扰动环境下机械车间任务低碳配置机制多尺度解析与协同调控

Low-carbon tasks allocation of mechanical workshop, including batch planning of workshop level, task assignment of process unit level, and job sequencing of equipment level. It has the characteristics of multi-disturbance, multi-level tasks allocation, multi-mode workshop entity, uncertain change of carbon efficiency, etc., as well as the characteristics of high coupling, mutual restraint, dynamic strong correlation and time variability among elements. As a result, its correlation and reflection mechanism are complex and difficult to distinguish, leads to difficulties in analysis and regulation. From the perspective of relational scale, the carbon efficiency correlation and disturbance response mechanism of task allocation is resolved. The carbon efficiency coupling network model of tasks allocation and the model of disturbance information transfer chain are established, so that an association network is formed. From the perspective of behavior and rule scale, the entity reflection mechanism of low carbon tasks allocation in the workshop, the entity multimodal transformation and information output rules are constructed, and the modal and rules which controlled by tasks allocation are located. Thus, a dynamic model of low-carbon tasks allocation based on multi-resolution entities is constructed to adapt to the disturbance ,and accurately reflect the changing of carbon efficiency of multi-level tasks allocation. Intelligent algorithms are designed to connect the dynamic model. The cooperative control method of multi-level low-carbon tasks allocation of model and rule-driven is established. The research results will provide theoretical and methodological support for lean management of carbon efficiency, as well as the optimal low-carbon tasks allocation for high materials and energy efficiency in workshops.

机械车间任务低碳配置，包含车间层批量规划、工艺单元层任务分配、设备层作业顺序部署，具有多扰动、多层级任务配置、多模态车间实体、碳效率不确定变化等特点，以及要素间高度耦合性、相互制约性、动态强相关性与时变性等特性，造成其关联与反映机制复杂难辨，导致分析与调控困难。本项目从关系尺度，解析任务配置的碳效率关联与扰动响应机制，建立任务配置的碳效率耦合网络模型与扰动信息传递链模型，形成关联网络；从行为与规则尺度，解析任务低碳配置的车间实体反映机制，构建实体多模态转换与信息输出规则，定位任务配置控制的模态与规则，由此创建基于多分辨率实体的任务低碳配置动态模型，以适应扰动、准确反映多层级任务配置的碳效率变化；设计智能算法与任务配置模型联动，建立模型与规则驱动的多层级任务低碳配置协同调控方法。研究成果将为机械车间碳效率精益化管理、以及面向高物料与能源效率的车间任务低碳优化配置，提供理论与方法支持。

本项目针对我国机械车间低碳约束增强、个性化生产任务增多、扰动环境日益复杂的现状，提出了一套扰动环境下机械车间任务低碳配置机制多尺度解析与协同调控方法。通过逐步解析扰动环境下任务低碳配置的碳效率关联与扰动响应机制、任务低碳配置的实体反映机制，建立了基于复杂网络分析的机械车间任务配置碳效率精细化评估模型，提出了基于多触发器深度学习的扰动评估与任务配置更新决策方法，揭示了“扰动-任务配置-车间实体-碳效率”的关联与作用网络，形成了多尺度解析扰动环境下任务低碳配置机制的方法体系；针对各层级任务配置 “动态-一致”、“独立-协同”的运行特征，设计了各层级任务配置实体的模块化、层次化模型构架、实体模型结构与运行机制，提出了多重扰动下机械车间动态碳效率智能监测模型构建方法，实现了扰动环境下各层级任务配置碳效率变化的动态监测与分析，为扰动环境下任务低碳配置提供了模型与技术支持；进一步设计了一种高频扰动下机械车间在线任务配置的规则挖掘算法，构建了碳效率优化规则库，提出了一种基于PDEVS的车间在线任务配置规则性能动态评估模型，实现了模型与规则驱动的任务低碳配置协同调控，为扰动环境下任务低碳优化配置提供了应用支持。项目部分研究成果已在企业开展了应用，对于提高机械车间任务低碳配置的物料资源利用率与能源效率、降低生产成本，以及支持企业低碳新产品与再制造产品的研发设计、低碳工艺设计等具有重要意义。