On Non Overlapping Segmentation of the Response Surfaces for Solving Constrained Programming Problems Through Super Convergent Line Series

Abstract

The solutions of Constrained Programming Problems (linear, quadratic, and cubic) by segmentation of the response surfaces through Super Convergent Line Series were obtained. The line search exchange algorithm was exploited. The response surfaces were explored and segmented up to four segments for linear, six for quadratic, and eight for cubic programming problems, respectively. It was verified and established that the number of segments, S, for which optimal solutions are obtained are two for linear, four for quadratic, and eight for cubic programming problems, respectively.

Keywords:

• Non-overlapping segmentation
• Optimal solutions
• Response surfaces
• Super convergent line series
• Support points

Mathematics Subject Classification:

• 62K
• 62