Dual, Feasible Direction Algorithms.
Abstract
An algorithm for the solution of nonlinearly constrained optimization problems with essentially nonnegativity constraints only. The quadratic subproblems are solved by principal pivoting or other fast quadratic methods. A new method for preventing jamming (or zigzagging) is proposed. Also a general convergence theorem for optimization algorithms is given using the concept of a general necessary optimality function and incorporating the antijamming feature. All algorithms work under a choice of a number of step size selection methods. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1972
- Accession Number
- AD0742908
Entities
People
- Olvi L. Mangasarian
Organizations
- University of Wisconsin–Madison