S-lemma with equality and its applications
Yong Xia,Shu Wang,Ruey-Lin Sheu
DOI: https://doi.org/10.1007/s10107-015-0907-0
IF: 2.7
2015-04-25
Mathematical Programming
Abstract:Let f(x)=xTAx+2aTx+c$$f(x)=x^TAx+2a^Tx+c$$ and h(x)=xTBx+2bTx+d$$h(x)=x^TBx+2b^Tx+d$$ be two quadratic functions having symmetric matrices A$$A$$ and B$$B$$. The S-lemma with equality asks when the unsolvability of the system f(x)<0,h(x)=0$$f(x)<0, h(x)=0$$ implies the existence of a real number μ$$\mu $$ such that f(x)+μh(x)≥0,∀x∈Rn$$f(x) + \mu h(x)\ge 0, ~\forall x\in \mathbb {R}^n$$. The problem is much harder than the inequality version which asserts that, under Slater condition, f(x)<0,h(x)≤0$$f(x)<0, h(x)\le 0$$ is unsolvable if and only if f(x)+μh(x)≥0,∀x∈Rn$$f(x) + \mu h(x)\ge 0, ~\forall x\in \mathbb {R}^n$$ for some μ≥0$$\mu \ge 0$$. In this paper, we show that the S-lemma with equality does not hold only when the matrix A$$A$$ has exactly one negative eigenvalue and h(x)$$h(x)$$ is a non-constant linear function (B=0,b≠0$$B=0, b\not =0$$). As an application, we can globally solve inf{f(x):h(x)=0}$$\inf \{f(x): h(x)=0\}$$ as well as the two-sided generalized trust region subproblem inf{f(x):l≤h(x)≤u}$$\inf \{f(x): l\le h(x)\le u\}$$ without any condition. Moreover, the convexity of the joint numerical range {(f(x),h1(x),…,hp(x)):x∈Rn}$$\{(f(x), h_1(x),\ldots , h_p(x)):x\in \mathbb R^n\}$$ where f$$f$$ is a (possibly non-convex) quadratic function and h1(x),…,hp(x)$$h_1(x),\ldots ,h_p(x)$$ are affine functions can be characterized using the newly developed S-lemma with equality.
mathematics, applied,computer science, software engineering,operations research & management science