Inverse problems of mathematical programming. Global optimization methods. (Q1432476): Difference between revisions

The author studies the problem of finding a pare of vectors \((x^*, u^*) \in \{ (x,u): f(x,u)\leq 0, w(x,u)=0, u \in R_u \}\), where \(f\) and \(w\) are continuous finite-dimensional vector functions, \(R_u \subset E^m\) a compact subset, and \(x^*\) solves the mathematical programming problem \(\min\{ \varphi(x,u^*): g(x,u^*) \leq 0, x \in R_x \}\); \(R_x \subset E^n\) is a compact subset. Let \(\varphi(x,u)\), \(f(x,u)\), and \(g(x,u)\) be jointly convex (vector) functions in \((x,u)\), \(w(x,u)\) an affine vector function, \(R_x\) and \(R_u\) convex compact sets. The original problem is reduced to a system of several convex inequalities and an inequality with a d.c. function. For solving the system, the author constructs an iterative process closely related to the Huard method.