betaFeedback

WorksMathematics

Dual-Mode Extremum Seeking for Diffusion Systems with Boundary-Cost Feedback

Abstract

Thermally tuned devices pose a basic control problem. A heater on one edge of a thin plate must hold a small element at an unknown optimal temperature, and the only signal available is the device's own performance. Boundary measurements of {such} an unknown performance objective can be used to tune a diffusion process without reconstructing its spatial state. This paper studies a dual-mode extremum-seeking controller for this setting, modeled as a semilinear diffusion equation on a square with nonuniform boundary actuation. The implementation contains two scalar states and exploits the leading attenuation and \rev{π/4π/4} phase \rev{lag} of the diffusion response, a property shared by every periodically heated surface. It requires neither numerical plant coefficients nor the sign of the input gain. An energy estimate first establishes exponential stability of a coupled linear diffusion and controller model. A local bound on the remaining reaction and cost nonlinearities then yields stability and disturbance estimates for the averaged system. Complete kernel and boundary-trace arguments justify a change of coordinates that compares the implemented and averaged trajectories for all time. The resulting practical stability guarantee concerns the field, input bias, and sensitivity estimate, and is unaffected by a constant cost offset under corresponding filter initialization. At fixed dither amplitude, the residual state bound decreases with the inverse square root of frequency, while the ultimate cost bound decreases with its inverse. A thermal optical-filter example illustrates the predicted trends and the conservatism of the sufficient conditions.

Introduction

Many devices perform well only within a narrow window of an internal variable that nobody measures directly. A thermally tuned optical filter is a representative case [1, 2]. Its transmission peaks at one temperature of the filter element. This temperature shifts with fabrication tolerances and ageing. The signal that is convenient to measure online is the transmitted optical power itself. The actuator, a resistive heater along one edge of the supporting plate, is not collocated with the element. The heat it releases must diffuse across the plate while part of it leaks to the surroundings. The control task is therefore to drive the heater so that the filter sits at its transmission peak, through a medium governed by a diffusion equation, using one scalar performance signal and no thermometer.

Extremum-seeking control is the natural tool for a problem of this kind. It adjusts a plant input using measured performance when the optimizing operating condition is unknown. Input dithers reveal local sensitivity, while averaging and singular perturbation provide classical analysis tools [3]. Lie-bracket approximations address a related class of oscillatory optimization laws [4]. Proportional–integral designs combine sensitivity feedback with input-bias adaptation [5, 6]. Dual-mode designs also address the input needed to sustain the operating condition [7], and phasor-based methods exploit structural information about the plant response [8].

Distributed dynamics attenuate and phase-shift the perturbation between actuation and observation, and the sensitivity estimate must account for both. Earlier applications include convection–reaction reactors [9, 10]. For diffusion actuation, probing-signal design and boundary compensation have been combined with infinite-dimensional stability analysis [11]. Extensions address multivariable PDE dynamics [12], reaction–advection–diffusion equations [13], spatially integrated outputs [14], unbiased designs [15], and physics-informed probing [16]. Broader developments appear in [17, 18]. These designs compensate the actuation dynamics online from a known model of the PDE.

This paper takes a different route, suggested by the physics of the thermal problem. When the heater power is modulated sinusoidally, the temperature at the heated edge responds as a thermal wave. Its amplitude falls with the inverse square root of the frequency and its phase lags the heat flux by π/4\pi/4. These are classical properties of diffusion at a periodically heated surface [19]. They hold whatever the conductivity, the heat loss, and the plate size, which enter only through a single amplitude factor carrying the sign of the input gain. Placing the sensor on the actuated edge therefore hands the extremum-seeking loop a response whose phase is known before the controller is switched on, and the demodulation can be fixed once and for all. These properties permit a controller with two internal states using only the measured cost. The feedback law requires no numerical PDE coefficients, state-field measurement, or online PDE compensator. Signal injection plays a related role in finite-dimensional virtual-measurement constructions [20], but boundary actuation and point observation require additional trace estimates here.

