摘要
针对热处理调度中分炉分批任务,先利用数据挖掘方法,从历史案例中提取有价值的知识信息;再将知识运用于案例推理过程,利用规则引擎对待处理任务进行分炉分批求解,得到新实例;最后根据新产生的应用实例,对案例库进行扩充,实现从"案例-知识-规则-应用实例-案例"的闭环流程。历史记录和新产生的实例可用于生成并丰富知识,使得推理准确率随着知识的积累不断提升。最后通过实例验证了该方案的可行性与可用性。
To the batching tasks in heat treatment scheduling,this paper introduced a set of knowledge that can be extracted from historical cases by data mining techniques from the perspective of case-based reasoning.In addition,the rule templates used for rule engine and the reasoning routines were designed,and new batches are got after calculation.New batches will be added to historical cases.The whole procedure is a closed-loop process that is from "Case"to"Knowledge"to"Rule"to"Inferenced Batch"and back to"Case".The newly generated cases can enrich knowledge,so that the accuracy of this algorithm will increase with time.Finally,an example is used to verify the feasibility and availability of the solution.
引文
[1]佚名.热处理行业“十三五”发展规划[J].热处理技术与装备,2015(4):63-76.
[2] Ram B,Patel G.Modeling furnace operations using simulation and heuristics[C]//Simulation Conference Proceedings.[s.l.]:[s.n.],1999:957-963.
[3]张晓平.钢铁企业罩式炉装炉优化及煤气柜位预测问题研究与应用[D].大连:大连理工大学,2010:41-46.
[4]刘颖,刘全利,王伟.罩式退火炉装炉组合的优化模型及其算法[J].信息与控制,2009,38(2):211-217.
[5]刘荣花.基于拉格朗日松弛的罩式炉排产优化问题的研究[D].沈阳:东北大学,2006:48-50.
[6]朱戈.大型煤气炉合炉热处理工艺的研究[J].应用能源技术,2013(8):37-39.