Different load applied on each phase
Fig. 26 Unit 2 is reconnected with different load applied on each phase. (Phase a and b : linear resistive load. Phase c: single-phase non-linear load.)
OSU Mechatronics Lab. 33
V. Conclusions
Standalone AC Power Supply
Design of Three Controllers: Voltage, Current, and Power Controller
Good Performance for Single DGS Unit and Two DGS Units
Nearly zero steady-state voltage (RMS) error
Low Total Harmonic Distortion (THD)
Fast/no-overshoot current response
Good voltage regulation
Good load sharing by P and Q control
OSU Mechatronics Lab.
Modeling and Control of a Fuel Cell Based Z-source Converter for Distributed Generation Systems
1. Conventional Topology: a DC-DC boost converter and a DC/AC inverter DSP controller: two controllers (DC/DC and DC/AC power converters)Power devices: ⇒ DC to DC boost converter: four power switches and four diodes⇒ DC to AC inverter: six power switchesSensors: DC input, DC output, and AC output
FuelCell(Vin)
S1 S3
S4 S2
S1 S3
S4 S6
3-phaseload
S5
S2
DC to DC boost converter DC to AC inverter
Fig. 1 Conventional system configuration.
OSU Mechatronics Lab. 35
2. Conventional Z-source Topology
Conventional Z-source converter: Impedance source (L-C) and a DC/AC inverterBoosted by shoot-though zero vectors (both switches turned-on)Open-loop control under only linear/heavy loadDynamic load applications like motorFuel cell modeled by a DC voltage source (battery)
Fig. 2 Conventional Z-source converter.
OSU Mechatronics Lab. 36
3. Proposed Z-source Topology
Proposed Z-source converter:Static load applications with fixed peak voltage/frequency (i.e., three-phase AC 208V and 60 Hz)Dynamic response of fuel cell considered System modeling/modified SVPWM implementation/closed-loop control system designGood performance under both linear load and nonlinear loadWide range of load, i.e., light load to a full load
Fig. 3 Proposed Z-source converter.
OSU Mechatronics Lab. 37
5. Entire Control Loop Structure
Fig. 5 Total control system block diagram.
where, Three controllers1. Discrete-time Optimal Voltage Controller2. Discrete-time Sliding Mode Current Controller (DSMC) 3. Discrete-time PI DC-link Voltage Controller
One observer1. Asymptotic Observer for estimation of load currents
OSU Mechatronics Lab.
Control of Power Converters for Distributed Generation
Ph.D. Student: Min DaiAdvisor: Prof. Ali Keyhani
OSU Mechatronics Lab. 39
Problem Statement (1)
• Power control of grid-connected inverters– Single unit with grid-connection
• No load connection• Disconnection• Reclose
– Multiple units• Parallel operation in island mode• Connect a 2nd unit to a grid-connected unit• Disconnections• Reclose
OSU Mechatronics Lab. 40
System Description (1)
• Power circuit topology– Single unit (3-ph 3 wire inverter plus transformer)
OSU Mechatronics Lab. 41
System Description (2)
• Power circuit topology– Multiple
units
OSU Mechatronics Lab. 42
PQ Control with Grid-Connection (1)
• Real & reactive power control in grid-connected mode
• Operating scenarios-single unit– Connection with no load:
• Synchronization before closing the switch
• Control strategy change while closing the switch
• Insignificant transient due to low inverter output impedance
InverterUnit
GridLocalLoad
jX
OSU Mechatronics Lab. 43
PQ Control with Grid-Connection (2)
InverterUnit
GridLocalLoad
jX– Disconnection with load:
• Three phase opened one by one due to zero-cross switching
– Reclose• Synchronization• Control strategy
switch• P,Q transient exists
due to non-zero X
InverterUnit
GridLocalLoad
jX
OSU Mechatronics Lab. 44
PQ Control with Grid-Connection (3)
• Operating scenarios-multiple units– Parallel in island mode:
• Synchronization• Load sharing after
connection• Harmonic load sharing
– Connect a 2nd unit• Synchronization• Load sharing operation
InverterUnit 1
GridLocalLoad
jX
InverterUnit 2
InverterUnit 1
GridLocalLoad
jX
InverterUnit 2
OSU Mechatronics Lab. 45
PQ Control with Grid-Connection (4)
– Disconnections with load:
• Three phase opened one by one due to zero-cross switching
InverterUnit 1
GridLocalLoad
jX
InverterUnit 2
InverterUnit 1
GridLocalLoad
jX
InverterUnit 2
OSU Mechatronics Lab. 46
PQ Control with Grid-Connection (5)
– Reclose• Synchronization in
parallel• Control strategy
switch• Fundamental and
harmonic load sharing before and after switching
• P,Q transient exists due to non-zero X
InverterUnit 1
GridLocalLoad
jX
InverterUnit 2
OSU Mechatronics Lab. 47
PQ Control with Grid-Connection (6)
• Control goals– Low steady state PQ tracking error– Relatively fast transient response– Steady state decoupling between P and Q
controls– Small PQ transient at reclose– Load sharing in island mode including harmonic
load sharing– Load sharing in grid-connected mode– As less control interconnections between the units
as possible
OSU Mechatronics Lab. 48
Mechatronics in Automotive Systems Mechatronics in Automotive Systems Embedded DSP/microcontrollersEmbedded DSP/microcontrollers