A design that uses no model of the plate must be shown to work on the plate, and the contribution is a constructive link between averaged stability and implemented performance. We first prove exponential stability of the averaged closed loop with its finite-bandwidth sensitivity filter. A longitudinal reaction contribution incorporates existing damping, and a local derivative bound controls the remaining nonlinearities. An exact harmonic lifting, spatial smoothing, and a boundary-trace argument then give trajectory closeness for all time. The resulting local practical exponential regulation has decreasing state and cost bounds as frequency increases, even at fixed dither amplitude. The comparison and its constants are developed directly, within the broader context of infinite-dimensional averaging [21] and input-to-state stability [22].

The objective is regulation toward an attainable local minimum of a boundary-output cost. The theorem is specific to the stated geometry and nonlinear margin. The thermal example illustrates controller performance and spatial regulation, while explicitly distinguishing its observed frequency trends from the conservative sufficient certificate.

Sections II–III specify the problem and controller. Averaged stability is established in Section IV. Sections V–VIII develop the lifting, regularity, comparison, and practical stability arguments. Frequency selection and the thermal filter example conclude the analysis. Readers interested mainly in the design and its performance can pass from Sections II–III directly to Sections VIII–X. Table 1 collects the constants that appear in the final estimates and says where each is defined.

SymbolMeaningDefined in
ε,ωlift\varepsilon,\omega_{\mathrm{lift}}ε=ω−1/2\varepsilon=\omega^{-1/2}; minimal lifting frequencySecs. III, V
M,σM,\sigmaovershoot and decay rate of the linear loop(20), (23)
GBG_Bgain from a boundary command to the state(24)
ℓ(R),α\ell(R),\alphanonlinear margin; α=σ−Mℓ(R)\alpha=\sigma-M\ell(R)(32)
Φ1,Φ2\Phi_1,\Phi_2bounds on φ′′\varphi'' near the optimumafter (32)
Lr,Lρ,LeL_r,L_\rho,L_elifted field, its trace, harmonic remainder(36)
KΔ,δωK_\Delta,\delta_\omegafilter correction; coordinate error(41), (42)
K0K_0L1L^1 norm of the trace kernel(61)
LzL_zweighted bound on the trace derivative(64)
KξK_\xifilter residual coefficient(76)
Rω\mathcal{R}_\omegaimplemented-vs-averaged comparison error(78)
βω,r⋆\beta_\omega,r_\starultimate bound; admissible initial radius(81)

Table 1. Main constants of the analysis and where they are defined.

Problem Formulation

Figure 1(a) shows the physical situation we have in mind. A thin square plate carries a small optical element at the midpoint of one edge. A resistive heater runs along that edge. The power it delivers may be shaped along the edge but is set by a single command uu. Heat spreads through the plate by conduction, leaks to the surroundings through the two broad faces, and the three remaining edges are insulated. The transmission of the element depends on its own temperature and peaks at a value that is not known in advance, Figure 1(b), and the only measurement is the transmitted optical power, summarized by a score JJ that is smallest at the peak. Neither the temperature field nor the temperature of the element is measured.

Figure 1. (a) Thin plate heated along one edge, with the optical element at the midpoint of that edge. The three other edges are insulated and heat leaks through the two broad faces. (b) Normalized transmission of the element and the score J=1/T−1J=1/T-1 against the detuning of the element temperature from its optimum, in units of the temperature scale Θ\Theta. Only JJ is measured.

Scaling lengths by the side of the plate, time by its diffusion time, and temperature by the width of the transmission peak turns the plate into the unit square and the temperature excess ww above ambient into the state of a semilinear diffusion equation with boundary actuation. Section X gives the details. In this dimensionless form, and allowing for a reaction term and a fixed heat load that the thermal model does not need but that cost nothing in the analysis, the plant reads as follows. Let Ω=(0,1)2\Omega=(0,1)^2, s=(s1,s2)s=(s_1,s_2), and consider

wt=κΔw−μ0w+F(s,w)+d(s),w_t=\kappa\Delta w-\mu_0w+F(s,w)+d(s),
−κws1(0,s2,t)=bq(s2)u(t),-\kappa w_{s_1}(0,s_2,t)=bq(s_2)u(t),
ws1(1,s2,t)=0,w_{s_1}(1,s_2,t)=0,
ws2(s1,0,t)=ws2(s1,1,t)=0,(1)w_{s_2}(s_1,0,t)=w_{s_2}(s_1,1,t)=0, \tag*{(1)}

where κ,μ0>0\kappa,\mu_0>0, b≠0b\ne0, u∈Ru\in\mathbb{R}, and q(s2)=1+ϑcos⁡(πs2)q(s_2)=1+\vartheta\cos(\pi s_2) with ϑ∈R\vartheta\in\mathbb{R}. The fixed source dd may be unknown. Here κ\kappa is the scaled diffusivity, μ0\mu_0 accounts for the heat lost through the faces, qq is the distribution of heater power along the actuated edge, and bb is the input gain. The sign of bb is unknown such that the design also covers cooling actuators and sensors of either polarity. Only the noise-free cost

