Analytic Functions Optimizing Competing Constraints

This paper analyzes a problem in which one optimizes objective functions over the space \(H_N^\infty\) of vector-valued functions \(f = (f_1, \ldots, f_N)\) defined on the unit circle, \(T\), where each coordinate function \(f_j\) belongs to \(L^\infty (T)\) and extends to be analytic on the entire unit disk. The objectives to be optimized are described in terms of nonnegative continuous functions \(\Gamma\) defined on \(T \times C^N\). Given positive functions \(\Gamma_j (e^{i \Theta}, z)\) in \(C^1 (T, C^N), j = 1, \ldots, p\), and a function \(f\) in \(H_N^\infty\), the authors define the \(p\) performances \(\gamma_j (f) : = sup {\Gamma_j (e^{i \Theta}, f), \Theta \in T}, j=1, \ldots, p\). Under strong assumptions on the performance functions the authors prove that if we optimize over \(N\) functions in \(H^\infty\) and have \(p\) performance measures with \(p \leq N\), then at a nondegenerate Pareto optimum, every performance is frequency independent (flat). Besides flatness, there are other gradient-alignment conditions which must hold at an optimum. The authors present these conditions and thus give the precise first-derivative test for a natural class of \(H ^\infty\) Pareto optima.

Recommendations

Cited in

(5)

This page was built for publication: Analytic Functions Optimizing Competing Constraints

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4339157)