Research On Storage Robot Scheduling Problem Based On Dwell Point Policy In Horizontal Operation Mode

With the gradual development of the Internet economy,and the increasingly multi-category,small-batch,quantitative and dynamic characteristics of the e-commerce industry,more and more attention are drawn to the efficiency and automatic level of logistics system.In this study,we used the storage system in horizontal operation mode as the storage system,in which storage robots as a key automatic equipment for the picking process,associates with the main control system and sorting system to realize the circulation of "part-to-picker" process.Therefore,the control strategy of storage robots is very important for the high efficiency of the picking process.This thesis analyses the periodic orders and storage robots status of a e-commerce sorting warehouse distribution center,and predicts the probability of SKU picking task mission.Combining these analysis,a corresponding heuristic strategy was adopted to carry out an innovative study on the storage robots scheduling problem based on dwell point policy in the storage system in horizontal operation mode,as a basis for research on the storage robots scheduling problem in similar systems in the future.Firstly,after an in-depth analysis of the facility equipment configuration,warehouse layout,and "part-to-picker" operation process of the storage system in horizontal operation mode,the distance of the walking path of all picking tasks of the batch order within the completed period is selected as the standard for measuring the degree of optimization of the dwell point and storage robots scheduling problem.Secondly,we establish mathematical models corresponding to the description of the storage robots scheduling problem based on dwell point policy both in single and multiple storage robots modes,obtaining that a reasonable selection of dwell point location can effectively optimize the system reaction time and shorten the picking route distance.Heuristic policies for the dwell point selecting and particle swarm optimization algorithm for robot scheduling problem are proposed for each of the two modes,and the steps,flow charts of the policies and algorithm are described.Finally,the real order data of a e-commerce sorting warehouse distribution center is used to simulate,and the results are analyzed and compared.We obtain the ranking of walking route distance of all the picking tasks of batch orders within a single storage robot's completion cycle:Heuristic policy for dwell point selecting with reference to the maximum picking probability position;Heuristic policy for dwell point selecting with reference to the maximum picking probability area;The dwell point selection policy staying at the midpoint below the warehouse;The dwell point selection policy staying at the inbound and outbound ports;The dwell point selection policy staying at a random position;We also obtain the ranking of walking route distance of all the picking tasks of batch orders within a multiple storage robots'completion cycle:Heuristic policy for dwell point selecting with reference to the maximum picking probability area is better than the dwell point selection policy staying at a random position.We suggest that during the same busy hour each day,select the historical order situations in the previous multiple cycles to predict the probability of future SKU picking tasks and determine the dwell point location.Based on this,the rational scheduling optimization of the task execution sequence or walking path of the storage robots in the batch cycle in the same period is performed.This guides the actual picking of batches in the next cycle of the same period to better suit the order structure of such enterprises and the actual operation characteristics of the warehouse,and to achieve the effect of saving the operating costs of the enterprise and improving the efficiency of the storage system.