Research on Predictive Pricing Strategy Based on Time Series Forecasting and AS-ADMM Algorithm

This study explores the management challenges of vegetable commodities in fresh supermarket, focusing on the impact of short shelf life and transportation losses, as well as the relationship between sales volume and pricing on supermarket profits. First, the data were preprocessed by using descriptive analysis and difference standardization, and then Spearman correlation coefficient was used for correlation analysis. A clustering model is constructed, which combines artificial bee colonies and K-Means clustering algorithm to reveal the interrelationship between different categories and items. Wavelet analysis and Fourier transform are also used to forecast the periodic sales volume, and seasonal time series models are used to reveal the obvious cyclical relationship between sales volume and season. It is found that there is an unreasonable relationship between the total sales volume of each category and the pricing. The relationship between pricing and sales volume in different time periods of each category is analyzed, and the Bi-LSTM bidirectional long and short time memory network is used for time series prediction to reveal the relationship between pricing and sales volume. Based on the correlation between the sales volume of each category, the incremental sales volume factor was constructed, and the pricing strategy model was constructed for vegetable category. AS-ADMM algorithm was used for planning and solving, so as to obtain the best pricing strategy and maximize the profit of supermarket. Finally, the model is extended to the case of single product, and the daily single product pricing and sales volume data are refined by Bi-LSTM model to further optimize the objective function and constraint conditions. AS-ADMM algorithm was used again to construct a pricing strategy model for vegetable products. Through the discussion, analysis and multi-index evaluation of the pricing decision model, the results of the model are robust, which can help the supermarket optimize the management of vegetable commodities, improve sales efficiency, reduce costs, and ensure the quality of goods, so as to enhance competitiveness. This study provides strong decision support for fresh grocery stores to respond to changing market demands and challenges.