Daniel E. Miller

ORCID: 0000-0002-0223-9483
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Stability and Control of Uncertain Systems
  • Adaptive Control of Nonlinear Systems
  • Advanced Control Systems Optimization
  • Control Systems and Identification
  • Stability and Controllability of Differential Equations
  • Iterative Learning Control Systems
  • Control and Stability of Dynamical Systems
  • Fault Detection and Control Systems
  • Solid State Laser Technologies
  • Model Reduction and Neural Networks
  • Advanced Fiber Laser Technologies
  • Cell Image Analysis Techniques
  • Laser Design and Applications
  • Laser-Matter Interactions and Applications
  • Advanced Fluorescence Microscopy Techniques
  • Advanced Control Systems Design
  • Extremum Seeking Control Systems
  • Neural Networks Stability and Synchronization
  • Power System Optimization and Stability
  • Nonlinear Dynamics and Pattern Formation
  • Analog and Mixed-Signal Circuit Design
  • Economic theories and models
  • Stochastic processes and financial applications
  • Matrix Theory and Algorithms
  • Quantum chaos and dynamical systems

University of Waterloo
2015-2024

MIT Lincoln Laboratory
2010-2017

Massachusetts Institute of Technology
2011-2017

Texas A&M University
2010-2016

Cincinnati Children's Hospital Medical Center
2016

Changchun Institute of Optics, Fine Mechanics and Physics
2013

K Lab (United States)
2012

Mitchell Institute
2010

University of Toronto
1987-2003

La Roche College
1983

A model reference adaptive control problem is posed. In the problem, objective not usual one of forcing error between plant output and asymptotically to zero, but instead, it that this be less than a (arbitrarily small) prespecified constant after short) period time, with upper bound on amount overshoot. It shown achieve goal for stabilizable detectable, single-input single-output linear time-invariant (LTI) plant, necessary sufficient minimum phase. Knowledge an order, relative degree, sign...

10.1109/9.62269 article EN IEEE Transactions on Automatic Control 1991-01-01

Handling delays in control systems is difficult and of long-standing interest. It well known that, given a finite-dimensional linear time-invariant (FDLTI) plant controller forming strictly proper stable feedback connection, closed-loop stability will be maintained under small delay the loop, although most closed loop become unstable for large delays. One previously unsolved fundamental problem this context whether, FDLTI plant, an arbitrarily margin can achieved using LTI control. Here, we...

10.1109/tac.2007.900824 article EN IEEE Transactions on Automatic Control 2007-07-01

An adaptive controller that can provide exponential Lyapunov stability for an unknown linear time-invariant (LTI) system is presented. The only required a priori information about the plant order of LTI stabilizing compensator be known, although this reduced to assuming stabilizable and detectable at expense using more complicated controller. This result extends work M. Fu B.R. Barmish (see ibid., vol.AC-31, p.1097-1103, Dec. 1986) in which it shown there exists provides if assumed upper...

10.1109/9.24228 article EN IEEE Transactions on Automatic Control 1989-06-01

A self-tuning regulator is presented which consists of a multivariable integral compensator together with tuning mechanism tunes the parameters this compensator. It shown that if reference and disturbance are constant signals, then controller plant states remain bounded for all t, asymptotic error regulation occurs. also can be modified slightly so it noise-tolerant. number simulation examples induced to illustrate behaviour when applied unknown plants.< <ETX...

10.1109/9.24203 article EN IEEE Transactions on Automatic Control 1989-05-01

We review recent progress in pulsed lasers using cryogenically-cooled Yb3+-doped gain media, with an emphasis on high average power. Recent measurements of thermo-optic properties for various host materials at both room and cryogenic temperature are presented, including thermal conductivity, coefficient expansion refractive index. Host reviewed include Y2O3, Lu2O3, Sc2O3, YLF, YSO, GSAG YVO4. report the performance several Yb operating 5-kHz pulse repetition frequency (PRF). A Q-switched...

10.1364/ome.1.000434 article EN Optical Materials Express 2011-06-24

We consider the use of linear time-varying controllers for simultaneous stabilization and performance. prove that every finite set plants, we can design a controller which provides not only closed loop stability, but also near optimal LQR type By adding mild nonlinearity in form nonlinear delay, end up with an improved performance each possible model as well case when plant changes occasionally.

10.1109/tac.2002.805687 article EN IEEE Transactions on Automatic Control 2002-12-01

We consider the use of linear time-varying controllers for simultaneous stabilization and performance. prove that every finite set plants, we can design a controller which provides not only closed loop stability, but also near optimal LQR type By adding mild nonlinearity in form nonlinear delay, end up with an improved performance each possible model as well case when plant changes occasionally.

10.1109/9.956050 article EN IEEE Transactions on Automatic Control 2001-01-01

Recently the use of a linear periodic controller has been proposed to solve model reference adaptive control problem. The resulting can handle rapid changes in plant parameters, and it provide nice transient behavior with arbitrarily good steady-state tracking using signal which remains modest size. However, also some undesireable features: i) sampled-data achieves performance by small sampling period, results large gains correspondingly poor noise tolerance, ii) rapidly varying is used, may...

10.1109/tac.2010.2042983 article EN IEEE Transactions on Automatic Control 2010-02-17

In classical model reference adaptive control, the goal is to design a controller make closed-loop system act like prespecified in face of significant plant uncertainty. Typically, consists an identifier (or tuner) which used adjust parameters linear time-invariant (LTI) compensator, and under suitable assumptions on uncertainty it proven that asymptotic matching achieved. However, typically highly nonlinear, closed loop can exhibit undesirable behavior, such as large transients or control...

10.1109/tac.2003.811251 article EN IEEE Transactions on Automatic Control 2003-05-01

Connectomics is the study of full connection matrix brain. Recent advances in high-throughput, high-resolution 3D microscopy methods have enabled imaging whole small animal brains at a sub-micrometer resolution, potentially opening road to full-blown connectomics research. One first such instruments achieve whole-brain-scale resolution Knife-Edge Scanning Microscope (KESM). KESM whole-brain data sets now include Golgi (neuronal circuits), Nissl (soma distribution), and India ink (vascular...

10.3389/fninf.2011.00029 article EN cc-by Frontiers in Neuroinformatics 2011-01-01

In this article, we consider the problem of tracking for a discrete-time plant with unknown parameters; assume knowledge an upper bound on order, and each admissible order compact set in which parameters lie. We carry out parameter estimation associated auxiliary model; indeed, dimension, cover by finite number convex sets use original-projection-algorithm-based estimator set. At point time, employ switching algorithm to determine model estimates are used pole-placement-based control law....

10.1109/tac.2021.3052745 article EN IEEE Transactions on Automatic Control 2021-02-25

Here we consider the problem of providing near optimal performance (in this context, "near performance" means as close to optimality desired) for a large set possible models. We adopt linear quadratic regulator (LQR) framework in single-input-single-output (SISO) setting, and prove that given compact controllable observable plant models fixed order, can construct single periodic controller (LPC) which provides LQR performance. Since is linear, it automatically has nice feature there some...

10.1109/tac.2006.880961 article EN IEEE Transactions on Automatic Control 2006-09-01

Classical discrete-time adaptive controllers provide asymptotic stabilization. While the original did not handle noise or unmodelled dynamics well, redesigned versions were proven to have some tolerance; however, exponential stabilization and a bounded gain on was rarely proven. Here we consider classical pole placement controller using projection algorithm rather than commonly modifed version; impose assumption that plant parameters lie in convex, compact set parameter estimates are...

10.1109/ccta.2017.8062745 article EN 2021 IEEE Conference on Control Technology and Applications (CCTA) 2017-08-01

Handling delays in control systems is difficult and of long-standing interest. It well known that, given a finite-dimensional linear time-invariant (LTI) plant LTI stabilizing controller, closed-loop stability will be maintained under small delay the feedback loop. However, some situations there large delay, perhaps arising from slow communications link or but variable computational (e.g., system which uses image processing). While are techniques available to design controller handle no...

10.1109/tac.2005.852562 article EN IEEE Transactions on Automatic Control 2005-08-01

Handling delays and uncertain parameters in control systems is difficult of long-standing interest. In this paper, we consider the problem bounded-input bounded-output stabilizing any single-input single-output linear time-invariant controllable/observable plant with gain time delay margins as large desired; to do so, propose use a non-linear varying (NLTV) controller. Despite being NLTV, proposed controller relatively simple variation periodic; furthermore, noise tolerant, guarantees...

10.1109/tac.2014.2321037 article EN IEEE Transactions on Automatic Control 2014-04-29

Byrnes et al. (1986) showed that there is no smooth, finite-dimensional, nonlinear time-invariant (NLTI) controller which asymptotically stabilizes every stabilizable and detectable, linear (LTI) plant (with a fixed number of inputs outputs). Here we construct finite-dimensional time-varying (NLTV) does exactly that; treat both the discrete-time continuous-time cases. With p equal to one in case outputs case, first show for detectable plant, exists p-dimensional (LTV) compensator provides...

10.1109/9.299621 article EN IEEE Transactions on Automatic Control 1994-07-01

In this paper, we consider the use of periodic controllers for simultaneous stabilization and step tracking, including case when there are occasional, though persistent, plant changes. We a finite set admissible models; provide design procedure that yields controller which stabilizes each such model while providing near optimal LQR tracking. then demonstrate has facility to tolerate occasional (but persistent) switches between these models without an undue effect on performance. The is with...

10.1109/tac.2010.2051065 article EN IEEE Transactions on Automatic Control 2010-05-28

We demonstrate a 149 W Yb:LiYF(4) laser with diffraction-limited beam quality at 995 nm. The laser, Q-switched 10 kHz pulse repetition frequency, produces linearly polarized 52 ns pulses slope efficiency of 73%. combination cryogenic cooling and low (3.5%) quantum defect results in minimal thermo-optic effects high thermal efficiency. measured heat load to the cryogen is 0.15 per watt output. These show potential for significant power scaling Yb:YLF lasers excellent quality.

10.1364/ol.38.004260 article EN Optics Letters 2013-10-14

Here we consider the problem of LQR optimal control in context decentralized systems. It has been proven that if system is centrally controllable and observable graph associated with strongly connected, then there exists a linear periodic controller which provide performance as close desired to centralized performance. demonstrate, via an example, connected assumption violated can be arbitrarily bad comparison

10.1109/tac.2015.2494381 article EN IEEE Transactions on Automatic Control 2015-10-28
Coming Soon ...