J(t)=φ(Cbw(t)),Cbw=w(0,12),(2)J(t)=\varphi(C_b w(t)),\qquad C_b w=w\left(0,\frac{1}{2}\right), \tag*{(2)}

is measured. The field and its boundary value are not separately available.

Since the heater acts through the boundary and the initial temperature need not match the boundary condition, the analysis needs a precise notion of solution. Let X=C(Ω‾;R)\mathcal{X}=C(\overline{\Omega};\mathbb{R}) be the continuous real-valued fields, with ∥f∥∞=max⁡s∈Ω‾∣f(s)∣\lVert f\rVert_\infty=\max_{s\in\overline{\Omega}}|f(s)|. The distinct symbol Cb:X→RC_b:\mathcal{X}\to\mathbb{R} denotes point observation, so ∥Cb∥=1\lVert C_b\rVert=1. For 0<γ<10<\gamma<1, use

∥f∥Cγ=∥f∥∞+[f]γ,[f]γ=sup⁡s≠s~∣f(s)−f(s~)∣∣s−s~∣γ.\lVert f\rVert_{C^\gamma}=\lVert f\rVert_\infty+[f]_\gamma,\qquad[f]_\gamma=\sup_{s\ne\widetilde{s}}\frac{|f(s)-f(\widetilde{s})|}{|s-\widetilde{s}|^\gamma}.

Let A=κΔ−μ0A=\kappa\Delta-\mu_0 be the homogeneous Neumann generator, S(t)S(t) its semigroup, S−1,S−1(t)S_{-1},S_{-1}(t) their extrapolations. Let BB be the scalar boundary-command operator in the extrapolation space. Write KB(t)=S−1(t)B\mathcal{K}_B(t)=S_{-1}(t)B and F(x)=F(⋅,x(⋅))F(x)=F(\cdot,x(\cdot)). Solutions have the mild interpretation

w(t)=S(t−t0)w0+∫t0tS(t−r)[F(w(r))+d] drw(t)=S(t-t_0)w_0+\int_{t_0}^{t}S(t-r)[F(w(r))+d]\,\mathrm{d}r
+∫t0tKB(t−r)u(r) dr.(3)\qquad+\int_{t_0}^{t}\mathcal{K}_B(t-r)u(r)\,\mathrm{d}r. \tag*{(3)}

The kernel estimates below establish local integrability in X\mathcal{X}. This formulation permits continuous initial fields without Neumann compatibility [23].

Assumption 1. The source dd is continuous. The functions F,FwF,F_w are jointly continuous near the equilibrium field range and its prescribed enlargement. There are HF,LF<∞H_F,L_F<\infty and 0<γ<10<\gamma<1 such that

∣F(s,ζ)−F(s~,ζ~)∣≤HF∣s−s~∣γ+LF∣ζ−ζ~∣(4)|F(s,\zeta)-F(\widetilde{s},\widetilde{\zeta})| \le H_F|s-\widetilde{s}|^\gamma +L_F|\zeta-\widetilde{\zeta}| \tag*{(4)}

on that region. The cost φ\varphi is C2C^2 on an open interval containing the corresponding boundary values.

Assumption 1 asks only for Hölder continuity in space and Lipschitz continuity in temperature of the reaction, and for a twice differentiable cost. Both are mild for the thermal application, where the reaction is absent and the cost is a smooth function of the element temperature.

Assumption 2. There is a stationary pair (w⋆,u⋆)(w^\star,u^\star) of (1) whose observed value z⋆=Cbw⋆z^\star=C_b w^\star satisfies φ′(z⋆)=0\varphi'(z^\star)=0 and h:=φ′′(z⋆)>0h:=\varphi''(z^\star)>0.

