逐步缩小搜索范围的遗传算法

THE GENETIC ALGORITHM THAT GRADUALLY REDUCES THE SEARCHING RANGES

本文介绍在遗传算法的运行过程中如何逐步缩小搜索范围，使搜索速度与计算精度同时得到提高的方法。文中提出了真值邻域的概念，利用这个概念得到了缩小搜索范围的压缩公式。只要合理选择初始搜索范围（这一点并不难做到），上述压缩公式能确保问题的解始终保留在压缩后的搜索范围内，通过理论模型的检验及与原有遗传算法结果的比较，证实了方法的有效性。

This paper introduced the method that reduces the searehing range gradually and raises searching speed and calculating precision simultaneously in the run-time of Genetie Algorithm. The conception of neighborhood of the real solution was presented, and the formula used for reducing the searching range was obtained by means of the above-mentioned conception. The formula can ensure the solution of a problem throughout staies in the range after reducing if only the initial searching range was selected suitablely (it is easy to do). The effectiveness of the method was verified throughout calculation on the theoretical models and comparision with the results of the original Genetic Algorithm.