A Study of Pricing and Replenishment Decisions Based on Wavelet Neural Network Prediction

Management of shelf life becomes a critical issue in the sale of vegetables in fresh produce supermarkets. Using the provided data, this study provides an in-depth analysis of the sales patterns among vegetable categories and predicts the restocking volume and pricing strategy of vegetable categories in the coming week and on specific dates. In the sales distribution analysis, it was found that none of the six categories of vegetables conformed to normal distribution, and the pattern of sales volume affected by season and epidemic was explored by plotting graphs and calculating Spearman's correlation coefficient. Further, the relationship between total sales volume and pricing was investigated through a multiple linear regression model and combined with a wavelet neural network to predict sales volume, a multi-objective nonlinear planning model was established, and a linear approximation bundle optimization method was applied to solve for maximum profit. For the replenishment strategy on a specific date, a simulated annealing algorithm is used to solve the problem by combining the data features and model constraints, and the replenishment volume and pricing strategy of a specific product are effectively predicted.