基于混沌初始化和高斯变异的鸽群算法  

Pigeon cluster algorithm based on chaotic initialization and Gaussian mutation

作　　者：魏超 韦修喜 黄华娟[2] WEI Chao;WEI Xiu-xi;HUANG Hua-juan(College of Electronic Information,Guangxi University for Nationalities,Nanning 530006,China;College of Artificial Intelligence,Guangxi University for Nationalities,Nanning 530006,China)

机构地区：[1]广西民族大学电子信息学院,广西南宁530006 [2]广西民族大学人工智能学院,广西南宁530006

出　　处：《计算机工程与设计》2023年第4期1112-1121,共10页Computer Engineering and Design

基　　金：国家自然科学基金项目(61662005);广西自然科学基金项目(2018GXNSFAA294068、2021GXNSFAA220068);广西民族大学科研基金项目(2019KJYB006)。

摘　　要：针对鸽群优化算法在求解非线性优化问题中,容易陷入局部最优,收敛精度不高的问题,提出改进的鸽群优化算法。采用混沌映射中的立方映射方法对鸽群位置进行初始化,增加种群的多样性;引入高斯变异算子,弥补鸽群算法容易陷入局部最优的不足,提高算法的全局搜索能力和搜索效率;在地标算子中添加递减因子,能够有效避免算法由于过早收敛而陷入局部最优,提高算法的收敛精度。测试19个基准函数和电力系统经济调度工程应用实验的结果表明,改进后算法与其它群智能算法相比,拥有更好的寻优能力。Aiming at the problems that the pigeon optimization algorithm is easy to fall into the local optimum and the convergence accuracy is not high when solving the nonlinear optimization problem,an improved pigeon optimization algorithm was proposed.The cubic mapping in the chaotic mapping method was used to initialize the position of the pigeon group to increase the diversity of the population.The introduction of the Gaussian mutation operator made up for the lack of the pigeon group algorithm that is easy to fall into the local optimum,and the algorithm’s global search ability and search efficiency were improved.A decrement factor was added to the landmark operator to prevent the algorithm from falling into the local optimum due to premature convergence,and the convergence accuracy of the algorithm was improved.The results of testing 19 benchmark function and power system economic dispatching engineering application experiments show that the improved algorithm has better optimization capabilities than other swarm intelligence algorithms.

关 键 词：鸽群优化算法 非线性优化 混沌映射 立方映射 高斯变异算子 递减因子 电力经济调度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP301.6[自动化与计算机技术—控制科学与工程]