分支定界分支方法_分支定界

分支定界分支方法

分支定界 (Branch and bound)

• Branch and bound is a state space search method that can be termed as an improved form of backtracking.

  分支定界是一种状态空间搜索方法，可以称为回溯的一种改进形式。

• It is suitable for solving the combinatorial optimization problem.

  适用于解决组合优化问题。

• These problems have an enumerable finite feasible solution.

  这些问题具有有限的可行解。

• In this approach, a set of feasible solutions are partitioned and the subsets that do not have an optimal solution are deleted for further consideration. Thus, the branch and bound has two major steps:

  在这种方法中，对一组可行解进行了划分，并删除了没有最优解的子集，以供进一步考虑。 因此，分支和绑定有两个主要步骤：

  1. Branching
  2. Bounding

The following are the steps that are repeated to solve any problem are:

以下是解决任何问题重复执行的步骤：

1)分支 (1) Branching)

Branching is the first step of the branch and bound technique. This technique involves division of the main problem into two or more subproblems. These subproblems are exclusive and independent of each other. In addition, they are similar to the original problem but smaller in size i.e. the operation applied to the main problem can also be applied to these subproblems.

分支是分支定界技术的第一步。 该技术涉及将主要问题分为两个或多个子问题。 这些子问题是排他的，彼此独立。 此外，它们与原始问题相似，但尺寸较小，即，应用于主要问题的操作也可以应用于这些子问题。

2)边界 (2) Bounding)

Bounding operation restricts the state space tree to grow exponentially.

边界操作限制状态空间树成倍增长。

Branch and bound is the extension of backtracking which is used to solve combinatorial optimize problem. This is more useful because in this the branches are more optimally find as it uses the breath first search so, that our search is more optimal and state space tree is more compact. This can be solved by using the following:

分支定界是回溯的扩展，用于解决组合优化问题。 这更有用，因为在这种情况下，由于它使用呼吸优先搜索，因此可以更好地找到分支，从而使我们的搜索更加优化，状态空间树也更加紧凑。 可以使用以下方法解决：

1. FIFO - (First In First Out): The list of live nodes is a first - in - first - out the list. Example - queue.

   FIFO-(先进先出)：活动节点列表是先进先出列表。 示例队列 。

2. LIFO - (Last In Last Out): The list of the live nodes is a last - in - first - out the list. Example - stack.

   LIFO-(后进后出)：活动节点的列表是后进先出列表。 示例堆栈 。

3. LCS - (Least Cost Search): In this, we give the preference to an answer node with minimum cost.

   LCS-(最低成本搜索)：在这种情况下，我们将优先考虑的是成本最低的答案节点。

The branch and bound is that state space method in which all the children’s of the E-node are generated before any other live node can become the E-node. The two graph strategies i.e. the Breath first search and depth-first search, in which the exploration of a new node cannot begin until the node currently being explored is fully explored. Bounding function is used to avoid the generation of subtrees that do not contain an answer node. Branch and bound check the state space tree complexity for an optimal value and potential solutions are always obtained. This branch and bound is mostly used as it makes more compact state space tree.

分支定界是一种状态空间方法，其中在任何其他活动节点都可以成为E节点之前生成E节点的所有子节点。 两种图策略，即“呼吸优先”搜索和“深度优先”搜索，其中直到充分探索当前正在探索的节点之前，才可以开始探索新节点。 边界函数用于避免生成不包含答案节点的子树。 分支定界检查状态空间树的复杂性以获得最佳值，并始终获得潜在的解决方案。 这个分支和边界最常用，因为它使状态空间树更紧凑。

翻译自: https://www.includehelp.com/algorithms/branch-and-bound.aspx

分支定界分支方法