SlideShare a Scribd company logo
1 of 79
Download to read offline
An investigation on physical limitations in vibration energy
harvesting
Rafael Rojas
Profesore guida: Antonio Carcaterra
Relatore: Aldo Sestieri
July 23, 2015
Index
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 2 of 40
Outline
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 3 of 40
Introduction
EH: Extracting environmental energy using non conventional
devices or sources .
Thermal.
Solar.
Electromagnetic.
Vibrations.
EH can be use to supply dierent power requirements.
Low power electronics , from acoustics/vibrations, solar,
electromagnetic sources: 10
−3W.
High power electronics , sea waves, wind, etc. : 10
6W.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 4 of 40
Motivation
Upper bounds to energy extraction is a key problem in history of
technology.
Optimal Control Theory (OCT) has good chances to produce
technological advances in this eld.
In fact, some adjustable parameters of the EH plant can be
requested to track the best energy storage performance, and this is
possible with actual technologies.
Our scope is to control of device-environment interaction to
maximize energy extraction.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 5 of 40
Prototype Equation
System Dynamics.
n DOF linear oscillator.
Internal dissipation.
u, controllable parameters.
Controllable dissipation term.
M¨q + C0 ˙q + Kq + C(u)˙q = f
Key Performance Index.
Running Cost: Power + Control
eort term
Final cost.
I =
T
0
˙q C(u)˙q − u Ru dt
I = I + F(T)
Objective
Control the EH process, to maximize our KPI.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 6 of 40
Traditional OCT Approaches
Pontraying Maximum Principle (1960's)
Lagrange Equations + Constrains + Maximum principle.
LQR.
LQG.
Hamilton Jacobi Bellman (1960's)
Hamilton-Jacobi equation of CV with constraints.
Direct Methods
Algebraic Optimization.
Model Predictive Control, (MPC).
An investigation on physical limitations in vibration energy harvesting July 23, 2015 7 of 40
Key Problems
Local max− min (PMP and Direct Methods).
Mixed boundary conditions.
Dimensionality Curse (HJB).
Closed-Open loop dichotomy.
Possible singularities / degeneracy (singular arcs in PMP).
An investigation on physical limitations in vibration energy harvesting July 23, 2015 8 of 40
Krotov's Method
No engineering applications developed, but this OCT method
combined with EH has advantages:
Absolute maximum energy storages is attacked.
Benchmarking of operating EH devices in actual use.
No mixed boundary conditions (we only solve i.v. ODE's).
Absolute min− max (very good).
Elimination of certain types of singularities/degeneracies (singular
arcs).
An investigation on physical limitations in vibration energy harvesting July 23, 2015 9 of 40
Outline
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 10 of 40
Krotov Iteration Method
Using an arbitrary rst guest control law u0 we will generate the
state trajectory x0 starting at z.
System
u0 x0
An investigation on physical limitations in vibration energy harvesting July 23, 2015 11 of 40
Krotov Iteration Method
Using an arbitrary rst guest control law u0 we will generate the
state trajectory x0 starting at z.
The trajectory x0 and control u0 will construct an equivalent
problem where we can easily construct a new control u1 better
than u0.
System
u0 Krotov's
Algorithm
x0 u1
An investigation on physical limitations in vibration energy harvesting July 23, 2015 11 of 40
Krotov's Method
The new control u(t, x) is constructed using an algebraic formula
that uses our bounding function:
u(t, x) = arg max
u∈U
∂ϕ(t, x)
∂x
f (t, x, u) − f 0
(t, x, u)
System
u0 Krotov's
Algorithm
x0 u1
An investigation on physical limitations in vibration energy harvesting July 23, 2015 12 of 40
Krotov's Method
The new control u(t, x) is constructed using an algebraic formula
that uses our bounding function:
u(t, x) = arg max
u∈U
∂ϕ(t, x)
∂x
f (t, x, u) − f 0
(t, x, u)
In other words, from an arbitrary control law u0 we will construct a
control u1(t, x) such that the process (x1, u1) is better than (x0, u0)
for the particular initial condition z
I(u0, z)  I(u1, z)
System
Krotov's
Algorithm
x0 ui+1ui
An investigation on physical limitations in vibration energy harvesting July 23, 2015 12 of 40
Krotov's Method
The new control u(t, x) is constructed using an algebraic formula
that uses our bounding function:
u(t, x) = arg max
u∈U
∂ϕ(t, x)
∂x
f (t, x, u) − f 0
(t, x, u)
In other words, from an arbitrary control law u0 we will construct a
control u1(t, x) such that the process (x1, u1) is better than (x0, u0)
for the particular initial condition z
I(u0, z)  I(u1, z)
This sequence is always improving and converges to the global
maximum.
System
Krotov's
Algorithm
x0 ui+1ui
An investigation on physical limitations in vibration energy harvesting July 23, 2015 12 of 40
Krotov's Method
At the end of the iterative process we will obtain a bounding
function ¯ϕ(t, x) that allows to compute the optimal controller
¯u(t, x) for the initial condition z.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 13 of 40
Variation of the Initial Conditions
Once obtained the solving bounding function ¯ϕ we wish to know
how the value of I(u, z) behaves.
Theorem: Region on Improvement of a control u0
There exist a neighbourhood of z where the bounding function
¯ϕ(t, x) allows to compute a control that is improving w.r.t. the
control u0.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 14 of 40
Constraints in the Control
The constrains in the control are common all engineering
applications.
u ∈ U(t, x)
The Krotov's method perfectly introduce time depended control
constraints by the same way that PMP does.
u = arg max
u∈U(t)
R(t, x, u)
To introduce time and state depended control constraints
u = arg max
u∈U(t,x)
R(t, x, u)
the only requirement is that U(t, x) must be wide enough to
guarantee improvements at each iteration.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 15 of 40
Constraints in the State
Also a set of constraint in the state variables can be of interest in
