scientific article; zbMATH DE number 1524343

From MaRDI portal
Publication:4512231





The paper deals with the following mathematical programming problem \[ \min\limits_{u\in U}\left\{f_0(u): f_{i}(u)\leq 0, i\in J=\{1,2,\ldots,l\}\right\}, \] \[ U=\left\{u\in {\mathbb R}^{n}:-\infty<a^{i}\leq u^{i}\leq b^{i}<\infty, i\in \{1,2,\ldots,n\}\right\}, \] where \(f_{i}(u), i=0,\dots, l\) are continuously differentiable functions. On every iteration of a linearization method for the solution of the given problem it is necessary to find a solution to the following auxiliary quadratic programming problems \[ \min\limits_{p\in Q(u,\mu)}\{\langle f_0(u),p\rangle+\tfrac 12\langle p,Ap\rangle:\langle f'_{i}(u),p\rangle+f_{i}(u)\leq 0, i\in J_{\delta}(u)\}, \] where \(A\) is a symmetric positive definite matrix; \(Q(u,\mu)\) is a special set in \({\mathbb R}^{n}\); \(J_{\delta}(u)=\{i\in J: f_{i}(u)\geq F(u)-\delta\}, \delta>0\); \(F(u)=\max\{0,f_1(u),\ldots,f_{l}(u)\}\). The authors propose an algorithm of the linearization method for finding a solution of the given mathematical programming problem.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4512231)