Heiko J. Zwart

Stability and Stabilization of a Class of Boundary Control Systems

We study a class of partial differential equations on a one dimensional spatial domain with control and observation at the boundary. For this class of systems we describe how to obtain an impedance energy-preserving system, as well as scattering energy-preserving system. For the first type of systems we consider (static and dynamic) feedback stabilization by means of boundary control. For the scattering energy-preserving systems we give conditions for which the system is either asymptotically or exponentially stable.

LPMLE3: A novel 1‐D approach to study water flow in streambeds using heat as a tracer

We introduce LPMLE3, a new 1-D approach to quantify vertical water flow components at streambeds using temperature data collected in different depths. LPMLE3 solves the partial differential equation for coupled water flow and heat transport in the frequency domain. Unlike other 1-D approaches it does not assume a semi-infinite halfspace with the location of the lower boundary condition approaching infinity. Instead, it uses local upper and lower boundary conditions. As such, the streambed can be divided into finite subdomains bound at the top and bottom by a temperature-time series. Information from a third temperature sensor within each subdomain is then used for parameter estimation. LPMLE3 applies a low order local polynomial to separate periodic and transient parts (including the noise contributions) of a temperature-time series and calculates the frequency response of each subdomain to a known temperature input at the streambed top. A maximum-likelihood estimator is used to estimate the vertical component of water flow, thermal diffusivity, and their uncertainties for each streambed subdomain and provides information regarding model quality. We tested the method on synthetic temperature data generated with the numerical model STRIVE and demonstrate how the vertical flow component can be quantified for field data collected in a Belgian stream. We show that by using the results in additional analyses, nonvertical flow components could be identified and by making certain assumptions they could be quantified for each subdomain. LPMLE3 performed well on both simulated and field data and can be considered a valuable addition to the existing 1-D methods.

Frequency domain sample maximum likelihood estimation for spatially dependent parameter estimation in PDEs

The identification of the spatially dependent parameters in Partial Differential Equations (PDEs) is important in both physics and control problems. A methodology is presented to identify spatially dependent parameters from spatio-temporal measurements. Local non-rational transfer functions are derived based on three local measurements allowing for a local estimate of the parameters. A sample Maximum Likelihood Estimator (SMLE) in the frequency domain is used, because it takes noise properties into account and allows for high accuracy consistent parameter estimation. Confidence bounds on the parameters are estimated based on the noise properties of the measurements. This method is successfully applied to the simulations of a finite difference model of a parabolic PDE with piecewise constant parameters.

Estimation of the thermal diffusion coefficient in fusion plasmas taking frequency measurement uncertainties into account

In this paper, the estimation of the thermal diffusivity from perturbative experiments in fusion plasmas is discussed. The measurements used to estimate the thermal diffusivity suffer from stochastic noise. Accurate estimation of the thermal diffusivity should take this into account. It will be shown that formulas found in the literature often result in a thermal diffusivity that has a bias (a difference between the estimated value and the actual value that remains even if more measurements are added) or have an unnecessarily large uncertainty. This will be shown by modeling a plasma using only diffusion as heat transport mechanism and measurement noise based on ASDEX Upgrade measurements. The Fourier coefficients of a temperature perturbation will exhibit noise from the circular complex normal distribution (CCND). Based on Fourier coefficients distributed according to a CCND, it is shown that the resulting probability density function of the thermal diffusivity is an inverse non-central chi-squared distribution. The thermal diffusivity that is found by sampling this distribution will always be biased, and averaging of multiple estimated diffusivities will not necessarily improve the estimation. Confidence bounds are constructed to illustrate the uncertainty in the diffusivity using several formulas that are equivalent in the noiseless case. Finally, a different method of averaging, that reduces the uncertainty significantly, is suggested. The methodology is also extended to the case where damping is included, and it is explained how to include the cylindrical geometry.