【作者】 宁树实；

【导师】 王伟；

【作者基本信息】 大连理工大学 ， 控制理论与控制工程， 2006， 博士

【摘要】 本文综述了钢铁企业CIMS中MES的重要作用和钢铁行业生产调度研究的现状以及一体化生产调度的特点，在此基础上，基于国家“863计划”重点课题，结合我国钢铁企业生产实际情况，针对钢铁企业炼钢一连铸一热轧一体化生产调度中存在的若干问题，进行了下列研究。 提出了一体化生产计划体系下的热轧批量计划编制问题的数学模型及其算法。模型以热轧生产实际要求的单元计划平均轧制长度和热装比作为优化目标，将生产实际中的时间和温度要求作为约束条件。根据实际生产中优先考虑单元计划平均轧制长度的情况，构造了用于热轧批量计划编制的两阶段变邻域搜索算法，首先以单元计划平均轧制长度作为优化目标，然后以热装比作为优化目标求解。 建立了一体化生产计划体系下的炼钢—连铸生产计划编制问题的数学模型，并根据炼钢和连铸工序衔接紧密的特点，提出了一种统一求解炼钢—连铸生产计划的方法，使得炼钢炉次计划编制和连铸浇次计划编制交互进行而不是分别孤立求解，达到两种计划同时优化的效果。首先根据炼钢计划编制的多目标优化模型求出炼钢炉次计划的候选解集，然后以此作为输入条件，根据连铸浇次计划编制的数学模型求出连铸浇次计划，最后再由浇次计划确定炉次计划的生产顺序。求解炼钢—连铸生产计划编制问题的过程中使用了多目标模拟退火算法和变邻域搜索算法。 对于处于连铸和热轧两道工序之间的重要热工设备——步进式连续加热炉的生产调度问题进行了研究，建立了以最小化板坯达到轧制温度后在加热炉中停留的总时间作为优化目标的数学模型，以在保证满足热轧机生产需求的前提下尽量降低加热炉运行的能量消耗。同时构造了用于求解加热炉调度问题的遗传局部搜索算法。 提出基于准时制的动态调度思想并据此建立了炼钢—连铸生产动态调度的数学模型。为了最大限度减少动态调度对于生产全流程的影响，基于准时制的动态调度以减少动态调度方案与原调度方案的时间偏差作为优化目标。求解算法采用了一种基于种群进化的进化迭代局部搜索算法。 所建立的生产调度数学模型和相应的算法均经过生产实际数据验证，并在此基础上开发了炼钢—连铸—热轧—体化生产调度仿真系统，系统运行表明可以提高钢铁企业生产管理水平，达到了节能、降耗和增效的目的。更多还原

【Abstract】 In this dissertation, the importance of the MES in the CIMS of Iron & Steel enterprise and the main role of production scheduling in the MES are reviewed. The sate of art of the steelmaking production scheduling is also reviewed and the characters of integrated production scheduling are analyzed. The research of this dissertation is supported by National High-Tech. R&D Program. Some problems existed in steelmaking production integrated scheduling are studied and solved by considering the actual production process in Iron & Steel enterprises. This dissertation has mainly carried on the following research.A mathematical model and an algorithm for hot rolling lot planning in integrated production scheduling system are proposed. The model is proposed with the aims of promoting average rolling length of rolling turn and increasing hot charge rate, and the time constraint and temperature constraint in industrial practices are considered. According to the fact that the average rolling length of rolling turn is considered as more important than hot charge rate, a two-stage variable neighborhood search algorithm is designed to solve the problem. The algorithm solves the problem by optimizing the average rolling length firstly and optimizing hot charge rate secondly.Mathematical models for steelmaking and continuous casting planning are formulated. An integrated method is proposed for the two planning problems considering the close relationship between steelmaking and continuous casting, so that the steelmaking plans and continuous casting plans are not solved separately, and global optimization is achieved. Firstly a set of potential furnace charge plans are obtained through solving a multi-objective model of furnace charge planning. Then the casting plans can be obtained by the solutions of mathematical model of continuous casting planning with the objective of minimizing the number of casting plans. Finally the furnace charge plans can be confirmed according to the casting plans. Multi-objective simulated annealing algorithm and variable neighborhood search algorithm are designed to solve the problems.Reheat furnace is important equipment between continuous casting machine and hot rolling machine. The reheat furnace scheduling problem is researched and mathematical model for the problem is formulated. Objective of the model is to minimize total extra time of slabs kept in reheat furnace after reaching the rolling temperature. As a result the energy更多还原