公开(公告)号：US6052678A

公开(公告)日：2000-04-18

申请号：US356979

申请日：1994-12-16

申请人： Hiroyasu Itoh ,  Akio Shinagawa ,  Toshihiro Nishimura ,  Hiroyuki Okada ,  Nobue Adachi ,  Hajime Ohi

发明人： Hiroyasu Itoh ,  Akio Shinagawa ,  Toshihiro Nishimura ,  Hiroyuki Okada ,  Nobue Adachi ,  Hajime Ohi

IPC分类号： G06N3/12 ,  G06N5/00

CPC分类号： G06N3/126 ,  G06N5/003

摘要： A problem solving operation apparatus obtains an optimum solution for a given problem through a combined use of a plurality of optimum solution search techniques of a genetic algorithm, and other search techniques like a hill climbing method. A combination of the optimum solution search techniques applied to each of a group of solution candidates varies according to each state. When a search efficiency in a single state is lowered, the problem solving operation apparatus immediately causes a state transition to the next state, thus enabling a search of an optimum solution through the use of another combination of more efficient optimum solution search techniques. Because a state transition constantly maintains a high search efficiency, this problem solving operation apparatus can reach an optimum solution in a short amount of time.

摘要翻译： 问题解决操作装置通过综合使用遗传算法的多个最优解搜索技术以及诸如爬山方法的其他搜索技术来获得给定问题的最优解。 应用于一组解决方案候选者中的每一个的最佳解决方案搜索技术的组合根据每种状态而变化。 当单一状态的搜索效率降低时，问题解决操作装置立即使状态转换到下一状态，从而通过使用更有效的最优解搜索技术的另一组合来搜索最佳解。 由于状态转换始终保持较高的搜索效率，所以该问题解决操作装置可以在短时间内达到最佳解决方案。