组合优化（Ⅱ）——对称差分解法的又一应用

ON COMBINATORIAL OPTIMIZATION (Ⅱ) ANOTHER APPLICATION OF SYMMETRICAL DIFFERENCE METHOD

本文先讨论函数的增量与微分对于连续型最优化问题的作用，析出有益的启发。用之于组合优化，得到了求解问题的一个方法——对称差（的）分解法。文献［２］对它作了讨论并得到不少应用。本文提出两个赋权凸锥独立集合问题。它们是典型的组合优化问题，分别与线性规划中两个互为对偶模型等价；用对称差分解法进行求解。

The article,first,analyses the core essence of the increment and differential of continuous function to the optimum problems of continuous type.Applying the essence to combinatorial optimization problems,it deduces a method,called the symmetrical difference method. has already discussed the method and presented several applications.Secondly,it gives two weighted convex cone independent set problems.They are typical problems of combinatorial optimization.They are equivalent to two such models of linear programming that they have the primal dual relation.An algorithm is deduced by the symmetrical difference method for them.It reveals that the algorithm obtained is simply the well known revised simplex algorithm for linear programming.