Presentation by CZH
0 likes242 views
This document summarizes Zhaohui Cen's research work in four areas: 1) Test, diagnosis and reliability of electronics and renewable energy systems, 2) Fault prognosis and recovery for aircrafts and UAVs, 3) Quadrotor UAV and thrust-vectoring UAV modeling and control, and 4) Applications of neural networks in modeling and fault identification. It provides relevant publications, deliverables, and figures for each research area.
![1. Test, diagnosis and reliability for Electronics, Renewable energy systems and Smart-grid
Component
failures
Related deliverables:
[1] Z. Cen. Condition Parameter Estimation for Buck Converter based on
Model Observer. IEEE Transactions on Industrial Electronics. 2015. (in
revision)
[2]Z. Cen, Abdelkader Bousselham. Fault diagnosis for Photo-Voltaic
Power Converter based on Model Observers. The International
Conference on Advances in Computing, Communication and Information
Technology -CCIT 2014, London, UK, June 2014. (Best Paper Award)
[3]Y. Cao, Z. Cen and J. Wei, "FDSAC-SPICE: Fault diagnosis software for
analog circuit based on SPICE simulation," in International Conference on
Space Information Technology 2009.
Registered Software:
Wei Jiaolong, Cen Zhaohui, JIang rui, Fault simulation software for
aircraft attitude control system. (Registered No: 2010SR004895).
Zhaohui Cen2015/3/18 1
Fig. 1-1 faulty electronic components and typical failures
Fig. 1-2 ATE environment hardware for eletronics test and
diagnosis
Fig. 1-3 “Fault doctor” software operation panel
Fig. 1-4 Diagnosis reasoning-logic procedure Fig. 1-5 Fault-tree analysis
Fig. 1-6 self-diagnosis and condition parameter estimation for power
electronics devices
Fig. 1-7 experiment platform for power electronics based on NI compact
RIO and labview](https://image.slidesharecdn.com/7f83a8d9-805f-4952-92e4-91973018a59a-151215084805/85/Presentation-by-CZH-1-320.jpg)
![2. Fault prognosis and recovery for Aerocrafts and Unmanned Aerial Vehicles
Fig.2-4 diagram of Satellite Attitude Control system
Disturbance
Controller
Reaction
Wheel
Attitude
Dynamics Model
Attitude
Determine
Model
Attitude
Sensors
Model
Attitude Motion
Model
- +
Refer
Attitude
Fault
inu outu
Selected deliverables:
[1] Z.Cen, H.Noura, T.Bagus, Al Younes. Robust Fault Diagnosis for Quadrotor
UAVs Using Adaptive Thau Observer*J+ , Journal of Intelligent & Robotic Systems,
January 2014, Volume 73, Issue 1-4, pp 573-588.
[2] Z. Cen, J. Wei and R. Jiang, A Grey-Box Neural Network based Model
Identification and Fault Estimation Scheme for Nonlinear Dynamic Systems*J+,
International journal of neural system. 23(6), 2013, 1350025. (Currently IF=6.056
and a rank of 3 out of 114 in the Computer Science and Artificial Intelligence
category).
[3] J. Wei, Z. Cen and R. Jiang. A sensor fault-tolerant observer for satellite attitude
control*J+. Journal of System Engineering and Electronics. 2012. Vol. 23, No. 1,
February 2012, pp.99–107.
Patents:
[1]Wei Jiaolong, Cen Zhaohui, JIang rui, Fault tolerant observing method of
sensor for satellite attitude control system. (Authorized No: ZL200910060816.2).
[2]Wei Jiaolong, Cen Zhaohui, JIang rui, Blind system fault detection and isolation
method for real-time signal processing of spacecraft. (Authorized No:
ZL200910272265.6).
