基于双重距离的多目标粒子群优化算法

多目标粒子群优化(multi-objective particle swarm optimization,MOPSO)算法在维护收敛性的同时搜索分布良好的最优解集较为费力.为此,提出一种基于双重距离的MOPSO,由种群的平均距离定义粒子的邻域空间,邻域粒子数为粒子的等级,数量越多,粒子的等级越大.当等级相同时,算法结合粒子的拥挤距离选择最优粒子,并更新外部归档集.此外,算法结合粒子的变异行为避免陷入局部最优.在对比实验中,该算法在收敛性和多样性上可取得较优结果.最后,将该算法应用到电力系统的环境/经济调度模型(environmental/economic dispatch,EED),也可获得性能较好的解集.

Multi-objective particle awarm optimization algorithm based on dual distances

Multi-objective particle swarm optimization(MOPSO) algorithm is laborious to choose well-distributed optimal solutions while maintaining convergence. In this paper, an improved MOPSO based on dual distance is proposed to alleviate the above issues. It defines the neighborhood of a particle according to the average distance of the population, and thus the number of neighbors is the level of the particle. The more there are neighbors, the higher the level is. The average distance of two particles is equal so that with combining the crowding distance of swarm, the optimal particle is chosen and the extend archive is updated. In addition, this algorithm combines the mutation behavior of particles to avoid local space. In the experiment, the proposed algorithm is compared with several state-of-the-art algorithms, demonstrating that this algorithm achieves good performance in distribution and diversity of swarm. Finally, the algorithm is applied to the environmental/economic scheduling model in power systems, which can obtain well-distributed optimal solutions.