Research on replenishment decision of vegetable commodities based on genetic algorithm

Abstract

Fresh produce superstores, are an important component of the modern economy, however, they face multiple challenges in vegetable merchandise management, including how to forecast accurate sales trends, develop reasonable pricing strategies, and effective replenishment programs. Therefore, this study aims to provide a comprehensive set of vegetable merchandising strategies for superstores to help them better solve their problems. Firstly, this paper constructs an integrated fitting model to fit a polynomial function to each single vegetable item; secondly, taking amaranth as an example, it constructs an image of the parameters of the amaranth fitting function; based on the ARIMA model of various categories of vegetable demand and wholesale price prediction, it solves the amaranth demand using the ARIMA model, and uses the hyper-parameter optimization to obtain the model with the smallest BIC value; it establishes the ARIMA (5,1,6) model and forecast the demand for the next seven periods, with the objective of maximizing the revenue of the superstore, establish a linear programming model (LP), constraints, and solve the model based on genetic algorithm.