# -*-coding:utf-8 -*- #目标求解sin(x)最大值 import random import math import matplotlib.pyplot as plt #初始化种群 生成chromosome_length大小的population_size个个体的种群 def species_origin(population_size,chromosome_length): population=[[]] #one dimension represent a individual for i in range(population_size): temporary=[] #染色体暂存器 for j in range(chromosome_length): temporary.append(random.randint(0,1)) #随机产生一个染色体,由二进制数组成 population.append(temporary) #将染色体添加到种群中 return population[1:] # 将种群返回，种群是个二维数组，个体和染色体两维 #从二进制到十进制 #编码 input:种群,染色体长度 def translation(population,chromosome_length): temporary=[] for i in range(len(population)): total=0 for j in range(chromosome_length): total+=population[i][j]*(math.pow(2,j)) #从第一个基因开始，每位对2求幂，再求和 # 如：0101 转成十进制为：1 * 20 + 0 * 21 + 1 * 22 + 0 * 23 = 1 + 0 + 4 + 0 = 5 temporary.append(total) #一个染色体编码完成，由一个二进制数编码为一个十进制数 return temporary # 返回种群中所有个体编码完成后的十进制数 #from protein to function,according to its functoin value #a protein realize its function according its structure # 目标函数相当于环境 对染色体进行筛选，这里是sin(x) def function(population,chromosome_length,max_value): temporary=[] function1=[] temporary=translation(population,chromosome_length) # 暂存种群中的所有的染色体(十进制) for i in range(len(temporary)): x=temporary[i]*max_value/(math.pow(2,chromosome_length)-1) # function1.append(math.sin(x)) #这里将sin(x)作为目标函数 return function1 #定义适应度 def fitness(function1): fitness1=[] min_fitness=mf=0 for i in range(len(function1)): if(function1[i]+mf>0): temporary=mf+function1[i] else: temporary=0.0 # 如果适应度小于0,则定为0 fitness1.append(temporary) #将适应度添加到列表中 return fitness1 #计算适应度和 def sum(fitness1): total=0 for i in range(len(fitness1)): total+=fitness1[i] return total #计算适应度斐伯纳且列表 def cumsum(fitness1): for i in range(len(fitness1)-2,-1,-1): # range(start,stop,[step]) # 倒计数 total=0 j=0 while(j<=i): total+=fitness1[j] j+=1 fitness1[i]=total fitness1[len(fitness1)-1]=1 #3.选择种群中个体适应度最大的个体 def selection(population,fitness1): new_fitness=[] #单个公式暂存器 total_fitness=sum(fitness1) #将所有的适应度求和 for i in range(len(fitness1)): new_fitness.append(fitness1[i]/total_fitness) #将所有个体的适应度正则化 cumsum(new_fitness) # ms=[] #存活的种群 population_length=pop_len=len(population) #求出种群长度 #根据随机数确定哪几个能存活 for i in range(pop_len): ms.append(random.random()) # 产生种群个数的随机值 ms.sort() # 存活的种群排序 fitin=0 newin=0 new_population=new_pop=population #轮盘赌方式 while newinbestfitness): bestfitness=fitness1[i] bestindividual=population[i] return [bestindividual,bestfitness] population_size=500 max_value=10 # ?这个值是干什么的 chromosome_length=10 pc=0.6 pm=0.01 results=[[]] fitness1=[] fitmean=[] population=pop=species_origin(population_size,chromosome_length) #生成一个初始的种群 for i in range(population_size): function1=function(population,chromosome_length,max_value) fitness1=fitness(function1) best_individual,best_fitness=best(population,fitness1) results.append([best_fitness,b2d(best_individual,max_value,chromosome_length)]) #将最好的个体和最好的适应度保存，并将最好的个体转成十进制 selection(population,fitness1) crossover(population,pc) mutation(population,pm) results=results[1:] results.sort() X=[] Y=[] for i in range(500): X.append(i) Y.append(results[i][0]) plt.plot(X,Y) plt.show()
