Dual Applications of Proximal Bundle Methods, Including Lagrangian Relaxation of Nonconvex Problems (Q4509733)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1519728
Language Label Description Also known as
default for all languages
No label defined
    English
    Dual Applications of Proximal Bundle Methods, Including Lagrangian Relaxation of Nonconvex Problems
    scientific article; zbMATH DE number 1519728

      Statements

      Dual Applications of Proximal Bundle Methods, Including Lagrangian Relaxation of Nonconvex Problems (English)
      0 references
      0 references
      0 references
      19 October 2000
      0 references
      nondifferentiable optimization
      0 references
      convex programming
      0 references
      proximal bundle methods
      0 references
      Lagrangian relaxation
      0 references
      convexified relaxations
      0 references
      unit commitment
      0 references
      numerical examples
      0 references
      power production scheduling
      0 references
      Useful properties of proximal bundle methods for finding \(\min_Sf\), where \(f\) and \(S\) are convex, are proposed. The following convex constrained minimization problem: \(f_*:= \min\{f(x):x\in S\}\) (where \(S\) is a nonempty closed convex set in \(\mathbb{R}^n\), \(f: \mathbb{R}^n\to \mathbb{R}\) is a convex function), is considered. The nonemptyness of the optimal set: \(S_*:= \text{Argmin}_Sf\) is assumed. Main result: The authors show that the proximal bundle method finds asymptotically not only some \(x^\infty\in S_*\), but also the objective subgradients and constraint multipliers involved in the optimality conditions for \(\min_Sf\), and the multipliers of the objective pieces when \(f\) is a max-type function, and primal and dual solutions on Lagrangian decomposition of convex programs. As the application to Lagrangian relaxation of nonconvex programs, they find solutions to relaxed convexified versions of such programs. Numerical results are presented for unit commitment in power production scheduling.
      0 references

      Identifiers