In the thermal system, Assumption 2 says that some constant heater power holds the element exactly at the temperature where its transmission peaks, and that the peak is not flat. Thus the scalar output minimum is attainable at steady state. The objective is local regulation of the associated field and bias, not optimization of an arbitrary steady input-to-cost map. Neither the equilibrium nor the cost derivatives are known to the controller. For the nominal estimates, fix H0<∞H_0<\infty and require

∥w0−w⋆∥Cγ≤H0.(5)\lVert w_0-w^\star\rVert_{C^\gamma}\le H_0. \tag*{(5)}

The initial error will also be small in X\mathcal{X}. This fixed family controls the integrable startup layer of the measured trace.

The design relies on one physical observation. A sinusoidal modulation of the heater power launches a thermal wave into the plate, and at the frequencies of interest the temperature at the heated edge responds with an amplitude that decays like ω−1/2\omega^{-1/2} and a phase that lags the heat flux by π/4\pi/4, the classical signature of diffusion at a periodically heated surface [19]. Conductivity, heat loss, and plate size do not alter this leading behavior; they only set the amplitude. In the present notation, for the linear diffusion equation with F=d=0F=d=0, the midpoint transfer function is

Hb(iω)=bκμ0+iωcoth⁡μ0+iωκ.(6)H_b(i\omega)=\frac{b}{\sqrt{\kappa}\sqrt{\mu_0+i\omega}}\coth\sqrt{\frac{\mu_0+i\omega}{\kappa}}. \tag*{(6)}

Here the square roots are taken with positive real part, and the transverse cosine mode of qq does not appear because it vanishes at the midpoint. § 5 proves

Hb(iω)=χω−1/2e−iπ/4+O(ω−3/2)(7)H_b(i\omega)=\chi\omega^{-1/2}e^{-i\pi/4}+O(\omega^{-3/2}) \tag*{(7)}

where χ=bκ\chi=\frac{b}{\sqrt{\kappa}}. The controller uses the known diffusion factor after separating χ\chi. Changing the sign of bb changes the total transfer phase by π\pi, but does not change the phase of this factor. The demodulation phase is thus fixed in advance, and χ\chi is absorbed into a loop gain shown to be proportional to the positive quantity b2κ\frac{b^2}{\sqrt{\kappa}} in § 4.

Controller and Averaged Target

Fix a0>0a_0>0, a dither amplitude a∈(0,a0]a\in(0,a_0], and λ>0\lambda>0. Choose gains satisfying

0<kI<λkP.(8)0<k_I<\lambda k_P. \tag*{(8)}

Set θ=ω(t−t0)+θ0\theta=\omega(t-t_0)+\theta_0 and ε=ω−1/2\varepsilon=\omega^{-1/2}. The two-state controller is

