Abstract: In this paper, a competitive PSO is presented for MSA. Based on the best fitness of the particles, inertia weights and the velocity range of the particles are classified and adjusted dynamically with the evolution of particle swarm in the algorithm, which will improve the convergence speed and precision. To avoid the possible prematurity in the PSO, the re-initialization mechanism is introduced in the algorithm. The novel velocity model and the competitive strategy is presented, which increases the power of PSO. Experiment shows the algorithm is effective.

摘要： 本文提出了一种求解多序列联配的竞争粒子群优化算法，算法根据适应值分类设计了粒子群的惯性权重以及其飞行速度范围，并进行了动态调整，提高了算法的收敛速度和精度；引入了重新初始化机制，有效的避免粒子群优化算法可能出现的早熟现象；提出了一种全新的速度更新模式和竞争策略，增强了算法的能力。实验表明该算法是有效的。