Python:遗传算法实现

# Encode parameters into DNA    
def encode(self, parameters):
    dna = ‘‘
    for i in range(self.nGenes):
        geneSequenceDigit = (parameters[i]-self.parametersRange[i,0])/(self.parametersRange[i,1]-self.parametersRange[i,0])*(2**self.geneLength[i]-1)
        geneSequenceDigit = int(round(geneSequenceDigit))
        geneSequenceBin   = self.int2bin(geneSequenceDigit, self.geneLength[i])
        dna = dna + geneSequenceBin 
    dna = list(dna)  # Trun string to list
    return dna

# Decode DNA to parameters
def decode(self, dna):
    dna = ‘‘.join(dna)  # Trun list to string
    parameters = []
    for i in range(self.nGenes):
        geneSequenceBin   = dna[self.dnaIdx[i,0]:self.dnaIdx[i,1]+1]
        geneSequenceDigit = self.bin2int(geneSequenceBin)
        parameterI = geneSequenceDigit/(2**self.geneLength[i]-1)*(self.parametersRange[i,1]-self.parametersRange[i,0])+self.parametersRange[i,0]
        parameters.append(parameterI)
    return parameters

# Returns the binary string of integer n, using count number of digits
def int2bin(self, n, count=32):
    binStr = "".join([str((n >> y) & 1) for y in range(count-1, -1, -1)])
    return binStr

# Returns digit integer of binary string
def bin2int(self, n):
    ret = int(n,2)
    return ret

# Select individuals
def select(self):
    nCandidates = 2
    generation  = []
    for i in range(self.popGeneration):
        candidates = random.choices(self.generation, k=nCandidates)
        fitnessList = self.fitness(candidates=candidates)
        bestFitness = max(fitnessList)
        idx = fitnessList.index(bestFitness)
        generation.append(candidates[idx].copy())
    self.generation = generation
    return

# Elitist preservation
def elitistPreservation(self):
    bestFitness = max(self.fitnessList)
    if bestFitness > self.bestFitness:
        idx = self.fitnessList.index(bestFitness)
        self.bestFitness    = bestFitness
        self.bestIndividual = self.generation[idx].copy()
    else:
        worstIndividual = min(self.fitnessList)
        idx = self.fitnessList.index(worstIndividual)
        self.generation[idx]  = self.bestIndividual.copy()
        self.fitnessList[idx] = self.bestFitness
    return

import GeneticAlgorithm as GA # 导入GeneticAlgorithm模块
import math

# 根据具体的优化目标设计适应度函数，par为输入参数的列表，ret为适应度
def fitness(par):
    x = par[0]
    ret = x*math.sin(10*math.pi*x)+2
    return ret

maxGen = 500 # 迭代次数
popGeneration = 50 # 每个世代的个体数目
pCross = 0.1 # 发生交叉的概率
pMutation = 0.05 # 发生变异的概率
nGenes = 1 # 待优化的参数的个数
parametersRange = [[-1, 2, 0.01]] # 待优化参数的范围及要求的精度，多个参数写法：[[-1, 2, 0.01]，[-2, 3, 0.1]]

# 创建GA对象
ga = GA.GeneticAlgorithm(maxGen, popGeneration, pCross, pMutation, nGenes, parametersRange, fitness)
# 显示对象信息
ga.info()
# 执行优化
ga.run()
# 显示最优解
ga.result()