Zhaohui Cen2015/3/18 2
Fig. 2-1 FDD methods utilized in my research
Fig. 2-2 Studied satellite prototype
Fig. 2-3 Hardware-in-Loop Simulation Environment
Fig. 2-5 studied Quad-rotor UAV
Fig. 2-6 studied Hexi-rotor UAV](https://image.slidesharecdn.com/7f83a8d9-805f-4952-92e4-91973018a59a-151215084805/85/Presentation-by-CZH-2-320.jpg)
![3. Quadrotor UAV and Thrust-Vectoring UAV aerodynamics modeling and control
z zF T
y yF T
x xF T
a
e
r
Ty
Tz
T
F
,n p
,m q
,l r
6DOFNonlinearaircraft
Controlallocation
Fastdynamics(innerloop)
Fastloopcontroller
Slowdynamics(outerloop)
Slowerloopcontroller
slowerdynamics
(outererloop)
Slowerloopcontroller
Navigationdynamics
(outerestloop)
Navigationloopcontroller
+
+
+
+
+
+
+
+
+
+
+
Thrust Vector control+
V
yz
V
pqr
a
e
r
Ty
Tz
dp
dq
dr
cp
cq
cr
pq r
d
d
d
-
-
-
-
-
--
-
-
-
c
c
c
dV
d
d
cV
c
c
V
dy
dz
yz
cy
cz
-
-
cT
-200
0
200
400
600
800-40 -20 0 20
-80
-60
-40
-20
0
20
Y (m)
X (m)
Z(m)
-50
0
50
100
150
200
250 -20
0
20
-20
0
20
40
Y (m)
X (m)
Z(m)
Related deliverables:
[1] Z.Cen, H.Noura, Al Younes. Systematic Fault Tolerant Control based on
Adaptive Thau Observer Estimation for Quadrotor UAVs *J+ ,International
Journal of Applied Mathematics and Computer Science (AMCS), 2015, Vol. 25,
No. 1.
[2]Z. Cen, Tim Smith, Paul Stewart and Jill Stewart. Integrated flight/thrust
vectoring control for jet-powered unmanned aerial vehicles with ACHEON
propulsion *J+. Proceedings of the Institution of Mechanical Engineers, Part G:
Journal of Aerospace Engineering, July, 2014, DOI: 10.1177/0954410014544179.
Zhaohui Cen2015/3/18 3
Fig. 3-1 Quad-rotor UAV in hovering Fig. 3-2 Studied TV-UAV
Fig. 3-3 kinetics and dynamics of
Quad-rotors
Fig. 3-4 kinetics and dynamics of
Thrust-vectoring Fixed-wing aircrafts
Fig. 3-5 Proposed full Position Controller for TV-UAV
Fig. 3-6 Trajectory of UAVs under position control
Fig. 3-7 High-Attack-Angle control for TV-UAV
Fig. 3-8 Velocity-Vector-Roll control for TV-UAV](https://image.slidesharecdn.com/7f83a8d9-805f-4952-92e4-91973018a59a-151215084805/85/Presentation-by-CZH-3-320.jpg)
![4. Neural Networks and its applications in modeling and fault identification
0 200 400 600 800 1000
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
t(s)
resdual
tuned SPNN
expected
GBNNM
tuned SNN
tuned RNN
tuned PNN
0 200 400 600 800 1000
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
t(s)
PartialLOEfaultprameter(N.m)
estimation of IESONN
estimation of ESO
real fault value
d=0.1
d=0.2
d=0.15
INDEX SNN RNN PNN SPNN
GBNN
M
expected
R2 98.84% -21.25 46.98% 97.46% 99.99% 100%
RMSE 3.95e-2 1.73 2.692e-1 5.85e-2 2.7e-3 0
Selected deliverables:
[1] Z. Cen, J. Wei and R. Jiang, A Grey-Box Neural Network based Model
Identification and Fault Estimation Scheme for Nonlinear Dynamic Systems*J+,
International journal of neural systems. 23(6), 2013, 1350025. (Currently IF=6.056
and a rank of 3 out of 114 in the Computer Science and Artificial Intelligence
category).
[2]Z. Cen, J. Wei, R. Jiang, and X. Liu. Application of Mallat wavelet fast transforms
and IDRNN in real-time fault detection and identification for satellites *J+, Journal
of University of Science and Technology Beijing,2012. 32(1), 90-95.
[3] Z. Cen, J. Wei, R. Jiang, and X. Liu, "Real time fault diagnosis of Infrared Earth
Sensor using Elman neural network," Zhendong Ceshi Yu Zhenduan/Journal of
Vibration, Measurement and Diagnosis, vol. 30, pp. 504-509, 2010.
Patents:
[1]Wei Jiaolong, Cen Zhaohui, JIang rui, Blind system fault detection and isolation
method for real-time signal processing of spacecraft. (Authorized No:
ZL200910272265.6).
