def singleGA(populationCount, crossRate, murationRate, iteraNum, filename):
cvrp = CVRP()
name, dimension, capacity, CustomerDistance, Requirement = LoadData(filename)
ga = GA(GeneLenght=(dimension-1), PopulationCount=populationCount, CrossRate=crossRate, MurationRate=murationRate, DistanceMat=CustomerDistance, Requirement=Requirement, Capacity=capacity)
ga.initPopulation()
population = copy.copy(ga.population)
#population.append([21, 31, 19, 17, 13, 7, 26, 12, 1, 16, 30, 27, 24, 29, 18, 8, 9, 22, 15, 10, 25, 5, 20, 14, 28, 11, 4, 23, 3, 2, 6])
bestGene = []
minDistance = 100000
minItera = 0
for itera in range(iteraNum):
newPopulation = []
while True:
selectNum1, selectNum2, bestNum, distance, bestPath, bestFit = ga.selectGene(population)
child1, child2 = ga.cross(selectNum1, selectNum2)
newPopulation.extend(ga.muration([child1, child2]))
if len(newPopulation) > populationCount:
newPopulation.append(population[bestNum])
bestGene = copy.copy(population[bestNum])
if distance < minDistance:
minItera = itera
minDistance = distance
population = copy.copy(newPopulation)
break
print minItera, bestPath, minDistance, bestFit
return minItera, bestPath, minDistance, bestFit
def doubleGA(populationCount, crossRate1, murationRate1, crossRate2, murationRate2, iteraNum, filename):
cvrp = CVRP()
name, dimension, capacity, CustomerDistance, Requirement = LoadData(filename)
ga1 = GA(GeneLenght=(dimension-1), PopulationCount=populationCount, CrossRate=crossRate1, MurationRate=murationRate1, DistanceMat=CustomerDistance, Requirement=Requirement, Capacity=capacity)
ga2 = GA(GeneLenght=(dimension-1), PopulationCount=populationCount, CrossRate=crossRate2, MurationRate=murationRate2, DistanceMat=CustomerDistance, Requirement=Requirement, Capacity=capacity)
ga1.initPopulation()
ga2.initPopulation()
population1 = copy.copy(ga1.population)
population2 = copy.copy(ga2.population)
bestGene = []
minDistance = 100000
minItera = 0
for itera in range(0,iteraNum):
newPopulation1 = []
newPopulation2 = []
while True:
#选择两个父亲，生成两个子代后变异放入新种群
selectNum1, selectNum2, bestNum1, distance1, bestPath1, bestFit1 = ga1.selectGene(population1)
child1, child2 = ga1.cross(selectNum1, selectNum2)
newPopulation1.extend(ga1.muration([child1, child2]))
selectNum1, selectNum2, bestNum2, distance2, bestPath2, bestFit2 = ga2.selectGene(population2)
child1, child2 = ga2.cross(selectNum1, selectNum2)
newPopulation2.extend(ga2.muration([child1, child2]))
#当数量超过设定值，交换两个平行种群的解
if len(newPopulation1) > populationCount:
selectNum1, selectNum2, bestNum1, distance1, bestPath1, bestFit1 = ga1.selectGene(newPopulation1)
selectNum1, selectNum2, bestNum2, distance2, bestPath2, bestFit2 = ga2.selectGene(newPopulation2)
exPopulation1 = []
exPopulation2 = []
exNum1 = range(len(newPopulation1)-1)
exNum2 = range(len(newPopulation2)-1)
random.shuffle(exNum1)
random.shuffle(exNum2)
exPopulation1.append(newPopulation2[bestNum2])
exPopulation2.append(newPopulation1[bestNum1])
num = random.randint(5,len(newPopulation1)-5)
count = 0
for i in range(len(newPopulation1)):
if i != bestNum1:
if i in exNum1:
if count < num:
exPopulation2.append(newPopulation1[i])
else:
exPopulation1.append(newPopulation1[i])
else:
exPopulation1.append(newPopulation1[i])
count = count + 1
if newPopulation1[bestNum1] in exPopulation1:
exPopulation1.remove(newPopulation1[bestNum1])
count = 0
for i in range(len(newPopulation2)):
if i != bestNum2:
if i in exNum2:
if count < num:
exPopulation1.append(newPopulation2[i])
else:
exPopulation2.append(newPopulation2[i])
else:
exPopulation2.append(newPopulation2[i])
count = count + 1
if newPopulation2[bestNum2] in exPopulation2:
exPopulation2.remove(newPopulation2[bestNum2])
if distance1 < minDistance:
minItera = itera
minDistance = distance1
bestFit = bestFit1
bestPath =copy.copy(bestPath1)
if distance2 < minDistance:
minItera = itera
minDistance = distance2
bestFit = bestFit2
bestPath =copy.copy(bestPath2)
Population1 = copy.copy(exPopulation1)
Population2 = copy.copy(exPopulation2)
break
print minItera, bestPath, minDistance, bestFit
return minItera, bestPath, minDistance, bestFit