Active suspensioncontrol
Thermal managementsystem control
Electric motor drive controlin hybrid electric car
IC Engine control
Power steering andtraction control
Adaptive comfortcontrol :heat,ventilatiion,air condition
Active noise cancellation
OSU Mechatronics Lab. 49
Brake-By-Wire
• Brake-by-wire does everything:– Braking– ABS – Antilock brake system– Brake power assisting– Vehicle stability enhancement control– Parking brake control– Tunable pedal feeling
OSU Mechatronics Lab. 50
Application of Embedded System to Brake-By-Wire
• Plug-in modules for Brake-By-Wire
OSU Mechatronics Lab. 51
Application of Embedded System to Brake-By-Wire• EMB: Electromechanical Brake Actuators• BBWM: Brake-By-Wire Manager
OSU Mechatronics Lab. 52
Application of Embedded System to Brake-By-Wire• System structure
DSP based Controller
Motor Gear and Screw
Caliper
Force Sensor
TV FclFd
Position Sensor
OSU Mechatronics Lab. 53
Application of Embedded System to Brake-By-Wire
• Electromechanically actuated disk brake by ITT Automotive
OSU Mechatronics Lab. 54
Application of Embedded System to Brake-By-Wire• Control of brake-by-wire system
– Four-quadrant operation of servo-motor– Desired clamping force response– Torque ripple minimization– Elimination of rotor position sensor– Elimination of clamping force sensor– Fail-safe operation
OSU Mechatronics Lab. 55
Steer-By-Wire
• Not just electrically assisted power steering
• Steer-by-wire comes in two flavors:– Front steer– Rear wheels
• Cars with steer-by-wire may not even have a driver’s wheel
OSU Mechatronics Lab. 56
Application of Embedded System to Steer-By-Wire
• Only wires may relay signals from a car’s steering wheel to its front wheels in a front steer-by-wire system. And an electrically actuated motor, not a mechanical link with the steering wheel, turns the front wheel.
OSU Mechatronics Lab. 57
Research @ OSU• Experimental setup
OSU Mechatronics Lab.
Modeling and Control of SRM in Electromechanic Brake
System
Wenzhe Lu, Ph.D. studentAli Keyhani, Advisor
OSU Mechatronics Lab. 59
Project Goal and Objective• Overall project goal:
Develop and implement a low-cost drive system consisting of a sensorless switched reluctance motor (SRM) with power converter and controller.
• Objectives:1. Four-quadrant operation2. Clamping force response3. Torque ripple minimization4. Elimination of rotor position sensor5. Elimination of clamping force sensor6. High performance, low cost7. Fault tolerance
OSU Mechatronics Lab. 60
Inductance Model of SRM for Standstill and Online Operation
(a) Standstill (b) Online Operation
)k(v)1k(X)(C)1k(Y)k(w)k(u)(B)k(X)(A)1k(X
s
ss
++⋅θ=++⋅θ+⋅θ=+&
X=[i1, (,i2)] Y=[i] u=[V] θs=[R, L (,Rd, Ld)]w: process noise v: measurement noise
OSU Mechatronics Lab. 61
Parameter Identification of SRM Using Maximum Likelihood
Maximum Likelihood Estimation
phase winding to be estimated
phase winding model
maximum likelihood estimation
input u output y
estimated output
errorestimated
parameter R, L
+_
OSU Mechatronics Lab. 62
Experimental Setup with dSPACE DS1103
OSU Mechatronics Lab. 63
Parameter Estimation ResultsInductance and Flux Linkage
OSU Mechatronics Lab. 64
Sensorless ControlAn inductance model of SRM can be used to design sensorless controller for full speed range.When speed is above 20RPM, a sliding mode observer based sensorless controller can yield satisfactory results.At near zero speeds is, the turn-on position for next phase and the turn-off position for current phase can be determined by matching inductances.At standstill, rotor position can be estimated by applying voltage pulses and comparing peak current values.
I
I
ControllerController SRMSRM
SRMmodelSRMmodel
ObserverObserver
V
+_
θ ω
OSU Mechatronics Lab. 65
4-Q Torque and Force Control in EMB System
System structure
Four-quadrant operationForce control and torque ripple minimizationSensorless operation (no rotor position sensors)
switchingsignal
SRMSRMControllerController InverterInverter PlantPlant
V,I T,θFFcmd
ObserverObserverθ, ω
OSU Mechatronics Lab. 66
4-Q Torque and Force Control in EMB System
Two control loops + torque controlOuter loop: force control (PID)
Inner loop: current control (Hysteresis)
Force control (PID)
Torque control
Current control (Hysteresis)
Tcmd Icmd
VSRM and Brake system
Fcmd
Fcl
I
ObserverObserverθ, ω
OSU Mechatronics Lab. 67
4-Q Torque and Force Control in EMB – Simulation Results
0 0.05 0.1 0.15 0.2 0.250
500
1000
1500
2000
2500Force (N)
time (s )
(a) Clamping Force Response
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4-40
-20
0
20
40
60
80
100
120
140
160
T (Nm)
w (rad/s )
(b) Torque-Speed Curve
OSU Mechatronics Lab. 68
Conclusions
• Four Ph.D students will graduate in 2005.• We could continue our research if industry
could support us.• Our support has gone to zero since 2001.• Outsourcing?• Could we do some of your outsourcing
Research Projects?
top related