engineering problems
x ∈ X(t)
This kind of constraints cannot be directly imposed as the previous
case.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 16 of 40
Constraints in the State
Denition: Reachability Set
Given
˙x = f (t, x, u) u ∈ U(t, x) (*)
The reachability set is
A(x0, T, U) = {x(T) ∈ Rn : x(t) is solution of (*)}
An investigation on physical limitations in vibration energy harvesting July 23, 2015 17 of 40
Constraints in the State
We propose to solve the following problem to translate the state
constraints into control constraints:
Find ¯U(t, x) s.t. A(x0, T, ¯U) = X
Proposed Methodology
We have solved this problem in an intuitive way. Constructing upper
bounds of the variables of interest, we make an approximation of
R(t, x).
An investigation on physical limitations in vibration energy harvesting July 23, 2015 18 of 40
Outline
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 19 of 40
Piezo electric device
Mass
x
Piezo R
i
y(t)
Equation of the system:
¨x + ω2
n(x − y (t)) − g ωn ˙χ = 0
¨χ +
1
R
˙χ + g ωn ˙x = 0
Harvested Energy
KPI =
T
0
i
2
R dt
An investigation on physical limitations in vibration energy harvesting July 23, 2015 20 of 40
Piezo electric device
Mass
x
Piezo R
i
y(t)
Equation of the system:
¨x + ω2
n(x − y (t)) − g ωn ˙χ = 0
¨χ +
1
R
˙χ + g ωn ˙x = 0
Harvested Energy
KPI = max
R∈U
T
0
i
2
R dt
R ∈ [Rmin, Rmax]
Problem: Maximize KPI controlling
R
An investigation on physical limitations in vibration energy harvesting July 23, 2015 20 of 40
Piezo electric device
Mass
x
Piezo R
i
y(t)
Equation of the system:
¨x + ω2
n(x − y (t)) − g ωn ˙χ = 0
¨χ +
1
R
˙χ + g ωn ˙x = 0
Harvested Energy
KPI =
T
0
i
2
R dt
Our Control variable is
dened as u = 1
R
An investigation on physical limitations in vibration energy harvesting July 23, 2015 20 of 40
Piezo electric EH. y(t) = Rand(
ω
PDF
)
Iteration 0
As a rst control guess we use the constant value R wich maximises our KPI
t
Harvested Energy ( KPI)
t
Control u = 1/R
An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
Piezo electric EH. y(t) = Rand(
ω
PDF
)
Iteration 1
t
Harvested Energy ( KPI)
t
Control u = 1/R
An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
Piezo electric EH. y(t) = Rand(
ω
PDF
)
Iteration 10
t
Harvested Energy ( KPI)
t
Control u = 1/R
An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
Piezo electric EH. y(t) = Rand(
ω
PDF
)
Iteration 25
t
Harvested Energy ( KPI)
t
Control u = 1/R
An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
Piezo electric EH. Constrains on the state
If the harvested energy is too high the current owing on the circuit
can have an unacceptable magnitude.
ι = −
1
R
˙χ
To translate the constraints on ι to U we have considered the
following inequality
|ι| ιmax ⇐⇒ |−
1
R
˙χ| ιmax
An investigation on physical limitations in vibration energy harvesting July 23, 2015 22 of 40
Piezo electric EH. Behaviour of the Current.
Iteration 0
t
Current on the circuit
An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
Piezo electric EH. Behaviour of the Current.
Iteration 1
t
Current on the circuit
An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
Piezo electric EH. Behaviour of the Current.
Iteration 10
t
Current on the circuit
An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
Piezo electric EH. Behaviour of the Current.
Iteration 25
t
Current on the circuit
An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
Piezo electric EH. Behaviour of the Current.
Iteration
Desired Bouds
t
Current on the circuit
An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
Piezo electric EH. Constrains on the Current
We have computed a new admissible control set ¯U that limit the
current, and not deteriorate the KPI in a excessive way.
t
Current
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
Piezo electric EH. Constrains on the Current
We have computed a new admissible control set ¯U that limit the
current, and not deteriorate the KPI in a excessive way.
t
Current
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
Piezo electric EH. Constrains on the Current
We have computed a new admissible control set ¯U that limit the
current, and not deteriorate the KPI in a excessive way.
Desired Bouds
t
Current
t
Control Shape
An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
Piezo electric EH. Constrains on the Current
We have computed a new admissible control set ¯U that limit the
current, and not deteriorate the KPI in a excessive way.
Desired Bouds
t
Current
t
Control Shape
An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
Piezo Electric EH. Variation of the Initial Condition
We have computed ¯ϕ(t, x) for the specic initial condition z0 = 0
t
Displacement
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
Now we present how varies teh nbehaviour of the system for dierent
initial conditions z ∈ D
t
Displacement
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
Now we present how varies teh nbehaviour of the system for dierent
initial conditions z ∈ D
t
Displacement
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
We observe better performance becaouse z0 = 0 is a position of
minimum energy.
t
Displacement
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
We observe that after a transitory the trajectory of the system
converge to the optimal limit cycle for z0
t
Displacement
Parallel
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
We observe that after a transitory the trajectory of the system
converge to the optimal limit cycle for z0
t
Displacement
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
We observe that after a transitory the trajectory of the system
converge to the optimal limit cycle for z0
t
Displacement
Parallel
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
We observe that after a transitory the trajectory of the system
converge to the optimal limit cycle for z0
t
Displacement
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Piezo Electric EH. Variation of the Initial Condition
We observe that after a transitory the trajectory of the system
converge to the optimal limit cycle for z0