Zhaohui Cen2015/3/18 4
Fig 4-1“P2P” diagnosis strategy for general systems Fig. 4-2 Various NNs as a reference for residuals
nonlinear dynamic system
1
S2 ( )f
1( )f
3 ( )f
4 ( )f1
S
1
S
1
S2NN
1NN
3NN
4NN
1
S
1
S
Grey-box NN identification
u y
Fig. 4-3 Proposed “Grey-Box” NN concept
1
S
( , )F x u
( )x t
( , )h x x
( )x t
( )fy t
NN1
1 1( , , )g x u w
NN2
2 2( , , )g x x w
actuators
1
S
ˆ( )x t
ˆ( )x t
GBNNM
+
-
ˆ( )y t
( )r t
( )u t
Improved ESO based on Neural Network
NN1
1 1( , , )g x u w
NN2
2 2( , , )g x x w
1
S
-
2
2
ˆ ( , , )f g e
+
+
-
1
+
Estimation of
Fault parameter
GBNNM
Residual
Fig. 4-4 GBNNM application for fault estimation
Fig. 4-5 Fault identification results
Tab. 4-1 Modeling Comparison for Various NN and proposed GBNNM](https://image.slidesharecdn.com/7f83a8d9-805f-4952-92e4-91973018a59a-151215084805/85/Presentation-by-CZH-4-320.jpg)
Presentation by CZH
- 1. 1. Test, diagnosis and reliability for Electronics, Renewable energy systems and Smart-grid Component failures Related deliverables: [1] Z. Cen. Condition Parameter Estimation for Buck Converter based on Model Observer. IEEE Transactions on Industrial Electronics. 2015. (in revision) [2]Z. Cen, Abdelkader Bousselham. Fault diagnosis for Photo-Voltaic Power Converter based on Model Observers. The International Conference on Advances in Computing, Communication and Information Technology -CCIT 2014, London, UK, June 2014. (Best Paper Award) [3]Y. Cao, Z. Cen and J. Wei, "FDSAC-SPICE: Fault diagnosis software for analog circuit based on SPICE simulation," in International Conference on Space Information Technology 2009. Registered Software: Wei Jiaolong, Cen Zhaohui, JIang rui, Fault simulation software for aircraft attitude control system. (Registered No: 2010SR004895). Zhaohui Cen2015/3/18 1 Fig. 1-1 faulty electronic components and typical failures Fig. 1-2 ATE environment hardware for eletronics test and diagnosis Fig. 1-3 “Fault doctor” software operation panel Fig. 1-4 Diagnosis reasoning-logic procedure Fig. 1-5 Fault-tree analysis Fig. 1-6 self-diagnosis and condition parameter estimation for power electronics devices Fig. 1-7 experiment platform for power electronics based on NI compact RIO and labview
- 2. 2. Fault prognosis and recovery for Aerocrafts and Unmanned Aerial Vehicles Fig.2-4 diagram of Satellite Attitude Control system Disturbance Controller Reaction Wheel Attitude Dynamics Model Attitude Determine Model Attitude Sensors Model Attitude Motion Model - + Refer Attitude Fault inu outu Selected deliverables: [1] Z.Cen, H.Noura, T.Bagus, Al Younes. Robust Fault Diagnosis for Quadrotor UAVs Using Adaptive Thau Observer*J+ , Journal of Intelligent & Robotic Systems, January 2014, Volume 73, Issue 1-4, pp 573-588. [2] Z. Cen, J. Wei and R. Jiang, A Grey-Box Neural Network based Model Identification and Fault Estimation Scheme for Nonlinear Dynamic Systems*J+, International journal of neural system. 23(6), 2013, 1350025. (Currently IF=6.056 and a rank of 3 out of 114 in the Computer Science and Artificial Intelligence category). [3] J. Wei, Z. Cen and R. Jiang. A sensor fault-tolerant observer for satellite attitude control*J+. Journal of System Engineering and Electronics. 2012. Vol. 23, No. 1, February 2012, pp.99–107. Patents: [1]Wei Jiaolong, Cen Zhaohui, JIang rui, Fault tolerant observing method of sensor for satellite attitude control system. (Authorized No: ZL200910060816.2). [2]Wei Jiaolong, Cen Zhaohui, JIang rui, Blind system fault detection and isolation method for real-time signal processing of spacecraft. (Authorized No: ZL200910272265.6). Zhaohui Cen2015/3/18 2 Fig. 2-1 FDD methods utilized in my research Fig. 2-2 Studied satellite prototype Fig. 2-3 Hardware-in-Loop Simulation Environment Fig. 2-5 studied Quad-rotor UAV Fig. 2-6 studied Hexi-rotor UAV
