一、五种算法简介
1、哈里斯鹰优化算法HHO
2、鲸鱼优化算法WOA
3、灰狼优化算法GWO
4、蜣螂优化算法DBO
5、粒子群优化算法PSO
二、5种算法求解CEC2013
(1)CEC2013简介
参考文献:
[1] Liang J J , Qu B Y , Suganthan P N , et al. Problem Definitions and Evaluation Criteria for the CEC 2013 Special Session on Real-Parameter Optimization. 2013.
(2)部分python代码
from CEC2013.cec2013 import * import numpy as np from WOA import WOA from GWO import GWO from PSO import PSO from HHO import HHO from DBO import DBO import matplotlib.pyplot as plt plt.rcParams['font.sans-serif']=['Microsoft YaHei'] #主程序 #主程序 function_name =1 #测试函数1-28 SearchAgents_no = 50#种群大小 Max_iter = 100#迭代次数 dim=10#维度 10/30/50/100 lb=-100*np.ones(dim)#下限 ub=100*np.ones(dim)#上限 cec_functions = cec2013(dim,function_name) fobj=cec_functions.func#目标函数 BestX1,BestF1,curve1 = WOA(SearchAgents_no, Max_iter,lb,ub,dim,fobj)#问题求解 BestX2,BestF2,curve2 = GWO(SearchAgents_no, Max_iter,lb,ub,dim,fobj)#问题求解 BestX3,BestF3,curve3 = PSO(SearchAgents_no, Max_iter,lb,ub,dim,fobj)#问题求解 BestX4,BestF4,curve4 = HHO(SearchAgents_no, Max_iter,lb,ub,dim,fobj)#问题求解 BestX5,BestF5,curve5 = DBO(SearchAgents_no, Max_iter,lb,ub,dim,fobj)#问题求解 #画收敛曲线图 Labelstr=['WOA','GWO','PSO','HHO','DBO'] Colorstr=['r','g','b','k','c'] if BestF1>0: plt.semilogy(curve1,color=Colorstr[0],linewidth=2,label=Labelstr[0]) plt.semilogy(curve2,color=Colorstr[1],linewidth=2,label=Labelstr[1]) plt.semilogy(curve3,color=Colorstr[2],linewidth=2,label=Labelstr[2]) plt.semilogy(curve4,color=Colorstr[3],linewidth=2,label=Labelstr[3]) plt.semilogy(curve5,color=Colorstr[4],linewidth=2,label=Labelstr[4]) else: plt.plot(curve1,color=Colorstr[0],linewidth=2,label=Labelstr[0]) plt.plot(curve2,color=Colorstr[1],linewidth=2,label=Labelstr[1]) plt.plot(curve3,color=Colorstr[2],linewidth=2,label=Labelstr[2]) plt.plot(curve4,color=Colorstr[3],linewidth=2,label=Labelstr[3]) plt.plot(curve5,color=Colorstr[4],linewidth=2,label=Labelstr[4]) plt.xlabel("Iteration") plt.ylabel("Fitness") plt.xlim(0,Max_iter) plt.title("cec2013-F"+str(function_name)) plt.legend() plt.savefig(str(function_name)+'.png') plt.show() #