t
Displacement
Parallel
t
Harvested Energy
An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
Outline
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 26 of 40
Fluttering Wing
An investigation on physical limitations in vibration energy harvesting July 23, 2015 27 of 40
Fluttering Wing: System Dynamic
We want to maximize the energy dissipated by the generator G
m¨z + CA ˙z + KAz = LAer − FG
J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ)
Lift Force LAer =
1
2
ρU2
(t)SW CW θrel
Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t)
Generator Force FG = β ˙z + ˙θbHS cos(θ)
An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
Fluttering Wing: System Dynamic
This energy is KPI =
T
0 FGvH
m¨z + CA ˙z + KAz = LAer − FG
J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ)
Lift Force LAer =
1
2
ρU2
(t)SW CW θrel
Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t)
Generator Force FG = β ˙z + ˙θbHS cos(θ)
An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
Fluttering Wing: System Dynamic
Our Control Variables in the gain of the motor G u = β
m¨z + CA ˙z + KAz = LAer − FG
J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ)
Lift Force LAer =
1
2
ρU2
(t)SW CW θrel
Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t)
Generator Force FG = β ˙z + ˙θbHS cos(θ)
An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
Fluttering Wing: System Dynamic
KPI =
T
0
FG ˙z + ˙θbHS cos(θ)
m¨z + CA ˙z + KAz = LAer − FG
J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ)
Lift Force LAer =
1
2
ρU2
(t)SW CW θrel
Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t)
Generator Force FG = β ˙z + ˙θbHS cos(θ)
An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
Fluttering Wing: System Dynamic
max
u
T
0
u ˙z + ˙θbHS cos(θ)
2
dt
u ∈ [βmin, βmax]
m¨z + CA ˙z + KAz = LAer − FG
J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ)
Lift Force LAer =
1
2
ρU2
(t)SW CW θrel
Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t)
Generator Force FG = β ˙z + ˙θbHS cos(θ)
An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
Fluttering Wing: KPI
An investigation on physical limitations in vibration energy harvesting July 23, 2015 29 of 40
Simulation U = Const
Iteration 0
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
Simulation U = Const
Iteration 2
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
Simulation U = Const
Iteration 8
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
Simulation U = Const
Iteration 11
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
Symulation U = Const + Rand(
ω
PDF
)
Iteration 0
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
Symulation U = Const + Rand(
ω
PDF
)
Iteration 2
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
Symulation U = Const + Rand(
ω
PDF
)
Iteration 21
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
Symulation U = Const + Rand(
ω
PDF
)
Iteration 33
t
Harvested Energy ( KPI)
t
Control u = β
An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
Outline
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 32 of 40
Energy harvesting from MEMS
V
R1
i1
R2
i3i2
mass
y(t)
An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
Energy harvesting from MEMS
V
R1
i1
R2
i3i2
mass
y(t)
Power from the source V P1 = R1i1
An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
Energy harvesting from MEMS
V
R1
i1
R2
i3i2
mass
y(t)
Power from R2 P1 = R2i3
An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
Energy harvesting from MEMS
V
R1
i1
R2
i3i2
mass
y(t)
Control Variables R1 and R2
An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
System Dynamic
Liner oscillator ¨x + 2ζωn (˙x − ˙y(y)) + ω2
n(x − y(t)) =
q2
εA
Circuit ˙q =
1
R1
V −
1
R1
+
1
R2
q
(d − c)
Aε
An investigation on physical limitations in vibration energy harvesting July 23, 2015 34 of 40
System Dynamics
Power from the main source P1 = i2
1 R1.
Power from the resistor R2 P2 = i2
3 R2
Our KPI is the dierence between P1 and P2
KPI =
T
0
P2 − P1 dt
Our Control variables are the resistance R1 and R2
An investigation on physical limitations in vibration energy harvesting July 23, 2015 35 of 40
Symulation V = Const
Iteration 0
t
Harvested Energy
t
Control 1 u1 = 1/R1
Control 2 u2 = 1/R2
An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
Symulation V = Const
Iteration 1
t
Harvested Energy
t
Control 1 u1 = 1/R1
Control 2 u2 = 1/R2
An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
Symulation V = Const
Iteration 3
t
Harvested Energy
t
Control 1 u1 = 1/R1
Control 2 u2 = 1/R2
An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
Symulation V = Const
Iteration 27
t
Harvested Energy
t
Control 1 u1 = 1/R1
Control 2 u2 = 1/R2
An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
Outline
1 Introduction
2 A survey on Krotov's method
3 Application to Piezo electric EH.
4 Fluttering Wing
5 Capacitor EH.
6 Conclusions
An investigation on physical limitations in vibration energy harvesting July 23, 2015 37 of 40
Conclusion
Using a tunable parameter we have proposed a control law that
improves the harvested energy in a considerable way, w.r.t. the
classical passive approach. This is the best possible performance
for this device for a particular initial condition z.
In a neighbourhood of z these controllers are suboptimal but
improving w.r.t. an initial control u0.
We have successfully introduced constraints in both, the control
and the state variables.
We have found upper bounds to the energy harvesting process for
our particular piezo-electric model that reveals an encouraging
horizon in using OCT in EH.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 38 of 40
Perspectives
Direc Use , Krotov's methods for implementation of MPC's. This
requires
Stability.
Robustness.
Tracking accuracy.
Indirect Use Give upper bound for new proposed technologies and
benchmarking of current used technologies of EH.
An investigation on physical limitations in vibration energy harvesting July 23, 2015 39 of 40
Thank you
An investigation on physical limitations in vibration energy harvesting July 23, 2015 40 of 40