- 3. 3. Quadrotor UAV and Thrust-Vectoring UAV aerodynamics modeling and control z zF T y yF T x xF T a e r Ty Tz T F ,n p ,m q ,l r 6DOFNonlinearaircraft Controlallocation Fastdynamics(innerloop) Fastloopcontroller Slowdynamics(outerloop) Slowerloopcontroller slowerdynamics (outererloop) Slowerloopcontroller Navigationdynamics (outerestloop) Navigationloopcontroller + + + + + + + + + + + Thrust Vector control+ V yz V pqr a e r Ty Tz dp dq dr cp cq cr pq r d d d - - - - - -- - - - c c c dV d d cV c c V dy dz yz cy cz - - cT -200 0 200 400 600 800-40 -20 0 20 -80 -60 -40 -20 0 20 Y (m) X (m) Z(m) -50 0 50 100 150 200 250 -20 0 20 -20 0 20 40 Y (m) X (m) Z(m) Related deliverables: [1] Z.Cen, H.Noura, Al Younes. Systematic Fault Tolerant Control based on Adaptive Thau Observer Estimation for Quadrotor UAVs *J+ ,International Journal of Applied Mathematics and Computer Science (AMCS), 2015, Vol. 25, No. 1. [2]Z. Cen, Tim Smith, Paul Stewart and Jill Stewart. Integrated flight/thrust vectoring control for jet-powered unmanned aerial vehicles with ACHEON propulsion *J+. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, July, 2014, DOI: 10.1177/0954410014544179. Zhaohui Cen2015/3/18 3 Fig. 3-1 Quad-rotor UAV in hovering Fig. 3-2 Studied TV-UAV Fig. 3-3 kinetics and dynamics of Quad-rotors Fig. 3-4 kinetics and dynamics of Thrust-vectoring Fixed-wing aircrafts Fig. 3-5 Proposed full Position Controller for TV-UAV Fig. 3-6 Trajectory of UAVs under position control Fig. 3-7 High-Attack-Angle control for TV-UAV Fig. 3-8 Velocity-Vector-Roll control for TV-UAV
- 4. 4. Neural Networks and its applications in modeling and fault identification 0 200 400 600 800 1000 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 t(s) resdual tuned SPNN expected GBNNM tuned SNN tuned RNN tuned PNN 0 200 400 600 800 1000 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 t(s) PartialLOEfaultprameter(N.m) estimation of IESONN estimation of ESO real fault value d=0.1 d=0.2 d=0.15 INDEX SNN RNN PNN SPNN GBNN M expected R2 98.84% -21.25 46.98% 97.46% 99.99% 100% RMSE 3.95e-2 1.73 2.692e-1 5.85e-2 2.7e-3 0 Selected deliverables: [1] Z. Cen, J. Wei and R. Jiang, A Grey-Box Neural Network based Model Identification and Fault Estimation Scheme for Nonlinear Dynamic Systems*J+, International journal of neural systems. 23(6), 2013, 1350025. (Currently IF=6.056 and a rank of 3 out of 114 in the Computer Science and Artificial Intelligence category). [2]Z. Cen, J. Wei, R. Jiang, and X. Liu. Application of Mallat wavelet fast transforms and IDRNN in real-time fault detection and identification for satellites *J+, Journal of University of Science and Technology Beijing,2012. 32(1), 90-95. [3] Z. Cen, J. Wei, R. Jiang, and X. Liu, "Real time fault diagnosis of Infrared Earth Sensor using Elman neural network," Zhendong Ceshi Yu Zhenduan/Journal of Vibration, Measurement and Diagnosis, vol. 30, pp. 504-509, 2010. Patents: [1]Wei Jiaolong, Cen Zhaohui, JIang rui, Blind system fault detection and isolation method for real-time signal processing of spacecraft. (Authorized No: ZL200910272265.6). Zhaohui Cen2015/3/18 4 Fig 4-1“P2P” diagnosis strategy for general systems Fig. 4-2 Various NNs as a reference for residuals nonlinear dynamic system 1 S2 ( )f 1( )f 3 ( )f 4 ( )f1 S 1 S 1 S2NN 1NN 3NN 4NN 1 S 1 S Grey-box NN identification u y Fig. 4-3 Proposed “Grey-Box” NN concept 1 S ( , )F x u ( )x t ( , )h x x ( )x t ( )fy t NN1 1 1( , , )g x u w NN2 2 2( , , )g x x w actuators 1 S ˆ( )x t ˆ( )x t GBNNM + - ˆ( )y t ( )r t ( )u t Improved ESO based on Neural Network NN1 1 1( , , )g x u w NN2 2 2( , , )g x x w 1 S - 2 2 ˆ ( , , )f g e + + - 1 + Estimation of Fault parameter GBNNM Residual Fig. 4-4 GBNNM application for fault estimation Fig. 4-5 Fault identification results Tab. 4-1 Modeling Comparison for Various NN and proposed GBNNM