一种基于景观特征的浮点数编码遗传算法研究

遗传算法作为一种适应性搜索技术得到了普遍的应用,但其搜索效率不如启发式搜索.已有研究者将启发式知识用于二进制编码遗传算法,但浮点数编码在函数优化和约束优化领域明显有效于其它编码.本文基于算法运行时的景观特征作为启发式知识,用于浮点数编码遗传算法,力求提高其搜索效率、增强其局部搜索能力、拓展其应用领域.本文的理论研究和实验结果表明,将景观特征用于浮点数编码遗传算法,理论是可靠的,方法是可行的.

Research on Genetic Algorithm of Float Number Code Based on Landscape Character

Genetic algorithm(GA)was used widely as a adapted search technology.GA is not as good as heuristic search in search efficiency.Binary code GA was researched by scholars with heuristic knowledge at past.Float number code is superior to other codes in function optimization and restriction optimization.In this paper,landscape character in algorithm running was used to GA of float number code as heuristic knowledge.It was aimed by it that improving GA's search efficiency,enhancing local search performance,extending application fields.It was indicated by the results of it's theoretic research and experiment result that the theory is credible,the method is feasible,landscape character could be used to GA of float number code.