More Related Content

What's hot

Alternative architecture and control strategy july 2010 - joe beno
Alternative architecture and control strategy   july 2010 - joe benoAlternative architecture and control strategy   july 2010 - joe beno
Alternative architecture and control strategy july 2010 - joe beno
cahouser
 
[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...
[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...
[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...
Gian-Marco Rignanese
 
Kinetics kinematics
Kinetics kinematicsKinetics kinematics
Kinetics kinematics
shanth_95
 
Quantum force sensing with optomechanical transducers
Quantum force sensing with optomechanical transducersQuantum force sensing with optomechanical transducers
Quantum force sensing with optomechanical transducers
Ondrej Cernotik
 
www.gravity.psu.edu/events/conferences/inaugural/t... www.gravity.psu.edu/e...
www.gravity.psu.edu/events/conferences/inaugural/t... 	 www.gravity.psu.edu/e...www.gravity.psu.edu/events/conferences/inaugural/t... 	 www.gravity.psu.edu/e...
www.gravity.psu.edu/events/conferences/inaugural/t... www.gravity.psu.edu/e...
MedicineAndHealthCancer
 

What's hot (20)

Alternative architecture and control strategy july 2010 - joe beno
Alternative architecture and control strategy   july 2010 - joe benoAlternative architecture and control strategy   july 2010 - joe beno
Alternative architecture and control strategy july 2010 - joe beno
 
Maximum Power Extraction Method for Doubly-fed Induction Generator Wind Turbine
Maximum Power Extraction Method for Doubly-fed Induction Generator Wind TurbineMaximum Power Extraction Method for Doubly-fed Induction Generator Wind Turbine
Maximum Power Extraction Method for Doubly-fed Induction Generator Wind Turbine
 
Robust model predictive control for discrete-time fractional-order systems
Robust model predictive control for discrete-time fractional-order systemsRobust model predictive control for discrete-time fractional-order systems
Robust model predictive control for discrete-time fractional-order systems
 
Ch19
Ch19Ch19
Ch19
 
Several Pieces in Graph Theory
Several Pieces in Graph TheorySeveral Pieces in Graph Theory
Several Pieces in Graph Theory
 
[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...
[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...
[APS2020] Phonon-limited carrier mobilityin semiconductors : importance ofthe...
 
Calculating transition amplitudes by variational quantum eigensolvers
Calculating transition amplitudes by variational quantum eigensolversCalculating transition amplitudes by variational quantum eigensolvers
Calculating transition amplitudes by variational quantum eigensolvers
 
Novel approaches to optomechanical transduction
Novel approaches to optomechanical transductionNovel approaches to optomechanical transduction
Novel approaches to optomechanical transduction
 
Controlling the motion of levitated particles by coherent scattering
Controlling the motion of levitated particles by coherent scatteringControlling the motion of levitated particles by coherent scattering
Controlling the motion of levitated particles by coherent scattering
 
Real Time Code Generation for Nonlinear Model Predictive Control
Real Time Code Generation for Nonlinear Model Predictive ControlReal Time Code Generation for Nonlinear Model Predictive Control
Real Time Code Generation for Nonlinear Model Predictive Control
 
Kinetics kinematics
Kinetics kinematicsKinetics kinematics
Kinetics kinematics
 
Physics formulas list
Physics formulas listPhysics formulas list
Physics formulas list
 
Quantum force sensing with optomechanical transducers
Quantum force sensing with optomechanical transducersQuantum force sensing with optomechanical transducers
Quantum force sensing with optomechanical transducers
 
The Analytical/Numerical Relativity Interface behind Gravitational Waves: Lec...
The Analytical/Numerical Relativity Interface behind Gravitational Waves: Lec...The Analytical/Numerical Relativity Interface behind Gravitational Waves: Lec...
The Analytical/Numerical Relativity Interface behind Gravitational Waves: Lec...
 
Study of using particle swarm for optimal power flow
Study of using particle swarm for optimal power flowStudy of using particle swarm for optimal power flow
Study of using particle swarm for optimal power flow
 
Physics formula ICSE_Standard 10
Physics formula ICSE_Standard 10Physics formula ICSE_Standard 10
Physics formula ICSE_Standard 10
 
Chemical Bonding
Chemical BondingChemical Bonding
Chemical Bonding
 
www.gravity.psu.edu/events/conferences/inaugural/t... www.gravity.psu.edu/e...
www.gravity.psu.edu/events/conferences/inaugural/t... 	 www.gravity.psu.edu/e...www.gravity.psu.edu/events/conferences/inaugural/t... 	 www.gravity.psu.edu/e...
www.gravity.psu.edu/events/conferences/inaugural/t... www.gravity.psu.edu/e...
 
