摘要
文章以并行遗传算法为基础，根据生物遗传规律的特点，综合多种杂交方式和生物遗传技术（共享技术、受限交配、灾变等），提出了一种综合的并行搜索策略，用以克服遗传算法的早熟收敛现象，提高遗传寻优的效率。通过对剩余静校正估计的模型试验，其结果比常规的单一遗传算法有明显的改善，能够更有效地接近于真正的最优解。
In this paper, a hybrid parallel search strategy is proposed on the basis of parallel genetic algorithm. The strategy combines with a variety of crossover schemes and a new technology with the character of genetic low of biology. In a sense, it can eliminate premature convergence and enhance the searching efficiency of genetic algorithm. Through testing residual statics estimation, the result is obvoiusly improved compared with the conventional single genetic algorithm.
出处
《成都理工学院学报》
CSCD
1997年第2期75-80,共6页
Journal of Chengdu University of Technology
关键词
遗传算法
并行搜索策略
剩余静校正
地震勘探
genetic algotithm
hybrid parallel search strategy
premature convengence
residual statics
