Pricing and Replenishment Decisions for Vegetable Commodities Based on ARIMA Prediction and Particle Swarm Optimization Algorithms

Supermarkets need to make replenishment decisions for the vegetable category of the day without knowing exactly what the specific individual items and the price of the stock are. Accurate market demand analysis is crucial for replenishment and pricing decisions. In this paper, for the problems in automatic pricing and replenishment of vegetables, based on the idea of cost-plus pricing and linear programming, a model is established by determining the indicators of sales volume, selling price, and relevance of vegetables, with the goal of finding the relevance of each part and obtaining the maximum revenue, and the model is solved using the particle swarm algorithm. Firstly, considering the different selling prices and sales volumes of vegetable categories in different time periods, two-by-two correlation analysis is carried out by calculating the Pearson correlation coefficient, and the correlation heat map is used to reflect the correlation problem between each vegetable category. Then, the ARIMA model is established to predict the cost of each vegetable category, and the quadratic programming method is used to predict the total daily replenishment and pricing strategy, which gives the replenishment volume and pricing strategy of each category with the greatest benefit in the coming week. Finally, by establishing a planning model for single product selection and using particle swarm algorithm to solve the problem, different profit margins are chosen to calculate different single product pricing, and the replenishment quantity and pricing strategy for single product with the maximum revenue are formulated. The proposed model is fully integrated with reality and has the advantages of simplicity and easy to understand.