Automatic Pricing and Replenishment Decision Based on WSO-BiLSTM and Nonlinear Programming

Supermarkets maximize their revenues by formulating appropriate replenishment and pricing strategies to satisfy both supply-side and demand-side conditions. This paper focuses on the restocking and pricing decision-making problem of vegetable commodities, which is solved based on objective planning analysis using simulated annealing algorithm and multi-objective particle swarm optimization algorithm. First of all, a wavelet analysis model is established, the wavelet variance diagram and quarterly trend diagram are drawn, and Spearman correlation analysis is used to solve the correlation between the sales volume of different categories or single products. Then, the fitting effects of multiple fitting models are compared, the number of independent variables is changed, and the fitting model with the best fit is selected. According to the cyclical characteristics of the total sales volume, the seasonal ARIMA model is selected to predict the total sales volume of each category in the next 7 days, and the predicted value is used as the upper limit of the total sales volume constraint for the subsequent target planning, taking into account the influence of other factors in real life. Finally, taking the total sales volume and cost-plus pricing as independent variables, while considering the influence of the attrition rate and the total number of saleable individual items limit. The objective function of revenue is established, and the order quantity of each individual item needs to satisfy the minimum display quantity as a constraint. The multivariate planning model is established, and the multi-objective particle swarm optimization algorithm is used for model solving.