圆钢热轧生产计划优化方法研究

发布时间：2018-05-02 02:34

本文选题：热轧计划 + 圆钢生产　；参考：《北京科技大学》2015年博士论文

【摘要】：圆钢作为一类重要的钢铁产品,因其热轧生产过程的特殊性,使得其在编制热轧计划时需要考虑的因素较多,难度较大。为了实现圆钢热轧生产的精细化管控,编制一个合理有效的热轧计划在理论探索和实际应用方面都具有十分重要意义。迄今为止,结合圆钢特点的热轧计划问题的研究成果还相对较少,且因其轧制过程的特殊性,已有理论和方法不具有适用性。因此,本文结合圆钢的热轧生产特点,围绕静态和动态环境下圆钢热轧计划的问题特征和求解算法展开探索性研究。主要研究内容和创新点如下： (1)为有效减少圆钢热轧生产过程中的余料浪费,研究了面向订货需求的圆钢热轧批量计划问题,建立了以最小化钢坯余料浪费和相邻钢坯间属性跳跃惩罚为目标的数学模型,并基于约束满足技术设计了求解算法,来确定轧制单元、钢坯与圆钢之间的关系。算法通过变量选择和值选择规则对待排产钢坯进行选择、分组和序列生成操作；通过约束传播技术缩减搜索空间,并划分轧制单元；同时,将装箱启发式BFD嵌入到算法中,以优化钢坯余料浪费和钢坯属性跳跃惩罚。基于实际生产数据的仿真实验表明了模型和算法的可行性和有效性。 (2)考虑圆钢热轧生产过程中工艺规程和交货期对连续加工的影响,建立了以最小化设备调整时间、拖期生产惩罚和钢种跳跃惩罚为优化目标的圆钢热轧批量调度问题模型。针对问题的多目标特征,提出了串行优化策略,并设计了嵌入EDD规则和钢种排序规则的变邻域搜索算法。算法首先计算轧制单元的交货期,然后采用约束满足技术生成初始解；根据实际生产需求,将最小化设备调整时间作为主要目标,设计变邻域搜索算法实现目标优化,其中,运用混合算子构造邻域结构和局部搜索,并引入模拟退火接受准则来控制迭代过程中产生的新解；同时,为了最小化拖期惩罚和钢种跳跃惩罚,在求解过程中嵌入了EDD规则以及钢种排序规则。实验结果表明,模型和算法是可行且有效的。 (3)考虑圆钢热轧过程中轧辊使用寿命动态变化的因素,分析与轧辊相关的优化指标和约束条件,建立了以最小化设备调整时间和最大化轧辊连续轧制长度为优化目标的数学模型,并设计了基于Pareto最优的多目标变邻域搜索算法。算法首先根据轧制规程对钢坯集合和轧辊集合进行预处理操作;针对问题的多目标特征,构造基解选择策略来引导算法的搜索过程；引入基于分解与重构思想的后优化过程,增强非支配解的多样性,并根据轧制规程重新分配轧辊,基于Pareto最优思想评价解的优劣,进而更新全局最优集。基于实际生产数据的仿真实验结果表明了模型和算法的有效性。 (4)针对动态环境下圆钢热轧滚动计划优化问题,分析了随机事件对计划执行过程的扰动,提出了基于关键事件驱动的滚动优化策略。在此基础上,为保持与初始计划在产能和生产效率上的一致性,建立了以最小化滚动前后热轧计划长度的差异性为优化目标的动态约束满足模型,并提出了一种基于局部修复策略的动态约束满足算法。算法结合圆钢热轧的工艺特征,设计了变量选择和值选择规则,将待排产钢坯集合排入到失去可行值的轧制单元集合；在求解过程中,采用约束传播技术约减变量的初始值域,基于约束检查和回跳机制消解冲突约束,并用局部修复方法进一步优化滚动计划方案。仿真实验表明,算法在求解质量和求解效率方面都具有良好的效果。
[Abstract]:As a kind of important iron and steel products , round steel is a kind of important iron and steel products , because of its particularity in hot rolling production process , it is very important to develop a reasonable and effective hot rolling plan in the development of hot rolling plan . So far , the research results of hot rolling plan with the characteristics of round steel have less applicability . So far , with the characteristics of hot rolling production of round steel , the problem feature and solving algorithm of round steel hot rolling plan in static and dynamic environment are studied .

( 1 ) To effectively reduce the waste of spare materials during hot rolling production of round steel , a mathematical model is established to minimize the waste of steel billet and the property jump penalty between adjacent billet , and the relation between rolling unit , billet and round steel is determined based on constraint satisfaction technique .
the search space is reduced by the constraint propagation technology , and the rolling unit is divided ;
At the same time , the box heuristic method is embedded in the algorithm to optimize the billet waste and billet attribute jump penalty . Simulation experiments based on the actual production data show the feasibility and effectiveness of the model and algorithm .

( 2 ) In consideration of the influence of the technological procedure and the delivery time on the continuous processing in the hot rolling production process of round steel , a round steel hot rolling batch scheduling problem model is established to minimize the adjustment time of the equipment , the production penalty of the tow period and the skip penalty of the steel . In this paper , a serial optimization strategy is proposed , and a variable neighborhood search algorithm is designed to embed the edd rule and the steel ordering rule . The algorithm firstly calculates the delivery time of the rolling unit and then uses the constraint satisfaction technique to generate initial solution ;
According to the actual production demand , it will minimize the adjustment time of the equipment as the main objective , and design the neighborhood search algorithm to achieve the objective optimization , in which the neighborhood structure and local search are constructed by using the hybrid operator , and the simulated annealing acceptance criterion is introduced to control the new solution generated in the iteration process ;
At the same time , in order to minimize the penalty of tow - stage punishment and steel - seed jump , the rules of edd and the ordering of steel seeds are embedded in the solving process . The experimental results show that the model and algorithm are feasible and effective .

( 3 ) In consideration of the factors of the dynamic change of the roller ' s service life in the hot rolling process of the round steel , the optimization indexes and the constraint conditions related to the roll are analyzed , and a multi - objective variable neighborhood search algorithm based on Pareto optimal is established . First , the steel billet set and the set of rolls are preprocessed according to the rolling procedure , and the searching process of the algorithm is guided by constructing the basis solution selection strategy according to the multi - objective characteristic of the problem ;
A post - optimization process based on decomposition and reconstruction is introduced , the diversity of non - dominant solutions is enhanced , and the optimal set is updated based on Pareto optimal thought evaluation . The simulation results based on the actual production data show the validity of the model and algorithm .

( 4 ) Aiming at the optimization problem of rolling plan of round steel under dynamic environment , this paper analyzes the disturbance of random event to plan execution process , and puts forward a rolling optimization strategy based on key event driven . Based on this , a dynamic constraint satisfaction model based on the difference of the length of rolling plan before and after rolling is established , and a dynamic constraint satisfaction algorithm based on local repair strategy is established .
In the process of solving , the initial value domain of the reduction variable is adopted by the constraint propagation technique . Based on the constraint checking and back - jump mechanism , the conflict constraint is eliminated , and the rolling plan scheme is further optimized by the local repair method . The simulation experiment shows that the algorithm has a good effect in solving the quality and solving the efficiency .

【学位授予单位】：北京科技大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TG335;TG338

【参考文献】