Decentralization in Linear Programming Models
提出一种基于最优基矩阵结构的线性规划分解方法，将问题拆分为价格设定和数量设定两个子问题，适用于具有共同约束或共同变量的情形，并以乳制品生产计划为例进行说明。
ABSTRACT This paper proposes a decomposition of a linear programming problem based on the structure of the optimal basis matrix. If this matrix contains a zero matrix of appropriate dimensions, the problem may be decomposed into a price‐setting problem and a quantity‐setting problem. This decomposition is valid for a set of coefficients of the problem to be determined by parametric programming. It can be applied to problems with common constraints or common variables. An application to dairy production planning is discussed and a comparison with the Dantzig‐Wolfe decomposition principle is given.