高频扰动与多定制需求驱动的制造系统动态重构优化

Production line allocation and planning control of discrete manufacturing are sensitive to production disturbances such as emergency insertion, equipment failure and tool wear. Current virtual reconfiguration and scheduling schemes can not respond to all kinds of disturbances in time. Systematic modeling and intelligent optimization techniques lack in-depth research on Scheduling Optimization of reconfigurable systems, and has poor robustness as well as versatility. In most enterprises,virtual reconfiguration implementation of intelligent production line is limited to production line reconfiguration and resource allocation. The integration is deficient in terms of layout planning, real-time bottleneck analysis, dynamic capacity planning and other related technologies, and there is no general framework of virtual reconfiguration scheme.Reconfigurable virtual cell construction and scheduling technology tends to maturity, but it works in coordination with modeling and simulation needs to be explored in depth. Feasibility evaluation among reconfigurable schemes lacks quantitative indicators, and the process of reconfigurable optimization lacks clear direction. Therefore, this project intends to carry out research on dynamic reconfiguration optimization of manufacturing system driven by high frequency disturbance and customization requirements.Based on dynamic reconfiguration and optimization of intelligent production line,intelligent algorithm and core model could be realized , and the related prototype systems are planned to develop and validate.On the basis of the above,relevant core system module development and widely applied validation could be carried out subsequently.

离散制造的产线配置及计划控制对紧急插单、设备故障、刀具磨损等生产扰动敏感。当前的虚拟重构与调度方案无法及时响应各类扰动，系统化建模和智能优化技术在可重构系统调度优化层面缺乏深入研究，鲁棒性与通用性差。绝大多数企业对于智能产线虚拟重构的实施仅限于产线重构与资源配置，对于布局规划、实时瓶颈分析、产能动态规划等相关技术整合度差，缺乏通用的虚拟重构方案框架。可重构虚拟单元构建与调度技术趋于成熟，但与建模仿真的协同规划需深入研究，重构方案间可行性评估缺乏量化指标，重构优化过程缺乏明确方向指引。因此本项目拟开展高频扰动与定制需求驱动的制造系统动态重构优化研究，实现基于智能生产线动态重构与优化的智能算法与核心模型，进行相关原型系统开发，完成相关的核心系统模块研发并开展广泛应用验证。

高频扰动与多定制需求是离散制造面临的重大挑战，如何动态高效地响应系统扰动从而进行重构与调度优化，其关键在于系统化建模方法与智能化优化技术的深度融合，实现面向智能生产重构优化的核心模型与算法。本项目的研究成果主要包括以下四个方面：（1）揭示了三种生产模式下瓶颈分布规律与动态漂移机理，探析瓶颈指数模型以及基于瓶颈网络特征的典型离散制造系统优化调控方法，从核心模型、计划调度与控制机理等本质特征上覆盖复杂离散制造系统的显著特点，提出了基于瓶颈自适应调控的动态调度与重构优化方法。（2）建立了生产调度与物流相集成的全系统协同重构机制，支持计划调度与物流环节协同的全流程统一建模。围绕制造系统重构优化需求，基于模块化组合技术建立集成物流环节的多阶段调度模型，设计了面向制造系统运行配置及策略重构的多尺度仿真优化方法。（3）探索了多定制需求驱动的单元虚拟重构优化方法，针对多定制需求下瓶颈漂移特性建立产线动态产能分析模型，基于相似性设计了有效的产品任务分组方法，在混流制造单元层次上对多工艺路径下的虚拟动态单元进行划分重组，提出了多品种混流路径重构规划技术思路。（4）提出了高频扰动下的智能产线自适应重构优化方法，面向高频扰动环境下模糊性与随机性，研究基于混合智能的模糊推理系统与数据驱动的自动化仿真建模技术，集成改进的多目标群智能优化算法，通过系统配置与调度策略调控实现系统总体效能优化。本项目主要研究成果包括发表标注项目资助的论文17篇，其中14篇SCI源刊和3篇中文核心，申请发明专利6项，授权5项。本项目研究丰富了高频扰动与多定制需求驱动的制造系统动态重构优化理论，相关成果分别在南方航发、胜利精密、兴森快捷等企业获得局部应用验证，展现出了在制造系统智能重构及运营优化领域的广阔应用前景。