论文部分内容阅读
炼钢-连铸排程是NP完全问题,文献中遗传算法、模拟退火等算法的计算速度无法满足进度控制的要求.目的就在于为改进现有算法提供标杆(benchmark).将任务委派和任务排序问题描述为黑球和白球的组合问题,大幅改进了从大量组合中找出可行组合的速度.以B企业两座转炉冶炼51炉为例,所给方法在CPU为双核4600+,内存为2 GHZ的主机上运行,仅用1.375 s从6 765 201个组合中,首先找出834个排列可行解,然后找出209个时间可行解,最后找出5个满意解,作为改进算法的标杆.
Steelmaking - continuous casting scheduling is a NP complete problem, the literature algorithms such as genetic algorithms, simulated annealing algorithm can not meet the schedule control requirements.The purpose is to improve the existing algorithm to provide a benchmark.Director delegation and tasks The ordering problem is described as the combination problem of black ball and white ball, which greatly improves the speed of finding a feasible combination from a large number of combinations.Based on the example of 51 furnaces of two smelters in B company, the given method is that the CPU is dual-core 4600+ and memory is 2 GHZ host, using only 1.375 s from 6 765 201 combinations, find 834 feasible solutions first, and then find 209 feasible solutions, and finally find 5 satisfactory solutions as the benchmark for improving the algorithm .