Optimization of positive linear systems via geometric programming
1. Optimization of Positive Linear Systems
via Geometric Programming
Masaki Ogura
Osaka University, Japan
1
M. Ogura, M. Kishida, and J. Lam, “Geometric programming for optimal positive linear
systems,” IEEE Transactions on Automatic Control (accepted for publication), 2020.
Shenzhen University
12/26/2019
2. Positive linear systems
Parameter Tuning Problem
Synthesis by geometric programming
(Hidden convexity)
Numerical example
Outline 2
5. Dynamical systems with positive variables
Lotka-Volterra equation
Rantzer, Valcher, “A tutorial on positive systems and large scale control,”
2018 IEEE Conference on Decision and Control, 2018.
Positive systems 5
Buffer network
13. Parametrized positive linear system
Parameter tuning problem
Problem statement (p. 8) 13
𝑑𝑥
𝑑𝑡
= 𝐴(𝜃)𝑥 + 𝐵(𝜃)𝑤
𝑦 = 𝐶 𝜃 𝑥,
Σ𝜃: 𝜃 ∈ Θ ⊂ ℝ𝑛𝜃
𝐿(𝜃)
minimize
𝐽 Σ𝜃 ≤ 𝛾
𝜃 ∈ Θ
subject to Σ𝜃 is internally stable
14. Convexity results
Assumption: Off-diagonals of 𝐴 fixed. 𝐵 and 𝐶 fixed.
Convexity of 𝑥 𝑇 [Colaneri 2014, Automatica]
Convexity of 𝐻2 or 𝐻∞ norm [Dhingra 2018, IEEE TCNS]
Don’t apply to epidemic containment (and other problems)
Research question
Related works 14
𝑑𝑥
𝑑𝑡
= 𝐴(𝜃)𝑥 + 𝐵(𝜃)𝑤
𝑦 = 𝐶 𝜃 𝑥,
Σ𝜃:
Is it possible to optimally tune the off-diagonals of 𝑨
as well as 𝑩 and 𝑪 matrices?
18. Reduction to a convex optimization
Posynomial
ℝ𝑛
ℝ+
𝑛
ℝ+
ℝ
exp log
Convex
Geometric program Convex optimization
Log transformation
18
19. Geometric program
Variable transformations: 𝑥 = 𝑒𝑡, 𝑦 = 𝑒𝑠
Equivalent convex optimization
An example
minimize 𝑥−1/3
𝑦−1/4
subject to 𝑥1/2 + 𝑦1/2 ≤ 1
k
𝑥
𝑦
minimize 𝑒
−
𝑡
3−
𝑠
4
subject to 𝑒𝑡/2
+ 𝑒𝑠/2
≤ 1
minimize −
𝑡
3
−
𝑠
4
subject to 𝑠 ≤ 2 log(1 − 𝑒𝑡/2)
𝑡
𝑠
19
20. Circuit design
Applications 20
Other fields
Boyd, Kim, Vandenberghe, Hassibi, “A tutorial on geometric programming,” Optimization and
Engineering, vol. 8, no. 1, pp. 67–127, 2007.
Geometric programming to solve parameter tuning
problem for positive linear systems?
25. Parametrized positive linear system
Parameter tuning problem
Assumption: cost function 25
𝑑𝑥
𝑑𝑡
= 𝐴(𝜃)𝑥 + 𝐵(𝜃)𝑤
𝑦 = 𝐶 𝜃 𝑥,
Σ𝜃: 𝜃 ∈ Θ ⊂ ℝ𝑛𝜃
𝐿(𝜃)
minimize
𝐽 Σ𝜃 ≤ 𝛾
𝜃 ∈ Θ
subject to Σ𝜃 is internally stable
Posynomial
26. If 𝐽 Σ𝜃 is either
𝐻2 norm;
𝐻∞ norm;
Worst-case growth rate under structural uncertainty;
Main result: An overview 26
Σ𝜃
Δ
27. Main result: An overview 27
https://www.slideserve.com/helene/introduction-to-the-hankel-based-model-order-
reduction-for-linear-systems
Hankel operator
LTI SYSTEM
X (state)
t
u
t
y
Hankel operator
Past
input
(u(t>0)=0)
Future
output
2 2
: ( ,0) (0, )
H L L
Hankel operator:
- maps past inputs to future system outputs
- ignores any system response before time 0.
- Has finite rank (connection only by the state at t=0)
- As an operator on L2 it has an induced norm (energy
amplification)!
If 𝐽 Σ𝜃 is either
𝐻2 norm;
𝐻∞ norm;
Worst-case growth rate under structural uncertainty;
Hankel norm;
Schatten 𝑝-norm;
then, the parameter tuning problem can be exactly solved by
geometric programming.
No need to fix the elements of 𝑨, 𝑩, 𝑪 matrices!
28. 35
𝐿(𝜃)
minimize
𝐽 Σ𝜃 ≤ 𝛾
𝜃 ∈ Θ
subject to Σ𝜃 is internally stable
Strategy
Reduction to a geometric program 28
Algebraic
How to rewrite into a
constraint in terms of
posynomials?
Posynomial
Original 𝐽 Σ𝜃
Linear systems theory Perron-Frobenius
Matrix theory tools
30. Maximum singular value lemma
Let 𝑀 be a nonnegative matrix and 𝛾 be a positive number. Then, the
following conditions are equivalent:
𝑀 < 𝛾
∃ positive vectors 𝑢 and 𝑣 such that 𝑀𝑢 < 𝛾𝑣 and 𝑀𝑣 < 𝛾𝑢
Matrix inversion lemma
Let 𝐹 be a Metzler matrix, 𝑔 be a nonnegative vector, and 𝐻 be a
nonnegative matrix. The following conditions are equivalent:
𝐹 is Hurwitz stable and −𝐻𝐹−1𝑔 < 𝑣
∃ positive vector 𝜒 such that 𝐻𝜒 < 𝑣 and 𝐹𝜒 + 𝑔 < 0
Lemmas 30
42. Linearlized SIS model w/ uncertainty
Cost function
𝐿 𝜃 =
𝑖=1
𝑁
(𝑓 𝛽𝑖 + 𝑔(𝛿𝑖))
Problem: maximize the worst-case decay rate of the size of
infected population geometric programming
Example: Epidemic Containment 42
𝑑𝑥
𝑑𝑡
= diag 𝛽1, … , 𝛽𝑁 (𝑨𝑮+𝚫) − diag 𝛿1, … , 𝛿𝑁 𝑥 + 𝐵𝑤
𝑦 = 𝐶𝑥,
Σ𝛽,𝛿:
43. PageRank and optimal investments
We need to invest more to prepare for worst case scenario
GP allows us to find cost-efficient medical resource allocation
Numerical simulation: results
Without uncertainty With uncertainty
43
Larger investment
45. Parameter tuning of positive linear systems
Geometric programs w/ finitely many variables
Hidden convexity
Future research directions
Switched linear systems
Discrete-time systems
Nonlinear systems
Research questions
Can we combine convexity results and log-convexity results
to better synthesize positive linear systems?
Conclusion 45