Showing papers by "Peter M. Young published in 1995"
TL;DR: This paper presents a readily computable formula for the real stability radius with respect to an arbitrary stability region in the complex plane.
337 citations
TL;DR: In this article, the performance characteristics of type-II InAs/InxGa1−xSb superlattices for long and very long-wave infrared detection are discussed.
Abstract: The performance characteristics of type‐II InAs/InxGa1−xSb superlattices for long and very long‐wave infrared detection are discussed. This system promises benefits in this wavelength range over conventional technology based on Hg1−xCdxTe, in part because of suppressed band‐to‐band Auger recombination rates which lead to improved values of detectivity. The formalism for calculating Auger rates in superlattices is developed and the physical origin of Auger suppression in these systems is discussed. Accurate K⋅p band structures are used to obtain radiative, electron–electron, hole–hole, and band‐to‐band Auger rules, as well as shallow trap level assisted Auger recombination rates for photodiodes. Theoretical limits for high temperature operation of ideal photovoltaic detectors are presented and compared with HgCdTe.
188 citations
TL;DR: The algorithm performance is very encouraging, in terms both of accuracy of the resulting bounds, and of growth rate in required computation with problem size, and it appears that one can handle medium-size problems with reasonable computational requirements.
Abstract: Robustness problems involving real parametric uncertainty can be reformulated as mixed μ problems, where the block-structured uncertainty description is allowed to contain both real and complex blocks. Upper and lower bounds for the mixed μ problem have recently been developed, and this paper examines the computational aspects of these bounds. In particular a practical algorithm is developed to compute the bounds. This has been implemented as a Matlab function (m-file), which is currently available in conjunction with the μ-Tools toolbox. Some results from our extensive numerical experience with the algorithm are presented. The algorithm performance is very encouraging, in terms both of accuracy of the resulting bounds, and of growth rate in required computation with problem size. In particular it appears that one can handle medium-size problems (fewer than 100 perturbations) with reasonable computational requirements.
89 citations
TL;DR: In this article, an approach to directly incorporate modeling errors into the control design process is presented that leads to less sensitive and more robust controllers with improved performance, which is used to synthesize controllers for the NASA Langley Mini-Mast experimental structure.
Abstract: The lightly damped nature of flexible structures can lead to controllers that are highly sensitive to modeling errors in structural natural frequencies and damping levels. An approach to directly incorporating these modeling errors into the control design process is presented that leads to less sensitive and more robust controllers with improved performance/This approach is used to synthesize controllers for the NASA Langley Mini-Mast experimental structure. The resulting designs obtain good performance in the presence of significant modeling errors in the structural natural frequencies.
53 citations
[...]
TL;DR: In this article, the authors present an algorithm and experimental evidence to suggest that this goal has, for the first time, been achieved and compare with other approaches to robustness analysis and considering potential extensions, including controller synthesis.
Abstract: This paper gives an overview of promising new developments in robust stability and performance analysis of linear control systems with real parametric uncertainty. The goal is to develop a practical algorithm for medium size problems, where medium size means less than 100 real parameters, and "practical" means avoiding combinatoric (nonpolynomial) growth in computation with the number of parameters for all of the problems which arise in engineering applications. We present an algorithm and experimental evidence to suggest that this goal has, for the first time, been achieved. We also place these results in context by comparing with other approaches to robustness analysis and considering potential extensions, including controller synthesis.
24 citations
TL;DR: In this article, robust stability against structured perturbations of bounded lp norm is studied in terms of convex optimization problems, and the results are extended to solve the corresponding robust performance problems.
13 citations
13 Dec 1995
TL;DR: This work considers the problem of robust performance analysis when some of the exogenous inputs acting on the system are assumed to be fixed and known, while others are unknown but bounded, and proposes a new sufficient condition for robust performance.
Abstract: We consider the problem of robust performance analysis when some of the exogenous inputs acting on the system are assumed to be fixed and known, while others are unknown but bounded. In the special case when all the exogenous inputs are fixed and known, we extend previous results for the unstructured case to the general MIMO case where the perturbation is allowed to be structured. In this case we present both upper and lower bounds for the measure of robust performance. In the more general case, where both fixed and worst case inputs act on the system, we propose a new sufficient condition for robust performance. All these conditions are readily computable, and yield much less conservative results than one would obtain from applying standard worst-case analysis methods.
6 citations
21 Jun 1995
TL;DR: In this paper, a sufficient condition for the lack of a duality gap for norm minimization problems in the infinite dimensional case was developed, and an exact characterization of primal and dual problems in this case was given.
Abstract: In this paper we study norm minimization problems, and in particular the l/sub 1/ optimal control problem, subject to linear matrix inequality (LMI) constraints. We show how LMI's can be recast as parametrized LP's, and from this develop a sufficient condition for the lack of a duality gap for such problems in the infinite dimensional case. We give an exact characterization of the primal and dual problems in this case. In the case that we use the l/sub 1/ norm then both primal and dual problems can be posed as LMI optimization problems.