Research On Key Problems For Optimizing Operation Of Robotic Mobile Fulfillment Systems

With the development of social economy and the progress of society,the consumers' consumption consciousness is upgrading and promoting the integration of online and offline retailing and the emergence of new retailing business model,which has brought huge opportunities but also challenges to logistics industry,and has accelerated its transformation.The warehousing is a key portion of logistics,involving the storage,delivering of commodities and the information circulation,which has a significant impact on the operating costs and operational efficiency of the entire logistics system.In the situation where the costs of resources such as land and labor continue to rise,and consumers demand more timeliness for commodity delivering,traditional logistics and warehousing systems have gradually revealed the defects of low space utilization and low operating efficiency,which is unable to meet the needs of new logistics.Fortunately,automated storage systems have effectively overcome these shortcomings and have gradually attracted the attention of enterprises and scholars,and also have seen widely application.Specifically,the robotic mobile fulfillment system adopts automatic guided vehicles to transport products,which saves the pickers' walking time and greatly improves the order picking efficiency in warehousing systems.And the storage and picking devices of the system are deployed on the ground,which makes it convenient for the adjustment of system layout and configuration,contributing to high flexibility and scalability of system.Besides,the order picking operation is carried out on the ground,making the human-device interaction more convenient and dedicating to improving the system's operation flexibility.At present,most researches on the robotic mobile fulfillment system are focused on the optimization of the classic Kiva system,while few optimize the system based on significantly improving and re-designing the system.However,the overall upgrade and optimization of the current robotic mobile fulfillment system is of great importance for adapting to the new logistics delivering mode and new logistics system in the context of New retailing business model.Therefore,this thesis focuses on the goal of optimizing the operation of robotic mobile fulfillment system,and studies several key optimization-related issues,including the optimization of picking device design,the system configuration and layout design and the order picking strategies.First,the thesis studies the design optimization of driving mode of order picking device,namely the automatic guided vehicle.The previous vehicles used in robotic mobile fulfillment systems are all differential driven vehicles.This study designs a diagonally installed double-steering driving mode for automatic vehicles,and a feasible motion control strategy is proposed for the double-steering vehicle based on the Ackerman steering geometry theory.In addition,a trajectory tracking algorithm based on the model predictive control is designed for the vehicle.The analysis shows that,compared to differential drive vehicles,the improved double-steering vehicle requires narrower paths,and has shorter turning time and higher turning stability.These features are conductive to improving the space utilization and overall operating efficiency of robotic mobile fulfillment systems,as well as the system stability.A real double-steering vehicle was built to carry out actual trajectory tracking experiments,which verified the effectiveness of the diagonally installed double-steering driving mode.Besides,the trajectory tracking results show that,the proposed double-steering vehicle can obtain accurate trajectory tracking performance by using the model predictive control-based trajectory tracking algorithm.The parking accuracy of the vehicle can meet the requirements of robotic mobile fulfillment systems.Second,the optimization problem of the design of robotic mobile fulfillment systems which use the "shelf-to-picker" operation mode is studied and a modular layout design for small and medium-sized systems is proposed,which adopts the operating mode of binding vehicles and aisles.The mode can effectively avoid traffic congestion.In addition,a buffer station is designed for the picking area to achieve the parallel operation of vehicles and pickers,which improves working efficiency.This study builds a theoretical model for the modular robotic mobile fulfillment system,to evaluate the system performance.Based on the model,the study proposes two methods for rapidly optimizing the system design,namely "Bisection method" and"Two-stage method".These two methods can rapidly determine an optimized system design solution which can meet order picking demands and costs lower.The effectiveness and accuracy of the theoretical mode are validated by discrete event simulation experiments in the research.Through the case study of actual system design,the effectiveness of the two optimization methods for system design is verified.The experimental results show that,"Two-stage method" is more suitable for the design optimization of small-scale systems."Bisection method" performs more stable for the design optimization of different scale systems,so it shows obvious advantages in large-scale systems design optimization.Third,the optimization problem of the design of robotic mobile fulfillment systems which use the "bin-to-picker" operation mode is studied.The "bin-to-picker"system introduced in this study uses high-density shelves to store products,and uses vehicles equipped with a lift and storage shelf to transport bins in batches to picking station for picking.An S-type accelerating method is proposed for vehicles,to enhance their stability during moving,and a new path design solution is proposed for the storage area of the system,to adapt to the batch order picking method.Besides,considering the batch service mode of the system,a semi-open queueing network model is formulated to evaluate system performance.In the research,the effectiveness of the theoretical mode is validated by simulation experiments through Arena.Through numerical analysis,the influence of vehicles'picking batch size,the quantity ratio of vehicles to picking stations,and the layout of storage area on system performance are studied,which provides theoretical support for the optimization of these system design related parametersFinally,the optimization problem of order picking strategies of robotic mobile fulfillment systems which use the "bin-to-picker" operation mode is studied.To begin with,an optimized order batching strategy is proposed.An integer programming model for the order batching problem within an order wave is formulated,considering the maximum number of orders that a picking station can handle at a time.A two-stage heuristic algorithm based on item coincidence degree of orders is designed to obtain an optimal order batching solution.Then,for the order picking of each batch,an optimized picking strategy of item grouping is proposed.Considering the vehicle service batch size,an integer programming model is formulated to item grouping problem.A heuristic algorithm based on location proximity degree of items is designed to obtain an optimal item grouping solution.The experimental results and analysis show that,the optimized order batching strategy proposed in this study can effectively reduce the number of bins that need to be delivered to picking stations,and the optimized item grouping picking strategy can effectively reduce time-consuming movements of vehicles such as turning by stopping,and acceleration or deceleration.These two optimized order picking strategies are both conductive to improving the efficiency of picking operations.