中文摘要：

在介绍分布估计算法和粒子群算法的基础上，提出一种动态融合的并行混合进化算法EDAs／PSO。该算法采用分布估计算法和粒子群算法两种模型，既保持了分布估计算法较强的全局搜索能力，又保持了粒子群算法较强的局部搜索能力；另一方面，又采用一种动态融合的并行策略，通过周期性地对子种群中的个体按照一定的迁移率进行迁移操作，保证了种群的多样性，从而防止了算法的早熟收敛，提高了解的精度。实验结果表明了该算法的有效性和正确性。

英文摘要：

Based on introducing the estimation of distribution algorithm and the particle swarm optimisation, we present a parallel hybrid evolutionary algorithm with dynamic fusion named as EDAs/PSO. The algorithm adopts two models of estimation of distribution algorithm and particle swarm optimisation, while maintaining the stronger global search capability of the estimation of distribution algorithm, the stronger local search ability of the particle swarm optimisation is also preserved. On the other hand, it also employs a dynamically fused parallel strategy and guarantees the diversity of the population by periodical migration operations on the individuals in sub-population according to a certain of mobility so as to prevent the premature convergence of the algorithm, this improves solution＇ s precision. Experimental results also show the validity and correctness of the algorithm.