搬运机器人路径规划与柔性作业车间生产调度集成优化研究

As an automation device that integrates loading-unloading functions and transport functions, transfer robot (TR) is applied in more and more modern flexible job shops, but the traditional production scheduling of flexible job shops only considers processing machines, processes and so on, it is independent of the path planning of TRs. So, this project mainly take research on the integrated optimization for path planning of TRs and production scheduling of flexible job shops from the aspects of mechanism layer, method layer, and application layer. In the mechanism layer, the classification and description methods of TRs properties and manufacturing task constraints based on improved K-Means algorithm are proposed, a mapping relationship representation mechanism model of them based on BPMN is constructed, and the corresponding mapping relationship knowledge base is built. In the method layer, the integrated optimization method is studied, which contains building the integrated optimization model based on mapping relationship knowledge base, designing the multi-objective optimization allocation method based on the strategy of Nash equilibrium game, developping dynamic scheduling strategy, designing hybrid algorithm, constructing evaluation and fault tolerance mechanism of integrated optimization method. In the application layer, a set of integrated scheduling simulation platform is designed and developed to carry out theoretical verification and practical application research. This project will provide theoretical and methodological support for integrated optimization for path planning of transfer robots and production scheduling of flexible job shop, and further enrich the existing scheduling optimization methods.

搬运机器人作为集装卸功能和运输功能为一体的自动化设备，被应用于越来越多的现代柔性作业车间中，而传统型柔性作业车间生产调度只考虑加工设备、工序等，与搬运机器人路径规划相独立。因此，本项目从机理层、方法层、应用层三个方面开展搬运机器人路径规划与柔性作业车间生产调度集成优化研究。在机理层，提出基于改进K-Means算法的搬运机器人属性与制造任务约束分类描述方法，构建基于BPMN的二者映射关系表征机理模型，并建立对应的映射关系知识库。在方法层，研究集成优化方法，包括搭建基于映射关系知识库的集成优化模型、设计基于Nash均衡博弈策略的多目标优化配置方法、制定动态调度策略、改进混合求解算法、构建集成优化方法评估机制和容错处理机制。在应用层，设计和开发一套集成调度仿真平台，开展理论验证和实践应用。本项目将为搬运机器人路径优化与柔性作业车间调度集成优化研究提供理论和方法支撑，进一步丰富现有调度优化方法。

搬运机器人作为集装卸功能和运输功能为一体的自动化设备，被应用于越来越多的现代柔性作业车间中，而传统型柔性作业车间生产调度只考虑加工设备、工序等，与搬运机器人路径规划相独立。因此，项目在前期研究基础上，通过理论分析、建模仿真、实例验证等相结合发方法，开展了搬运机器人路径规划与柔性作业车间生产调度集成优化研究。建立了基于改进K-means++算法的制造资源聚合模型，以描述搬运机器人和柔性作业车间制造任务属性。构建了实现AGV路径规划与柔性作业车间调度集成优化的融合调度模型，提出基于最先服务原则的AGV安排策略，并通过改进混合遗传鲸鱼优化算法对模型完成了求解；同时，建立了基于“请求-调度-反馈”模式的智能制造车间AGV动态规划模型，通过集成双层混合遗传算法和蚁群优化算法实现了模型求解，并将其与物联网集成来满足动态和实时的需求。另外，构建了面向数字孪生车间的制造任务语义建模和制造资源服务推荐框架，设计和开发了柔性作业车间APS管控系统，通过嵌入模型库、算法程序集为企业实际排产和调度提供决策支持，丰富了现有柔性作业车间AGV集成融合调度理论体系。