05 20261 real power loss reduction
05 20261 real power loss reduction05 20261 real power loss reduction
05 20261 real power loss reduction
 
Physics formulas
Physics formulasPhysics formulas
Physics formulas
 

Similar to main

MET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC Data
MET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC DataMET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC Data
MET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC Data
kuhanw
 
1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...
1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...
1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...
Aminullah Assagaf
 
Fir 05 dynamics 2-dof
Fir 05 dynamics 2-dofFir 05 dynamics 2-dof
Fir 05 dynamics 2-dof
nguyendattdh
 
Vibration energy harvesting under uncertainty
Vibration energy harvesting under uncertaintyVibration energy harvesting under uncertainty
Vibration energy harvesting under uncertainty
University of Glasgow
 
The Photoelectric Effect lab report
The Photoelectric Effect lab reportThe Photoelectric Effect lab report
The Photoelectric Effect lab report
Ethan Vanderbyl
 

Similar to main (20)

Random vibration energy harvesting
Random vibration energy harvestingRandom vibration energy harvesting
Random vibration energy harvesting
 
Hp05win 1224302948285022-9
Hp05win 1224302948285022-9Hp05win 1224302948285022-9
Hp05win 1224302948285022-9
 
Optimal Control of Electricity Production
Optimal Control of Electricity ProductionOptimal Control of Electricity Production
Optimal Control of Electricity Production
 
Voltage stability enhancement of a Transmission Line
Voltage stability  enhancement of a Transmission Line Voltage stability  enhancement of a Transmission Line
Voltage stability enhancement of a Transmission Line
 
Summary of masters work
Summary of masters workSummary of masters work
Summary of masters work
 
Eh4 energy harvesting due to random excitations and optimal design
Eh4   energy harvesting due to random excitations and optimal designEh4   energy harvesting due to random excitations and optimal design
Eh4 energy harvesting due to random excitations and optimal design
 
MET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC Data
MET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC DataMET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC Data
MET Energy Resolution in Pileup Minimum Bias Events using 7 TeV LHC Data
 
591 adamidis
591 adamidis591 adamidis
591 adamidis
 
Computational methods and vibrational properties applied to materials modeling
Computational methods and vibrational properties applied to materials modelingComputational methods and vibrational properties applied to materials modeling
Computational methods and vibrational properties applied to materials modeling
 
Integration of renewable energy sources and demand-side management into distr...
Integration of renewable energy sources and demand-side management into distr...Integration of renewable energy sources and demand-side management into distr...
Integration of renewable energy sources and demand-side management into distr...
 
1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...
1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...
1 Aminullah Assagaf_Estimation-of-domain-of-attraction-for-the-fract_2021_Non...
 
Nonnegative Matrix Factorization with Side Information for Time Series Recove...
Nonnegative Matrix Factorization with Side Information for Time Series Recove...Nonnegative Matrix Factorization with Side Information for Time Series Recove...
Nonnegative Matrix Factorization with Side Information for Time Series Recove...
 
Computational methods for nanoscale bio sensors
Computational methods for nanoscale bio sensorsComputational methods for nanoscale bio sensors
Computational methods for nanoscale bio sensors
 
Fir 05 dynamics
Fir 05 dynamicsFir 05 dynamics
Fir 05 dynamics
 
Fir 05 dynamics 2-dof
Fir 05 dynamics 2-dofFir 05 dynamics 2-dof
Fir 05 dynamics 2-dof
 
Vibration energy harvesting under uncertainty
Vibration energy harvesting under uncertaintyVibration energy harvesting under uncertainty
Vibration energy harvesting under uncertainty
 
Wireless Power Transmission for Implantable Medical Devices
Wireless Power Transmission for Implantable Medical DevicesWireless Power Transmission for Implantable Medical Devices
Wireless Power Transmission for Implantable Medical Devices
 
Circuit Network Analysis - [Chapter1] Basic Circuit Laws
Circuit Network Analysis - [Chapter1] Basic Circuit LawsCircuit Network Analysis - [Chapter1] Basic Circuit Laws
Circuit Network Analysis - [Chapter1] Basic Circuit Laws
 
The Photoelectric Effect lab report
The Photoelectric Effect lab reportThe Photoelectric Effect lab report
The Photoelectric Effect lab report
 
Linear response theory and TDDFT
Linear response theory and TDDFT Linear response theory and TDDFT
Linear response theory and TDDFT
 

main

  • 1. An investigation on physical limitations in vibration energy harvesting Rafael Rojas Profesore guida: Antonio Carcaterra Relatore: Aldo Sestieri July 23, 2015
  • 2. Index 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 2 of 40
  • 3. Outline 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 3 of 40
  • 4. Introduction EH: Extracting environmental energy using non conventional devices or sources . Thermal. Solar. Electromagnetic. Vibrations. EH can be use to supply dierent power requirements. Low power electronics , from acoustics/vibrations, solar, electromagnetic sources: 10 −3W. High power electronics , sea waves, wind, etc. : 10 6W. An investigation on physical limitations in vibration energy harvesting July 23, 2015 4 of 40
  • 5. Motivation Upper bounds to energy extraction is a key problem in history of technology. Optimal Control Theory (OCT) has good chances to produce technological advances in this eld. In fact, some adjustable parameters of the EH plant can be requested to track the best energy storage performance, and this is possible with actual technologies. Our scope is to control of device-environment interaction to maximize energy extraction. An investigation on physical limitations in vibration energy harvesting July 23, 2015 5 of 40
  • 6. Prototype Equation System Dynamics. n DOF linear oscillator. Internal dissipation. u, controllable parameters. Controllable dissipation term. M¨q + C0 ˙q + Kq + C(u)˙q = f Key Performance Index. Running Cost: Power + Control eort term Final cost. I = T 0 ˙q C(u)˙q − u Ru dt I = I + F(T) Objective Control the EH process, to maximize our KPI. An investigation on physical limitations in vibration energy harvesting July 23, 2015 6 of 40
  • 7. Traditional OCT Approaches Pontraying Maximum Principle (1960's) Lagrange Equations + Constrains + Maximum principle. LQR. LQG. Hamilton Jacobi Bellman (1960's) Hamilton-Jacobi equation of CV with constraints. Direct Methods Algebraic Optimization. Model Predictive Control, (MPC). An investigation on physical limitations in vibration energy harvesting July 23, 2015 7 of 40
  • 8. Key Problems Local max− min (PMP and Direct Methods). Mixed boundary conditions. Dimensionality Curse (HJB). Closed-Open loop dichotomy. Possible singularities / degeneracy (singular arcs in PMP). An investigation on physical limitations in vibration energy harvesting July 23, 2015 8 of 40
  • 9. Krotov's Method No engineering applications developed, but this OCT method combined with EH has advantages: Absolute maximum energy storages is attacked. Benchmarking of operating EH devices in actual use. No mixed boundary conditions (we only solve i.v. ODE's). Absolute min− max (very good). Elimination of certain types of singularities/degeneracies (singular arcs). An investigation on physical limitations in vibration energy harvesting July 23, 2015 9 of 40
  • 10. Outline 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 10 of 40
  • 11. Krotov Iteration Method Using an arbitrary rst guest control law u0 we will generate the state trajectory x0 starting at z. System u0 x0 An investigation on physical limitations in vibration energy harvesting July 23, 2015 11 of 40
  • 12. Krotov Iteration Method Using an arbitrary rst guest control law u0 we will generate the state trajectory x0 starting at z. The trajectory x0 and control u0 will construct an equivalent problem where we can easily construct a new control u1 better than u0. System u0 Krotov's Algorithm x0 u1 An investigation on physical limitations in vibration energy harvesting July 23, 2015 11 of 40
  • 13. Krotov's Method The new control u(t, x) is constructed using an algebraic formula that uses our bounding function: u(t, x) = arg max u∈U ∂ϕ(t, x) ∂x f (t, x, u) − f 0 (t, x, u) System u0 Krotov's Algorithm x0 u1 An investigation on physical limitations in vibration energy harvesting July 23, 2015 12 of 40
  • 14. Krotov's Method The new control u(t, x) is constructed using an algebraic formula that uses our bounding function: u(t, x) = arg max u∈U ∂ϕ(t, x) ∂x f (t, x, u) − f 0 (t, x, u) In other words, from an arbitrary control law u0 we will construct a control u1(t, x) such that the process (x1, u1) is better than (x0, u0) for the particular initial condition z I(u0, z) I(u1, z) System Krotov's Algorithm x0 ui+1ui An investigation on physical limitations in vibration energy harvesting July 23, 2015 12 of 40
  • 15. Krotov's Method The new control u(t, x) is constructed using an algebraic formula that uses our bounding function: u(t, x) = arg max u∈U ∂ϕ(t, x) ∂x f (t, x, u) − f 0 (t, x, u) In other words, from an arbitrary control law u0 we will construct a control u1(t, x) such that the process (x1, u1) is better than (x0, u0) for the particular initial condition z I(u0, z) I(u1, z) This sequence is always improving and converges to the global maximum. System Krotov's Algorithm x0 ui+1ui An investigation on physical limitations in vibration energy harvesting July 23, 2015 12 of 40
  • 16. Krotov's Method At the end of the iterative process we will obtain a bounding function ¯ϕ(t, x) that allows to compute the optimal controller ¯u(t, x) for the initial condition z. An investigation on physical limitations in vibration energy harvesting July 23, 2015 13 of 40
  • 17. Variation of the Initial Conditions Once obtained the solving bounding function ¯ϕ we wish to know how the value of I(u, z) behaves. Theorem: Region on Improvement of a control u0 There exist a neighbourhood of z where the bounding function ¯ϕ(t, x) allows to compute a control that is improving w.r.t. the control u0. An investigation on physical limitations in vibration energy harvesting July 23, 2015 14 of 40
  • 18. Constraints in the Control The constrains in the control are common all engineering applications. u ∈ U(t, x) The Krotov's method perfectly introduce time depended control constraints by the same way that PMP does. u = arg max u∈U(t) R(t, x, u) To introduce time and state depended control constraints u = arg max u∈U(t,x) R(t, x, u) the only requirement is that U(t, x) must be wide enough to guarantee improvements at each iteration. An investigation on physical limitations in vibration energy harvesting July 23, 2015 15 of 40
  • 19. Constraints in the State Also a set of constraint in the state variables can be of interest in engineering problems x ∈ X(t) This kind of constraints cannot be directly imposed as the previous case. An investigation on physical limitations in vibration energy harvesting July 23, 2015 16 of 40
  • 20. Constraints in the State Denition: Reachability Set Given ˙x = f (t, x, u) u ∈ U(t, x) (*) The reachability set is A(x0, T, U) = {x(T) ∈ Rn : x(t) is solution of (*)} An investigation on physical limitations in vibration energy harvesting July 23, 2015 17 of 40
  • 21. Constraints in the State We propose to solve the following problem to translate the state constraints into control constraints: Find ¯U(t, x) s.t. A(x0, T, ¯U) = X Proposed Methodology We have solved this problem in an intuitive way. Constructing upper bounds of the variables of interest, we make an approximation of R(t, x). An investigation on physical limitations in vibration energy harvesting July 23, 2015 18 of 40
  • 22. Outline 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 19 of 40
  • 23. Piezo electric device Mass x Piezo R i y(t) Equation of the system: ¨x + ω2 n(x − y (t)) − g ωn ˙χ = 0 ¨χ + 1 R ˙χ + g ωn ˙x = 0 Harvested Energy KPI = T 0 i 2 R dt An investigation on physical limitations in vibration energy harvesting July 23, 2015 20 of 40
  • 24. Piezo electric device Mass x Piezo R i y(t) Equation of the system: ¨x + ω2 n(x − y (t)) − g ωn ˙χ = 0 ¨χ + 1 R ˙χ + g ωn ˙x = 0 Harvested Energy KPI = max R∈U T 0 i 2 R dt R ∈ [Rmin, Rmax] Problem: Maximize KPI controlling R An investigation on physical limitations in vibration energy harvesting July 23, 2015 20 of 40
  • 25. Piezo electric device Mass x Piezo R i y(t) Equation of the system: ¨x + ω2 n(x − y (t)) − g ωn ˙χ = 0 ¨χ + 1 R ˙χ + g ωn ˙x = 0 Harvested Energy KPI = T 0 i 2 R dt Our Control variable is dened as u = 1 R An investigation on physical limitations in vibration energy harvesting July 23, 2015 20 of 40
  • 26. Piezo electric EH. y(t) = Rand( ω PDF ) Iteration 0 As a rst control guess we use the constant value R wich maximises our KPI t Harvested Energy ( KPI) t Control u = 1/R An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
  • 27. Piezo electric EH. y(t) = Rand( ω PDF ) Iteration 1 t Harvested Energy ( KPI) t Control u = 1/R An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
  • 28. Piezo electric EH. y(t) = Rand( ω PDF ) Iteration 10 t Harvested Energy ( KPI) t Control u = 1/R An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
  • 29. Piezo electric EH. y(t) = Rand( ω PDF ) Iteration 25 t Harvested Energy ( KPI) t Control u = 1/R An investigation on physical limitations in vibration energy harvesting July 23, 2015 21 of 40
  • 30. Piezo electric EH. Constrains on the state If the harvested energy is too high the current owing on the circuit can have an unacceptable magnitude. ι = − 1 R ˙χ To translate the constraints on ι to U we have considered the following inequality |ι| ιmax ⇐⇒ |− 1 R ˙χ| ιmax An investigation on physical limitations in vibration energy harvesting July 23, 2015 22 of 40
  • 31. Piezo electric EH. Behaviour of the Current. Iteration 0 t Current on the circuit An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
  • 32. Piezo electric EH. Behaviour of the Current. Iteration 1 t Current on the circuit An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
  • 33. Piezo electric EH. Behaviour of the Current. Iteration 10 t Current on the circuit An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
  • 34. Piezo electric EH. Behaviour of the Current. Iteration 25 t Current on the circuit An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
  • 35. Piezo electric EH. Behaviour of the Current. Iteration Desired Bouds t Current on the circuit An investigation on physical limitations in vibration energy harvesting July 23, 2015 23 of 40
  • 36. Piezo electric EH. Constrains on the Current We have computed a new admissible control set ¯U that limit the current, and not deteriorate the KPI in a excessive way. t Current t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
  • 37. Piezo electric EH. Constrains on the Current We have computed a new admissible control set ¯U that limit the current, and not deteriorate the KPI in a excessive way. t Current t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
  • 38. Piezo electric EH. Constrains on the Current We have computed a new admissible control set ¯U that limit the current, and not deteriorate the KPI in a excessive way. Desired Bouds t Current t Control Shape An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
  • 39. Piezo electric EH. Constrains on the Current We have computed a new admissible control set ¯U that limit the current, and not deteriorate the KPI in a excessive way. Desired Bouds t Current t Control Shape An investigation on physical limitations in vibration energy harvesting July 23, 2015 24 of 40
  • 40. Piezo Electric EH. Variation of the Initial Condition We have computed ¯ϕ(t, x) for the specic initial condition z0 = 0 t Displacement t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 41. Piezo Electric EH. Variation of the Initial Condition Now we present how varies teh nbehaviour of the system for dierent initial conditions z ∈ D t Displacement t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 42. Piezo Electric EH. Variation of the Initial Condition Now we present how varies teh nbehaviour of the system for dierent initial conditions z ∈ D t Displacement t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 43. Piezo Electric EH. Variation of the Initial Condition We observe better performance becaouse z0 = 0 is a position of minimum energy. t Displacement t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 44. Piezo Electric EH. Variation of the Initial Condition We observe that after a transitory the trajectory of the system converge to the optimal limit cycle for z0 t Displacement Parallel t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 45. Piezo Electric EH. Variation of the Initial Condition We observe that after a transitory the trajectory of the system converge to the optimal limit cycle for z0 t Displacement t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 46. Piezo Electric EH. Variation of the Initial Condition We observe that after a transitory the trajectory of the system converge to the optimal limit cycle for z0 t Displacement Parallel t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 47. Piezo Electric EH. Variation of the Initial Condition We observe that after a transitory the trajectory of the system converge to the optimal limit cycle for z0 t Displacement t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 48. Piezo Electric EH. Variation of the Initial Condition We observe that after a transitory the trajectory of the system converge to the optimal limit cycle for z0 t Displacement Parallel t Harvested Energy An investigation on physical limitations in vibration energy harvesting July 23, 2015 25 of 40
