复杂工况下柔性加工车间的碳足迹规划与调度

Modern manufacturing industries have been facing on the global market completion and the environmental pressure. It prompts that the flexible manufacturing system should be lower carbon, more efficient and more stable. For the flexible job shop under complex circumstances, the carbon footprint is quantized. Using the low carbon, efficiency and stability as target, the meta-heuristic algorithm has been applied to establish Meta-Rule automatically. The proposed methodology is realized from three aspects. (1) To quantize the carbon footprint of the operation and construct the carbon footprint model. (2) To construct the complex flexible manufacturing system model based on the mixed integer non-linear programming. (3)To design the multi-dimension prediction-reactive strategy with self-adaptability to respond the uncertain events based on the robust scheduling, the scheduling amendment and the completely rescheduling. Based on the research results, the encoding mechanism of gene expression programming and the framework of pattern search are hybridized. The meta-heuristic algorithm with self-adaptability is proposed to construct Meta-Rule automatically. The Meta-Rule can generate the scheduling scheme quickly and reliably, and satisfy the requirement of the schedulers in the manufacturing company due to its convenience in learning and application. If successful, this project can optimize the carbon footprint and the efficiency, and improve the stability of the manufacturing system concurrently. It also promotes the sustainable development of the manufacturing industry in China.

在全球市场竞争加剧、资源环境管制增强的形势下，复杂制造系统在环境排放、生产效率和运行状态方面，面临着更严格的要求。 本项目面向复杂工况下柔性加工车间的作业调度，量化其碳足迹，以低碳、高效和平稳为目标，采用元启发式算法构造Meta-Rule。从三个层面实现：①量化工序加工过程的碳足迹，构建碳足迹计算模型；②基于混合整型非线性规划，耦合碳足迹与传统调度指标，建立多目标集成优化模型；③联接预防调度、调度修复与补救、完全重调度方法，研制具有容纳、消除或修复不确定扰动能力的多维度自适应预-反应调度策略。在此基础上，结合基因表达式编程的显式解析优势和模式搜索的全局搜索能力，研制自适应元启发式算法，挖掘出更贴近问题属性的Meta-Rule，以便快速可靠地生成调度方案，更好地满足企业需求。 复杂工况下柔性加工车间的碳足迹规划与调度，优化了生产过程碳效，保证了车间运作稳健，促进了中国制造业的可持续发展。

在全球市场竞争加剧、资源环境管制增强的形势下，复杂制造系统在环境排放、生产效率和运行状态方面，面临着更严格的要求。. 项目组围绕制造车间生产系统，从加工工艺、设备运行与辅助生产三个视角，设计了能源消耗关键特性框架，推导了制造车间能源消耗量化公式，建立了制造车间多目标能耗集成优化模型，采用GAMS软件获得了小规模案例的能耗下界，检验了模型的正确性与通用性，为后续生产调度模型提供理论支撑；其次，调研了制造车间的实际情况，分析并凝炼了与生产能耗相关的各种不确定因素，提出了相应的鲁棒对等转换与随机机会约束转换方法，缓冲与削弱了不确定事件产生的消极影响；在此基础上，研制能快速响应实际生产复杂工况的元启发式算法，包括混合粒子群算法、教与学优化算法等，以探索生产调度时性能优越的解的属性；最后，结合元启发式算法的全局搜索能力和基因表达式编程的显式解析优势，挖掘了更贴近生产实际、更简单易行的Meta-Rule。. 针对柔性加工车间，挖掘出12种Meta-Rule，针对具有小规模属性的50组案例测试时，获得了18组案例的下界，所求解与下界的偏差均小于0.32；针对具有大规模属性的480组测试案例表明Meta-Rule性能显著优于经典规则如SPT、LSO、SRM、MWKR和GPrule1，可节约13%的能耗，甚至在极个别测试案例中的总节能比高达67%。. 上述研究成果发表在《Computer &Operation research》、《Journal of Cleaner Production》《Computers & Industrial Engineering》《ICIC Express Letters》、《计算机集成与制造》及《系统工程理论与实践》等重要刊物发表论文24篇，其中SCI/EI收录14篇；培养了5名硕士研究生。

内容获取失败，请点击重试