基于改进单亲遗传算法的炼钢最优炉次计划模型(英文)

王闯; 刘青; 李庆益; 王彬; 谢飞鸣; 王柏琳

引用本文:	王闯,刘青,李庆益,王彬,谢飞鸣,王柏琳.基于改进单亲遗传算法的炼钢最优炉次计划模型(英文)[J].控制理论与应用,2013,30(6):734~741.[点击复制]
	WANG Chuang,LIU Qing,LI Qing-yi,WANG Bin,XIE Fei-ming,WANG Bai-lin.Optimal charge plan model for steelmaking based on modified partheno-genetic algorithm[J].Control Theory and Technology,2013,30(6):734~741.[点击复制]

基于改进单亲遗传算法的炼钢最优炉次计划模型(英文)

Optimal charge plan model for steelmaking based on modified partheno-genetic algorithm

摘要点击 2240 全文点击 2428 投稿时间：2012-06-05 修订日期：2012-10-29

查看全文查看/发表评论下载PDF阅读器

DOI编号 10.7641/CTA.2013.12128

2013,30(6):734-741

中文关键词炼钢最优炉次计划模型改进单亲遗传算法仿真

英文关键词 steelmaking optimal charge plan model modified PGA simulation

基金项目 This work is supported by the Specialized Research Fund for the Doctoral Program of the Ministry of Education, China (No. 20090006110024), and the Fundamental Research Funds for Central Universities (No. FRF–BR–09–020B).

作者	单位	E-mail
王闯	北京科技大学钢铁冶金新技术国家重点实验室北京科技大学冶金与生态工程学院	wld059014197@163.com
刘青^*	北京科技大学钢铁冶金新技术国家重点实验室北京科技大学冶金与生态工程学院	qliu@ustb.edu.cn
李庆益	北京科技大学钢铁冶金新技术国家重点实验室北京科技大学冶金与生态工程学院
王彬	北京科技大学钢铁冶金新技术国家重点实验室北京科技大学冶金与生态工程学院
谢飞鸣	方大特钢科技股份有限公司
王柏琳	北京科技大学冶金与生态工程学院北京科技大学东凌经济管理学院

中文摘要

炉次计划在炼钢生产计划的编制过程中扮演着重要角色, 优化的炉次计划对炼钢厂的高效、稳定运行产生深远影响. 基于已有文献, 并根据小方坯连铸过程的特点, 考虑了钢种、断面、交货期等因素, 建立了新的炉次计划模型, 以期通过优化生产合同的组合而降低生产费用. 炉次计划问题是复杂的组合优化问题, 不可能在列举所有可能的求解结果. 因而, 采用了改进的单亲遗传算法寻求问题的最优/近优解. 在求解过程中, 通过分析比较, 得到了合理的算法参数. 最后, 通过采用遗传算法、单亲遗传算法和改进的单亲遗传算法对模型求解结果的比较, 验证了改进后单亲遗传算法的优越性.

英文摘要

The charge plan plays an important role in compiling the production plan for steelmaking, and an optimized charge plan will have a far-reaching effect on the stability and efficiency of a steelmaking workshop. According to the characteristics of billet continuous casting process, a new charge plan model is developed by taking into account three constraints: steel grades, dimensions, and due dates. By using this model, we reduce the production cost by optimizing the sequencing of the contracts. The problem is combinatorial in nature; the complete enumeration of all its possibilities is computationally prohibitive. Therefore, a modified partheno-genetic algorithm (PGA) is employed to search optimum/nearoptimum solutions. During the solving process, reasonable algorithm parameters are acquired through the analysis and comparison of different relative parameters. Furthermore, a comparative analysis of the results obtained by implementing the genetic algorithm (GA), PGA and modified PGA on the proposed model reveals that modified PGA outperforms GA or PGA in solving the charge planning problem.