A Study On The Efficiency Optimization Of AVS/RS And Picking In B2C E-Commerce Distribution Center

B2C E-commerce logistics has the characteristics of large variety of orders with small amount of frequent orders,great order variance,short response time,high number of break-bulk picking and selecting operation,large amount of returned goods and so on.The increased storage area and growing orders have resulted in longer picking routes,lower picking speed,and less accuracy.Therefore,selecting an appropriate and reasonable storage and order picking system as well as practicing storage allocation and order picking strategy is of great significance to B2C E-commerce logistics centers in improving system efficiency and accelerating order response.However,the current research on E-commerce logistics operators' hardware selection and their related management strategies is relatively rare.Based on the demand of B2C E-commerce logistics,the author selects four areas as the research object after viewing current domestic and foreign,researches and the practical dilemma faced by e.rprises in this study.First is scientifically and reasonably choosing an efficient storage and picking system under the constraints of order density and working space.Second is enhancing system efficiency through slot allocation strategy under the impact of system hardware,research background,sales model and order characteristics.The third is optimizing order batching in order to meet the requirements of B2C E-commerce firms' quick response to customers'orders while improving the imbalance of system operations caused by order fluctuations.The fourth is sequencing orders within the same batch aiming at reducing waiting time for both machines and customer orders.Regarding the above mentioned four topics,the main research contents and findings of this paper are as follows:(1)Through the analysis of the operation process of automation system and the analysis of equipment operation,the concentrated and dispersed AS/RS is proposed;the trajectory of the equipment under the combination of four kinds of operation instructions is analyzed;and single instruction and multiple instruction work time modeling under two systems pulled by random order are constructed.(2)The indicators and methods of system evaluation from the aspects of operation efficiency and reaction speed are defined;obtaining the index data of the two systems under different conditions through adjusting rack structure and order density based on proposed simulation model;the two systems are compared horizontally from different angles,and the data were analyzed and curve-fitted;multi-objective evaluations are performed on the two systems.(3)The influence of the distribution of item position on the system operation time is analyzed when the single order contains multiple items;the calculation method of similarity coefficient which is suitable for the B2C order is selected followed by calculating similarity coefficient matrix between items,and proposing two-stage K-means clustering methodology based on the similarity coefficient of items and multi-level heuristic clustering algorithm based on time-saving method.On the basis of clustering,the selected frequency of goods and the discount rate of commodity prices is introduced,and the concept of historical value and future value of item logistics is proposed to conduct the allocation optimization,(4)The adaptation of different batch strategies to B2C orders is discussed,and the calculation model of time window batching is established;the variable time window batch modeling constrained by order lines is built in order for improving the stability of batch order operation time;the refined classic algorithm in route optimization--seed algorithm and saving algorithm,is applied to sort the orders,which obviously reduce the waiting time of machines;,and improve order reaction speed.Finally,the paper presents the following conclusions:(1)Under multi-objective evaluation of the two systems,when the level of shelves keep constant while the order density and the number of shelves change between 10 and 100,and when the order interval is 0,the number of shelves is 100,the target value is the highest;when the number of columns is less than 100,the level of shelves is less than 10,while the order density is fixed,the target function value is the largest when the number of system columns is 100 and the number of floors is 9 layers.(2)The similarity coefficient calculation formula proposed by Russel and Rao in the field of industrial engineering is applicable to the calculation of the similarity coefficient of items under B2C E-commerce logistics circle.With respect to clustering method,the multi-level heuristic clustering algorithm based on time-saving method is better than the two-stage K-means clustering algorithm based on similarity coefficient of items.With regard to slot allocation method,the IFLV-based allocation optimization strategy is better than IHLV-based allocation optimization strategy,and the IHLV-based allocation optimization strategy is better than random allocation strategy.(3)Variable time window batch modeling constrained by order lines can effectively reduce the instability of system response time caused by imbalanced orders.(4)The modified seed algorithm and saving algorithm can be used for order sequencing to effectively reduce delivery time of batch orders and improve the timeliness of the system's response to orders,but the improvement effect on the reaction efficiency is inferior to that on delivery time.And the improvement effect of saving algorithm is better than seed algorithm.The above research has a sound optimization effect,fills in the gaps concerning AVS/RS shelf design,item clustering,slot location optimization and order sequencing in the B2C E-commerce field,and is of great value to practice.