摘要： 通过分析传统遗传算法和多亲遗传算法的不足，提出了一种多亲遗传算法的改进算法：基于共享存储器的多亲遗传算法，并对其进行了理论分析，讨论了GA的并行模型特点后，结合粗粒度并行模型和群体分组的并行方式，提出了一种MGASM的并行模型，该模型有利于改进MGASM的性能，提高其搜索效率。将MGASM-PPGA应用到了数据聚类问题中，进行了仿真实验，获得了理想的实验结果。

关键词: 多亲遗传算法, 共享存储器, 理论分析, 并行处理, 数据聚类

Abstract: By discussing the lack of traditional genetic algorithm(TGA) and multi-parent genetic algorithm(MGA), this paper proposes an improved algorithm called multi-parent genetic algorithm based on shared memory(MGASM), and analyses it in theory. Parallel model of GA is discussed. Combining parallel model based on thick grain with method of colony grouping, parallel model of MGASM is proposed. It is benefit to improve the performance of MGASM, and to raise the searching efficiency. MGASM-PPGA(psedo-parallel genetic algorithm) is applied in data-clustering problem. Emulational experimental figures show it has nice performance.

Key words: Multi-parent genetic algorithm(MGA), Shared memory, Theoretic analysis, Parallel processing, Data-clustering