References

  1. [1]L. Domash, M. Wu, N. Nemchuk, and E. Ma, “Tunable and switchable multiple-cavity thin film filters,” Journal of Lightwave Technology, vol. 22, no. 1, pp. 126–135, 2004.DOI
  2. [2]D. Holfeld and H. Zappe, “Thermal and optical characterization of silicon-based tunable optical thin-film filters,” Journal of Microelectromechanical Systems, vol. 16, no. 3, pp. 500–510, 2007.
  3. [3]M. Krstić and H.-H. Wang, “Stability of extremum seeking feedback for general nonlinear dynamic systems,” Automatica, vol. 36, no. 4, pp. 595–601, 2000.DOI
  4. [4]H.-B. Dürr, M. S. Stanković, C. Ebenbauer, and K. H. Johansson, “Lie bracket approximation of extremum seeking systems,” Automatica, vol. 49, no. 6, pp. 1538–1552, 2013.arxiv.org/abs/1109.6129
  5. [5]M. Guay, “A perturbation-based proportional integral extremum-seeking control approach,” IEEE Transactions on Automatic Control, vol. 61, no. 11, pp. 3370–3381, 2016.DOI
  6. [6]M. Guay and D. Dochain, “A proportional-integral extremum-seeking controller design technique,” Automatica, vol. 77, pp. 61–67, 2017.DOI
  7. [7]M. Guay and K. T. Atta, “Dual mode extremum-seeking control via Lie-bracket averaging approximations,” in 2018 Annual American Control Conference (ACC), 2018, pp. 2972–2977.DOI
  8. [8]K. T. Atta and M. Guay, “Fast proportional integral phasor extremum seeking control for a class of nonlinear system,” IFAC-PapersOnLine, vol. 50, no. 1, pp. 5724–5730, 2017.DOI
  9. [9]P. Cougnon, D. Dochain, M. Guay, and M. Perrier, “Real-time optimization of a tubular reactor with distributed feed,” AIChE Journal, vol. 52, no. 6, pp. 2120–2128, 2006.DOI
  10. [10]N. Hudon, M. Guay, M. Perrier, and D. Dochain, “Adaptive extremum-seeking control of convection-reaction distributed reactor with limited actuation,” Computers & Chemical Engineering, vol. 32, no. 12, pp. 2994–3001, 2008.DOI
  11. [11]J. Feiling, S. Koga, M. Krstić, and T. R. Oliveira, “Gradient extremum seeking for static maps with actuation dynamics governed by diffusion PDEs,” Automatica, vol. 95, pp. 197–206, 2018.DOI
  12. [12]T. R. Oliveira, J. Feiling, S. Koga, and M. Krstić, “Multivariable extremum seeking for PDE dynamic systems,” IEEE Transactions on Automatic Control, vol. 65, no. 11, pp. 4949–4956, 2020.DOI
  13. [13]——, “Extremum seeking for unknown scalar maps in cascade with a class of parabolic partial differential equations,” International Journal of Adaptive Control and Signal Processing, vol. 35, no. 7, pp. 1162–1187, 2021.DOI
  14. [14]P. H. S. Coutinho, T. R. Oliveira, and M. Krstić, “Extremum seeking control for scalar maps with distributed diffusion PDEs,” IEEE Transactions on Automatic Control, vol. 70, no. 7, pp. 4865–4872, 2025.arxiv.org/abs/2406.01564
  15. [15]C. T. Yilmaz, M. Diagne, and M. Krstić, “Unbiased extremum seeking for PDEs,” in 2024 IEEE 63rd Conference on Decision and Control (CDC), 2024, pp. 1199–1204.arxiv.org/abs/2403.16462
  16. [16]H. Guo, Z. Guo, J. Guo, and T. R. Oliveira, “Extremum seeking boundary control for PDE systems using PINNs,” IMA Journal of Mathematical Control and Information, vol. 43, no. 1, p. dnaf046, 2026.DOI
  17. [17]T. R. Oliveira and M. Krstić, Extremum Seeking Through Delays and PDEs. Philadelphia, PA, USA: SIAM, 2022.DOI
  18. [18]T. R. Oliveira, M. Krstić, and T. Başar, “Extremum and Nash equilibrium seeking with delays and PDEs: Designs & applications: Into the second century of extremum-seeking control,” IEEE Control Systems, vol. 46, no. 2, pp. 39–87, 2026.DOI
  19. [19]H. S. Carslaw and J. C. Jaeger, Conduction of Heat in Solids, 2nd ed. Oxford, UK: Oxford University Press, 1959.
  20. [20]P. Combes, A. K. Jebai, F. Malrait, P. Martin, and P. Rouchon, “Adding virtual measurements by signal injection,” in 2016 American Control Conference (ACC), 2016, pp. 999–1005.arxiv.org/abs/1510.00338
  21. [21]J. K. Hale and S. M. Verduyn Lunel, “Averaging in infinite dimensions,” Journal of Integral Equations and Applications, vol. 2, no. 4, pp. 463–494, 1990.DOI
  22. [22]A. Mironchenko and C. Prieur, “Input-to-state stability of infinite-dimensional systems: Recent results and open questions,” SIAM Review, vol. 62, no. 3, pp. 529–614, 2020.arxiv.org/abs/1910.01714
  23. [23]R. Nittka, “Inhomogeneous parabolic Neumann problems,” Czechoslovak Mathematical Journal, vol. 64, no. 3, pp. 703–742, 2014.
  24. [24]E. M. Ouhabaz, Analysis of Heat Equations on Domains, ser. London Mathematical Society Monographs. Princeton, NJ: Princeton University Press, 2005, vol. 31.DOI
  25. [25]G. H. Hardy, J. E. Littlewood, and G. Pólya, Inequalities, 2nd ed. Cambridge, UK: Cambridge University Press, 1952.

Paper details

vlnc:W_mkq9kmmb
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents