Quantum Control Theory; The basics
Transcript of Quantum Control Theory; The basics
![Page 1: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/1.jpg)
Quantum Control Theory;
Naoki Yamamoto
Keio Univ.
The basics
![Page 2: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/2.jpg)
Contents
Classical control theory
Quantum control theory --- Continuous meas.
![Page 3: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/3.jpg)
S
C
1. C control (1) : Various systems and purposes
SC
S
M
C
A
S
M
C
A
Signal (c-number)
![Page 4: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/4.jpg)
Thus choosing
Linear system
Feedback control law :
Design
(e.g.)is unstable when
, the system becomesVia the FB law
Target value :
1. C control (2) : Linear feedback control --- stabilization
so that the closed-loop
system is stable.
;
, the system is stabilized ;
![Page 5: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/5.jpg)
Linear system
Control purpose :
is a function ofSuppose the inputS
C
FB control law that minimizes the above cost function is given by :
Target trajectory :
1. C control (3) : Optimal linear feedback control
![Page 6: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/6.jpg)
A nonlinear system
Control purpose :
is a function ofSuppose
S
M
C
A
Set a non-negative function
FB control law yields
Thus always decreases in time.
In particular, we have
1. C control (4) : Nonlinear FB control --- Lyapunov method
then we have
![Page 7: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/7.jpg)
S
M
C
A
Stochastic system
Before discussing how to control…
White noise approximation:
Noise added in
Dynamics in = stochastic differential equation (SDE)
1. C control (5) : Stochastic FB control
= Wiener increment
Formally, and subjected to hence
![Page 8: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/8.jpg)
S
M
C
A
It would be a good idea to use the
estimate value of , say , and
design an estimate-based FB control
Need the conditional probability :
Actually,
1. C control (5) : Stochastic FB control
Stochastic system
We want to have the quantum version of this scheme.
![Page 9: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/9.jpg)
(i.e., )
Classical conditional probability : Dice as an example
Prob. distribution conditioned on
the result of “even”
2. Q control (1) : Prelimi (i) Conditional prob. & Measurement
Prob. distribution conditioned on
the result of “odd”
![Page 10: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/10.jpg)
Represent using quantum mechanics
state
observable
Projection
onto “even”
Projection
onto “odd”
2. Q control (1) : Prelimi (i) Conditional prob. & Measurement
![Page 11: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/11.jpg)
State reduction = conditional probability
even
odd
2. Q control (1) : Prelimi (i) Conditional prob. & Measurement
![Page 12: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/12.jpg)
State preparation and interaction
Projection measurement on the ancilla
Output probability
--- state reduction
2. Q control (1) : Prelimi (ii) Generalized measurement
![Page 13: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/13.jpg)
systemfield
Interaction with the vacuum field
Quantum white noise
Quantum Wiener increment
Formally, and the output prob. of
Interaction Hamiltonian in
2. Q control (3) : Continuous meas. --- field as an ancilla
is
![Page 14: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/14.jpg)
systemfield
interaction
2. Q control (3) : Continuous meas. --- interaction
Interaction Unitary in
![Page 15: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/15.jpg)
systemfield
Homodyne meas. of
the field
Projection onto the quadrature basis
(instantaneous value)
(increment value)
State reduction ; The unnormalized ket vector is given by
Recall:
2. Q control (3) : Continuous meas. --- projection meas.
=
![Page 16: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/16.jpg)
Recall:Output probability
is subjected to
Summary : time evolution of the unnormalized ket vector
2. Q control (3) : Continuous meas. --- output probability
State reduction ; The unnormalized ket vector is:
![Page 17: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/17.jpg)
Time evolution of the normalized ket vector = SSE
Time evolution of the density operator = SME
2. Q control (4) : Stochastic Schrodinger and Master Eqs.
![Page 18: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/18.jpg)
Time evolution of the conditional
probability density
S
M
C
A
Stochastic system
2. Q control (5) : General framework (classical)
:
Via a state-dependent FB control
We aim to attain a desirable state convergence:
![Page 19: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/19.jpg)
2. Q control (5) : General framework (quantum)
Via a state-dependent FB control
We aim to attain a desirable state convergence :
Time evolution of the cond. density operator:
system
field
A
![Page 20: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/20.jpg)
S
M
C
A
Classical stochastic system (SDE):
Estimation-based FB controller can be
a solution to e.g. some optimal control problem:
2. Q control (5) : General framework (C, Heisenberg pic.)
given by the following filtering equationTime evolution of the estimate of is
(use )
![Page 21: Quantum Control Theory; The basics](https://reader031.fdocuments.us/reader031/viewer/2022022723/621ab2351a4a7252db6f8496/html5/thumbnails/21.jpg)
given by the following filtering equation
(use )
S
M
C
A
Quantum stochastic system (QSDE):
Time evolution of the estimate of
2. Q control (5) : General framework (Q, Heisenberg pic.)
is
Estimation-based FB controller can be
a solution to e.g. some optimal control problem: