Optimization of the Operation of A Complex Water Resources System ; Part -I : Analysis of the Convergence Criterrion in A Solution by the Discrete Differential Dynamic Programming

Abstract

An iterative - solution procedure necessarily involved pre- specified convergence criteria to stop iteration . The Discrete Differential dynamic Programming procedure to solve optimization problems formulated by the Dynamic Programming is an iterative - solution procedure which , in its traditional form , involves two convergence criteria , namely , (α) and (β).The research used the optimum operation of an existing complex water - resources system as a case study . The objective function was formulated as the maximum real monetary return . The formulated optimization model was run for a total of (194) different operation cases . Beside the traditional (α) and (β), seven new styles for a unique convergence criterion were examined in the solution.Considering the monetary return and the number of performed iterations as the bases of comparison , the research showed that the new (ɣ) convergence criterion was the favorite among the tested convergence criteria.