CDS启发式算法

参考

gatt、makespan、Johnson启发式算法

CDS启发式算法

# -*- coding:utf-8 -*-
import numpy as np
import time
from .tool import makespan_value
from .johnson import johnson


class CDS:
	def group(self, data):
		data_group = np.zeros([data.shape[0] - 2, 3, data.shape[1]])
		for i in range(data_group.shape[0]):
			data_group[i, 0] = data[0]
			for j in range(data.shape[1]):
				data_group[i, 1, j] = np.sum(data[1:i + 2, j])
				data_group[i, 2, j] = np.sum(data[-i - 1:, j])
		return data_group
	
	def johnson(self, data_group):
		data_johnson = np.zeros([data_group.shape[0], data_group.shape[2]])
		for i in range(data_group.shape[0]):
			data_johnson[i] = johnson(data_group[i])
		return data_johnson
	
	def select(self, data, data_johnson):
		data_johnson = np.array(data_johnson, dtype=int) - 1
		data_best = data_johnson[0]
		for i in range(1, data_johnson.shape[0]):
			if makespan_value(data[:, data_best]) > makespan_value(data[:, data_johnson[i]]):
				data_best = data_johnson[i]
		data_best += 1
		return data_best


def cds(data, draw=0):
	"""
	:param data: n行m列，第一行工序编号，其他是加工时间
	:return:
	"""
	data = data[:, np.argsort(data[0])]
	new = CDS()
	start_time = time.time()
	data_group = new.group(data)
	data_johnson = new.johnson(data_group)
	data_best = new.select(data, data_johnson)
	end_time = time.time()
	print("Time used: %s" % (end_time - start_time))
	print("The minimum makespan: %s" % makespan_value(data[:, data_best - 1]))
	if draw:
		import matplotlib.pyplot as plt
		from .tool import gatt
		gatt(data[:, data_best - 1])
		plt.show()
	return data_best