改进的粒子群优化算法的研究与应用

Research and application of improved particle swarm optimization algorithm

分析避风型渔港规划问题的特点,将其抽象为离散型约束选址分配问题,建立该问题的数学模型。根据模型特性,结合基本粒子群算法,提出一种改进的粒子群优化算法,给出用于计算离散型粒子位置和速度的替换变异操作的定义,保持种群的多样性、提高粒子的适应性;设计一种基于贪婪原则的高效的适应度函数。以渔港和渔船坐标为实验数据,用传统算法和改进粒子群优化算法分别在小规模和大规模实验数据上进行测试,实验结果表明,改进粒子群算法在解决大规模数据的渔港规划问题上表现出较高效率,有一定准确性,为解决该问题提供了可靠的方法。

To analyze the features of planning problem of fishing port sheltered from typhoon, and to abstract it into discrete con- straint location allocation problem, then mathematical model for the problem was built. According to the characteristics of model and combining the basic particle swarm optimization （PSO） algorithm, an improved PSO algorithm was proposed, and the defini tion of alternate mutate （AM） operation used to calculate discrete particle position and velocity was described, the diversity of population and the adaptability of particles were ensured. An efficient fitness function based on the principle of greed was de- signed. Experimental data were from fishing port and fishing boat coordinates, experiments on the traditional algorithm and the improved PSO algorithm for small-scale and mass-scale experimental data were carried out respectively. Results of the experi- ments show the high efficiency and certain accuracy of improved PSO algorithm on solving the fishing port planning problem of mass-scale data, and it provides reliable solutions for the problem.