Modeling And Optimization Of "Parts-to-picker " Intelligent Picking Systems Based On Mobile Robots

One of the challenges that e-commerce retailers facing today is dealing with many time-pressed picking orders.Traditional "picker-to-parts" picking system has been unable to meet this demand.The goal of an intelligent automatic picking system is making the picker or the automatic picker to focus on handling the picking action rather than moving or searching the items.As the mainstay equipment of the "parts-to-picker"intelligent picking system,mobile robot usually assists the delivery of goods to the picker’s location,reducing the walking process of the manual searching process and focusing the picker’s or automatic picker’s main business on the extraction action with the highest value density.In this paper,the "parts-to-picker" intelligent picking system based on mobile robot has been studied.At present,how to evaluate the performance of "parts-to-picker" intelligent picking system accurately and efficiently has become the first important problem faced by such systems.The performance of intelligent picking system is affected by many factors such as structural parameters,operation strategy and system operation parameters.Based on this,how to build the performance evaluation model of different"parts-to-picker" intelligent picking systems is a new research hotspot.This paper aims to build a performance evaluation model for the intelligent picking systems and takes the fixed shelf type picking system and the mobile shelf type picking system as the main research objects,and uses mathematical modeling,queuing theory and other methods to conduct modeling and optimization research around the evaluation of system performance indicators and parameter configuration optimization.It is of great significance to realize performance evaluation of different intelligent picking systems and provides research basis and theoretical basis for structural analysis of this type systems.Firstly,in the research of storage equipment in the picking system,a structured quantitative analysis method of shelf is proposed.The structured comprehensive comparison has been studied through the utilization of shelf space,handling demand,initial input and operation cost and other aspects between the single-deep racks and gravity racks.As a starting point,the mapping relationship between product structure,cycle stock level,safety stock level and characteristics of different racking systems is studied.The space demand and the saving of transporting distance are introduced as the analysis indexes to improve the shelf evaluation system.Numerical experiments were conducted to analyze and verify the applicable conditions of single-deep rack and gravity rack respectively.The results show that when the cycle stock level is low,the space demand of gravity rack is smaller.When the cycle stock level is high,the advantage of single-deep rack is better.Then,as to the transportation equipment,we improves the combination of system evaluation model with the consideration of transportation equipment operation mode and shelf structure and evaluates the single-deep type rack（double operation mode）,gravity shelves system（single operation mode）,gravity shelves system（double operation mode）respectively by combined with the shelf structure and mobile robots’operation mode.In addition,this paper summarized system throughput,mobile robots’utilization,and reasonable allocation number of mobile robots as system evaluation indexes and discussed the influence of system structural parameters on system efficiency and mobile robots’ utilization in numerical experiments.The task completion rate is defined as the decision index of the reasonable allocation quantity of mobile robots,which provides a reasonable system evaluation system for system designers.Thirdly,as to the picking system structure layout research,this paper puts forward a variety of pick channel layout schemes and design optimization.The relationship among the position sequence of picking workstation,the shelf layout structure of the system and the expected running time of mobile robots is systematically studied.And this paper put forward a variety of picking station layout rules which includes picking channel external,picking channel internal,single picking channel,double picking channel,and two combinations of a variety of picking station layout scheme.At the same time,the objective function of the warehouse layout optimization model is designed by using the fragmented service time management method and the process of"splitting→ integration→ optimization".A warehouse layout optimization model based on SA algorithm is designed to find the optimal layout scheme of picking table for the system.Finally,in the aspect of the improvement and optimization of the picking system,a performance evaluation model of the composite shelf picking system was established.and a multi-objective optimization design model was built to solve the design parameters of the system.The composite system not only has the advantages of highdensity storage of fixed shelf picking system,but also has the efficient and flexible system configuration of mobile shelf picking system.In this paper,a discrete event system analysis method,queuing network,is used to model the composite shelf picking system.The system evaluation indexes such as system equipment cost,system throughput,system storage capacity,order fulfillment time and mobile robots’utilization rate are proposed.The multi-objective optimization design model of the above multiple indicators was built,and the optimal structural parameters,design parameters of mobile shelves,the number of mobile robots and the position sequence of picking tables of the system were solved by NAGS-Ⅱ method.