具有多父代重组的遗传算法解0-1背包问题

Solutions to 0-1 Knapsack Problem Using Multi-parent Genentic Algorithms

遗传算法的创始人最初是从自然界获取灵感的,但是后来的遗传算法的研究者也试图将生物界不存在的特征引入遗传算法,多父代重组(或称为N父代重组,N>2)就是其中的一种。有文献显示这种机制在解很多不同的问题时都能有较好的效果。本文用几种多父代重组方法(包括一种新的面向特定问题的方法)解0-1背包问题,结果显示多父代重组确实有较好的性能。

looking at nature＇ s solutions is a suoree of inspiration to the genetic algorithms＇ originator. But trying non - existing biological features has been done by several GAs＇ praetitionors. As a ease, they took multi - parent recombination （ or N - parent recombination, N 〉 2）. Their work had shown the beneficial effects under many different circumstances. Several multi - parent recombination mechanisms, including a new problem - specific method, are experimented under solving 0 - 1 knapsack problem in this paper. The results prove that such mechanism can reach good performance.