Vegetable Stocking and Pricing Model Based on Time Series Forecasting and Non-linear Programming

Abstract

This study aims to help fresh produce superstores develop vegetable replenishment and pricing strategies to maximize their benefits. By analyzing historical sales records and customer demand, the study used an ARIMA model to predict future demand and a planning model to optimize superstore revenues. The study also analyzed the sales data of 6 major categories and 246 individual types of vegetables and found that flower and leafy vegetables were the most popular in the market. By performing Spearman correlation analysis on the sales volume data, the study obtained the two pairs of vegetable categories with the highest degree of correlation. In addition, a positive correlation between sales volume and total price was fitted by linear regression and converted to a negative correlation between cost-plus pricing and sales volume. Finally, past costs, damage rates, price factors, and demand were taken into account to optimize each day's data individually to maximize the superstore's revenue, where the maximum profit on the first day was $1,520.6.