  • 49. Outline 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 26 of 40
  • 50. Fluttering Wing An investigation on physical limitations in vibration energy harvesting July 23, 2015 27 of 40
  • 51. Fluttering Wing: System Dynamic We want to maximize the energy dissipated by the generator G m¨z + CA ˙z + KAz = LAer − FG J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ) Lift Force LAer = 1 2 ρU2 (t)SW CW θrel Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t) Generator Force FG = β ˙z + ˙θbHS cos(θ) An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
  • 52. Fluttering Wing: System Dynamic This energy is KPI = T 0 FGvH m¨z + CA ˙z + KAz = LAer − FG J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ) Lift Force LAer = 1 2 ρU2 (t)SW CW θrel Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t) Generator Force FG = β ˙z + ˙θbHS cos(θ) An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
  • 53. Fluttering Wing: System Dynamic Our Control Variables in the gain of the motor G u = β m¨z + CA ˙z + KAz = LAer − FG J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ) Lift Force LAer = 1 2 ρU2 (t)SW CW θrel Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t) Generator Force FG = β ˙z + ˙θbHS cos(θ) An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
  • 54. Fluttering Wing: System Dynamic KPI = T 0 FG ˙z + ˙θbHS cos(θ) m¨z + CA ˙z + KAz = LAer − FG J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ) Lift Force LAer = 1 2 ρU2 (t)SW CW θrel Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t) Generator Force FG = β ˙z + ˙θbHS cos(θ) An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
  • 55. Fluttering Wing: System Dynamic max u T 0 u ˙z + ˙θbHS cos(θ) 2 dt u ∈ [βmin, βmax] m¨z + CA ˙z + KAz = LAer − FG J¨θ + CB ˙θ + KBθ = LAerbCS cos(θ) − FGbHS cos(θ) Lift Force LAer = 1 2 ρU2 (t)SW CW θrel Angle of attack θrel = −θ + atan2 ˙z − ˙θbLES, U(t) Generator Force FG = β ˙z + ˙θbHS cos(θ) An investigation on physical limitations in vibration energy harvesting July 23, 2015 28 of 40
  • 56. Fluttering Wing: KPI An investigation on physical limitations in vibration energy harvesting July 23, 2015 29 of 40
  • 57. Simulation U = Const Iteration 0 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
  • 58. Simulation U = Const Iteration 2 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
  • 59. Simulation U = Const Iteration 8 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
  • 60. Simulation U = Const Iteration 11 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 30 of 40
  • 61. Symulation U = Const + Rand( ω PDF ) Iteration 0 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
  • 62. Symulation U = Const + Rand( ω PDF ) Iteration 2 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
  • 63. Symulation U = Const + Rand( ω PDF ) Iteration 21 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
  • 64. Symulation U = Const + Rand( ω PDF ) Iteration 33 t Harvested Energy ( KPI) t Control u = β An investigation on physical limitations in vibration energy harvesting July 23, 2015 31 of 40
  • 65. Outline 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 32 of 40
  • 66. Energy harvesting from MEMS V R1 i1 R2 i3i2 mass y(t) An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
  • 67. Energy harvesting from MEMS V R1 i1 R2 i3i2 mass y(t) Power from the source V P1 = R1i1 An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
  • 68. Energy harvesting from MEMS V R1 i1 R2 i3i2 mass y(t) Power from R2 P1 = R2i3 An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
  • 69. Energy harvesting from MEMS V R1 i1 R2 i3i2 mass y(t) Control Variables R1 and R2 An investigation on physical limitations in vibration energy harvesting July 23, 2015 33 of 40
  • 70. System Dynamic Liner oscillator ¨x + 2ζωn (˙x − ˙y(y)) + ω2 n(x − y(t)) = q2 εA Circuit ˙q = 1 R1 V − 1 R1 + 1 R2 q (d − c) Aε An investigation on physical limitations in vibration energy harvesting July 23, 2015 34 of 40
  • 71. System Dynamics Power from the main source P1 = i2 1 R1. Power from the resistor R2 P2 = i2 3 R2 Our KPI is the dierence between P1 and P2 KPI = T 0 P2 − P1 dt Our Control variables are the resistance R1 and R2 An investigation on physical limitations in vibration energy harvesting July 23, 2015 35 of 40
  • 72. Symulation V = Const Iteration 0 t Harvested Energy t Control 1 u1 = 1/R1 Control 2 u2 = 1/R2 An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
  • 73. Symulation V = Const Iteration 1 t Harvested Energy t Control 1 u1 = 1/R1 Control 2 u2 = 1/R2 An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
  • 74. Symulation V = Const Iteration 3 t Harvested Energy t Control 1 u1 = 1/R1 Control 2 u2 = 1/R2 An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
  • 75. Symulation V = Const Iteration 27 t Harvested Energy t Control 1 u1 = 1/R1 Control 2 u2 = 1/R2 An investigation on physical limitations in vibration energy harvesting July 23, 2015 36 of 40
  • 76. Outline 1 Introduction 2 A survey on Krotov's method 3 Application to Piezo electric EH. 4 Fluttering Wing 5 Capacitor EH. 6 Conclusions An investigation on physical limitations in vibration energy harvesting July 23, 2015 37 of 40
  • 77. Conclusion Using a tunable parameter we have proposed a control law that improves the harvested energy in a considerable way, w.r.t. the classical passive approach. This is the best possible performance for this device for a particular initial condition z. In a neighbourhood of z these controllers are suboptimal but improving w.r.t. an initial control u0. We have successfully introduced constraints in both, the control and the state variables. We have found upper bounds to the energy harvesting process for our particular piezo-electric model that reveals an encouraging horizon in using OCT in EH. An investigation on physical limitations in vibration energy harvesting July 23, 2015 38 of 40
  • 78. Perspectives Direc Use , Krotov's methods for implementation of MPC's. This requires Stability. Robustness. Tracking accuracy. Indirect Use Give upper bound for new proposed technologies and benchmarking of current used technologies of EH. An investigation on physical limitations in vibration energy harvesting July 23, 2015 39 of 40
  • 79. Thank you An investigation on physical limitations in vibration energy harvesting July 23, 2015 40 of 40