Research On Energy-efficient Optimization Strategies In Drone-assisted Mobile Edge Computing Network

With the increasing competition,more and more retailers optimize their business processes with the help of intelligent technologies to enhance the efficiency of resource allocation and thus achieve the purpose of reducing operation costs and improving operation management.Inventory management can be carried out by determining the appropriate order quantity to replenish the product inventory for the stores and transporting the products from the central warehouse to the stores through the vehicles,therefore,a reasonable and accurate ordering plan plays an important role in the enterprise’s inventory management.However,the increase in the number of products,the shortage of historical data,and the irregularity of demand have led to complex and variable demand patterns,representative of which is the intermittent demand pattern,while the commonly used time series forecasting methods are mainly applicable to stable and continuous historical data.Therefore,products with intermittent demand patterns bring challenges to demand forecasting and inventory management.This study aims to optimize the inventory routing problem in apparel retailing enterprises.It investigates the inventory routing optimization problem with intermittent demand and proposes different modes of transshipment strategies to achieve the integrated optimization from demand forecast,replenishment plan,and distribution plan while improving the inventory management performance and reducing the total cost of the system.The research content of this paper is as follows:(1)To address the low accuracy of demand forecasting for products with intermittent demand patterns,this study takes demand forecasting as an object,and for the characteristics of uncertain demand occurrence and unstable demand quantities,proposes a demand aggregation forecasting model to predict the demand for the next period,constructs a demand occurrence classification model that predicts whether or not the demand occurs,establishes a decomposition model that decomposes the aggregated results,and a classification absolute error metric is designed for model evaluation.(2)To solve the inventory routing problem for products with intermittent demand patterns in transit between customers,an optimization model for intermittent demand in single-cycle same-level transshipment is developed based on the results of intermittent demand aggregation-disaggregation prediction,an external transshipment service is introduced to balance the inventory of the products at the customers,and the removal and insertion operators of the adaptive large-neighborhood search algorithm are improved to enhance the solution quality and to lay a foundation for the subsequent solution of the twolevel transshipment problem.(3)To solve the problem that products with intermittent demand are not only transshipped between customers but also need to be partially transshipped back to the central warehouse for temporary storage,this paper constructs a single-cycle two-level transshipment intermittent demand inventory routing optimization model by using the results of the intermittent demand combination prediction model and puts forward a clustered mode transshipment pickup strategy.Subsequently,the mutation and reorganization operations of the differential evolution algorithm are improved according to the characteristics of the problem to accelerate the convergence and solution quality of the algorithm,and the problem further enriches the inventory routing optimization domains under different transshipment scenarios and provides a foundation for the subsequent solution of the inventory transshipment problem under multi-level transshipment scenarios.(4)To solve the problem that some products need to be transported from the customer to the manufacturing center,which also includes the complex scenarios of products being transshipped between customers and products being transported from the customer to the warehouse,this paper investigates the multi-cycle multi-level transshipment intermittent demand inventory routing optimization problem and establishes a mixed-integer planning model aiming at the minimization of the total cost of the system.Considering the transshipment differences of products,a pickup strategy containing cluster transshipment mode and pick-and-deliver mode is proposed.In addition,a hybrid genetic algorithm is designed for the model with improved selection,crossover,and mutation operators to enhance the algorithm solving capability.The solution of this problem provides strong theoretical support for apparel retailers to solve the inventory routing optimization problem for intermittent products.This paper also develops a demand forecasting and inventory routing optimization system,which contains five functional modules,including a data collection module,data processing module,sales forecasting module,inventory routing module,and system management module.In the actual business context,the intermittent demand combination forecasting model and the multi-cycle multi-level transshipment intermittent demand inventory routing optimization model are embedded,the real dataset of the enterprise is collected through the data collection module and input into the sales forecasting module after data processing,and then the inventory routing optimization module automatically generates the replenishment plan and the distribution plan for the coming period.In addition,this paper also uses relevant evaluation metrics to evaluate the results of the system and then verifies the correctness and effectiveness of the program proposed in this paper.To improve the inventory management efficiency of apparel retail enterprises and reduce the total cost of enterprises,this paper designs an intermittent demand aggregation-disaggregation forecasting model,which aims at solving the inventory routing optimization problem in different scenarios,and develops an inventory routing optimization prototype system,which verifies the validity of the methodology proposed in this paper,and achieves the integrated optimization of three major businesses,i.e.,from the demand forecast,the replenishment plan,and the distribution plan,and thus provides the best results for the enterprise’s demand management and distribution plan.The integrated optimization of the three major businesses from demand forecasting,replenishment planning,and distribution planning is realized for the enterprise,which in turn provides scientific decision-making methodology support for the enterprise’s cost reduction and efficiency enhancement.