分时电价下配备自备电厂的钢铁制造业柔性调度方法

In this proposal, we study the problem of flexible scheduling of steelmaking industry with self-generation power plant under time-of-use electricity tariffs. It is the condition that the issue considers three kinds of electricity price as time-of-use electricity tariffs, the on-grid electrovalence and the self-generating power cost simultaneously that causes the rapidly increasing complexity of relation between electricity price, power generation and consumption as well as load demand and time. So there are some new challenges different from the traditional scheduling problems: three kinds of electricity price should be considered at the same time, the ability of response behaviors of power generation and consumption needs to be promoted urgently, the flexibility of scheduling scheme is required to be higher, machine restart decision mechanism becomes more complicated and factors like algorithm complexity and efficiency needed considering in modeling and solving the scheduling problem, etc. Aiming at these new challenges, we plan to conduct the research in four aspects: (1) method of flexible response behaviors of power generation and consumption of steelmaking industry with self-generation power plant under time-of-use electricity tariffs; (2) integrated flexible scheduling methods for steelmaking industry under new time-of-use electricity tariffs; (3) heuristic rule, quality evaluation and tradeoff method of solutions; (4) the application and validation of models and algorithms in steelmaking enterprises. The proposal will establish theories and methods of flexible scheduling for steelmaking industry with self-generation power plant under time-of-use electricity tariffs to solve a matching problem of the decision object and time object with different information granularity, so as to promote the ability of steelmaking industry to optimize the allocation of resource, reduce the operation cost and improve scheduling decision-making level of China's steelmaking industry.

本项目以分时电价下配备自备电厂的钢铁制造业柔性调度问题为研究背景。该问题由于同时具有分时电价、上网电价以及自发电成本三种电价导致电价、发用电量、负荷需求与时间之间的关系复杂性骤增，因此具备一些不同于传统调度问题的新挑战：需同时考虑三种电价、发用电响应能力亟须提升、调度方案的柔性要求更高、设备重启决策机制更加复杂以及调度问题的建模与求解需考虑复杂度、求解效率等。针对这些新挑战，本项目拟从四个方面展开研究：(1)分时电价下配备自备电厂的钢铁制造业发用电柔性响应方法；(2)新分时电价下钢铁制造业一体化柔性调度方法；(3)启发式规则、解的质量评价与权衡方法；(4)模型、算法在企业中的应用验证。本项目将建立分时电价下配备自备电厂的钢铁制造业柔性调度理论与方法，解决在多粒度时间对象下多粒度决策对象的资源优化配置问题，从而降低我国钢铁制造业的成本、提高调度决策水平。

本项目以分时电价下配备自备电厂的钢铁制造业柔性调度问题为研究背景，从四个方面展开了研究：(1)分时电价下配备自备电厂的钢铁制造业发用电柔性响应方法；(2)新分时电价下钢铁制造业一体化柔性调度方法；(3)启发式规则、解的质量评价与权衡方法；(4)模型、算法在企业中的应用验证。. 项目取得了如下研究成果：提出了一种配备自产电设备的钢铁企业分时电价下高效调度方法；一种分时电价下多工位炼钢-连铸调度方法；一种分时电价下钢铁企业的电力负荷跟踪调度方法；一种求解设置时间和运输时间相关的节能型柔性作业车间调度问题的启发式方法；一种基于分时电价的两阶段并行速度调整机器调度方法；一种带缓冲器的多状态可修可重构制造系统的动态和稳态性能分析方法；一种新的求解分时电价下单机调度的带多级过滤机制的贪心插入启发式算法；一种考虑能耗的柔性作业车间调度与多排布局集成的多目标双层规划方法；一种基于奇异值惩罚对抗域自适应的滚动轴承故障诊断方法；一种基于增强生成对抗网络的滚动轴承不平衡故障诊断方法；一种新的自适应混合马尔可夫话题模型在社交网络中兴趣点推荐方法；一种基于协同过滤的新用户冷启动推荐相似度计算方法。. 本项目将分时电价、自发电成本以及上网电价同时引入钢铁制造业的调度决策中，提出了分时电价下的钢铁制造业柔性调度决策理论与方法。该研究成果亦能在其他高能耗企业中推广应用。将运作管理、能源管理、智能优化及机器学习等理论相结合，系统研究了分时电价下的钢铁制造业发用电柔性响应方法、一体化柔性调度模型、启发式规则以及算法设计。解决了在多粒度时间对象下多粒度决策对象的资源优化配置问题，从而降低了我国钢铁制造业的成本、提高了调度决策水平。