在物流领域中，仓储作为订单完成作业中必不可少的一环，其研究内容逐渐从纯人工拣选拓展到机器人拣选等。随着智能化的发展与对人因的考虑，人机协同模式出现在大众视野中，该模式能有效解决机器人柔性低拓展性差，而纯人工效率低成本高的弊端。本文针对随机存储策略下的人机协同仓储系统开展了行程时间模型的推导，同时搭建了仿真平台，对行程时间模型进行验证对比的同时研究了该系统在不同场景下的性能表现，文章具体可分为以下内容：（1）人机协同系统的文献综述与系统流程分析：本文讨论了国内外机器人仓储系统的发展情况，指出人机协同系统在学术领域上的缺失，通过分析系统的组成架构与运作流程，明确了订单完成的所有步骤，对后续订单周转时间等系统表现的期望计算奠定基础。（2）人机协同拣选的协作模式与行程时间模型建模：根据订单的绑定状态，提出了四种可能的协作模式，基础模式、区域负责制下的人机一对多与人机多对一和理想状态下的就近匹配制。本文先使用均匀分布的次序统计量，在拣选员待命点在巷道端点的情况下，对订单周转时间与拣选员有效行走距离进行期望的求解。（3）人机协同拣选的离散型行程时间模型：针对拣选员待命点在上批次结束点的情况，使用等概率的离散分布的次序统计量对系统关键指标进行计算，通过仿真验证该模型的科学性。（4）基于人机协同系统的数值实验：搭建人机协同的仿真平台，针对订单周转时间、等待时长、服务时长、拣选员行走距离等系统性能指标，修改订单到达率、人机配比、批次订单大小等参数进行数值实验，得到不同场景下的适用协作模式。本文的研究为仓库的规划设计提供理论依据与辅助决策，同时填补了仓储领域中人机协同拣选方式的空白，利用次序统计量规定订单拣选顺序给出了该系统在不同协作模式下的连续与离散的行程时间模型，具有较高的学术意义与工程价值。
In the field of logistics, warehousing is an indispensable part of order fulfillment. The research content is gradually expanded from pure manual picking to automatic picking. With the development of intelligence and the consideration of human factors, the human robot cooperation mode appears in the public view. This mode can effectively overcome the disadvantages of low flexibility, poor expansibility, low work efficiency, and high cost.In this paper, the travel time model of the collaborative picking system under the random storage strategy is deduced, and a simulation platform is built to verify the accuracy of the model. The performance of the system in different scenarios is studied by the simulation as well. The article can be divided into the following contents:(1) Literature review and system process analysis of collaborative picking system: This paper discusses the development of robotic storage systems in china and abroad, points out the lack of collaborative picking system in the academic field, and defines all steps of order fulfillment by analyzing the composition and operation process of the system and lays the foundation for the calculation of expectations such as order turnover time.(2) Cooperation modes of Collaborative picking system and continuous travel time model: According to the binding state of orders, four possible cooperation modes are proposed: basic mode, human-robot one-to-many, and many-to-one under regional responsibility mode, and nearest matching mode under ideal state. Firstly, this paper uses theorder statistics of uniform distribution to solve the expected order turnover time and the effective walking distance of the picker when the picker dwell point is at the end of the aisle.(3) Discrete travel time model of collaborative picking system: Based on the assumption that the picker dwell point is the last storage position of the previous batch, the order statistics of discrete distribution with equal probability are used to calculate the key indexes of the system, and the scientificity of the model is verified by simulation.(4) Numerical experiment based on collaborative picking system: We build a simulation platform to compare the performance of four cooperation modes with different order arrival rates, human-robot ratios, and batch order size and obtain the suitable cooperation modes in different scenarios.The research of this paper provides a theoretical basis and auxiliary decision-making for the planning and design of warehouses. At the same time, it fills the gap in the collaborative picking system in the field of warehousing. The order statistics are used to specify the order picking list, and the related expectations under different cooperation modes in both continuous and discrete are given, which has high academic significance and engineering value.
