Sequential Joint Maximization

摘要: This paper describes a new method by which competitive market economy may be represented as the optimal solution to planning problem The resulting sequential joint maximization (SJM) algorithm solves sequence of “partial equilibrium relaxations” underlying general model partial submodels can solved nonlinear complementarity problems or constrained optimization in either primal dual form. introduces three SJM algorithms, evaluates performance and examines convergence theory. Computational tests demonstrate that is not always efficient methods. A counterexample presented demonstrates local cannot guaranteed Although have neither theoretical pedigree fixed-point nor rate Newton algorithm, usefulness demonstrated range real models for procedure has been successfully applied benefits from availability robust, large-scale programming codes. Consequently, large scale with inequalities “best” theory, but it works extremely well practice.