线性规划中大M法的参数M估值问题

在线性规划的单纯形法中,为求初始的可行基有著名的大M法,即惩罚因子法.在通常的运筹学教材中,只说明当M充分大时,大M法是有效的,并没有给出参数M的确切估计值.现给出一个确定的常数M0,并证明当M>M0时,大M法收敛于原问题的最优解.     

3阶幂图的广义导出匹配可扩性

A graph G is induced matching extendable if every induced matching of G is included in a perfect matching of G. A graph G is generalized induced matching extendable if every induced matching of G is included in a maximum matching of G. A graph G is claw-free, if G dose not contain any induced subgraph isomorphic to K1,3. The k-th power of G, denoted by Gu, is the graph with vertex set V（G） in which two vertices are adjacent if and only if the distance between them is at most k in G. In this paper we show that, if the maximum matchings of G and G3 have the same cardinality, then G3 is generalized induced matching extendable. We also show that this result is best possible. As a result, we show that if G is a connected claw-flee graph, then G3 is generalized induced matching extendable.