Research On Optimization Strategy Of FMCG Based On Time Series Analysis

In recent years,e-commerce business gains prosperous development,and the influence of logistics industry becomes more and more profound.As the most eye-catching commodity genre,the mode of less quantity and more times of fast moving consumer goods provides extremely big test for the sorting and distribution of goods.Most researches use the heuristic algorithm to improve and optimize efficiency.Optimized researches based on relevance through order are relatively few,and cannot fast accommodate to the ever-changing consumption trend.In order to make the goods allocation optimized model able to adjust to the ever-changing orders,this paper designs the new goods allocation distribution optimized model,which efficiently improves the efficiency of sorting operation.The major research contents are as follows:(1)Two anomaly detection algorithms are designed and improved,pulseconsumption to influence optimized model is avoided.Through analysis on the consumption characteristics of FMCG and research on frequently-sued detection algorithm,the goods are divided into two categories.Aiming at the numerous common goods,the K-iForest algorithm based upon iForest algorithm and K-means algorithm is proposed,which guarantees the efficiency of algorithm and controls the sensitivity of algorithm.Regarding the luxury products,SVR model introducing sliding windows is introduced,making the algorithm has good inclusivity on the small fluctuations.(2)Through GSP algorithm,the goods allocation optimized model between theturnover and the correlation is built.By combining with the products characteristics,build GPS algorithm model to calculate the correlation factors on commodity,and get the correlation index of the goods by combining seasonal factor,warehouse layout,shelf characteristics and turnover condition,build storage location distribution optimization model,place the goods in reasonable storage location for convenient picking and sorting.(3)Based upon the above research result,the example experiment is combinedto further propose optimization suggestion on model.Based on historical order's test and assessment on the improved algorithm,verify the optimization effect of model and propose suggestion on the optimization direction of the model.The example experiment proved the goods allocation distribution optimized model based on temporal sequence has obvious effect on improving sorting efficiency.Then according to the ever-changing order,it can provide reference for improving the efficiency of warehouse workers and lowering down the operation cost.