ON ENTROPY-BASED METHODS FOR NONLINEAR PROGRAMMING PROBLEMS

Abstract

This paper presents further theoretical analysis and developments of entropy-based methods for mathematical programming. A new generalized multiplier function is given which approximates uniformly the maximum function, and from which the aggregate function is derived in a different way. Some new methods for minimax and constrained nonlinear programming problems are proposed. The convergence properties for these methods are discussed.

KEY WORDS:

• Entropy-based optimization
• aggregate function
• multiplier function
• minimax and constrained optimization.