Short-term Sales Prediction Method of City Gas

Short-term sales predicate of city gas is a key link in the management of energy supply and demand balance and the safe operation of natural gas pipeline networks. Accurate prediction models are of great significance for optimizing resource allocation and ensuring people's livelihood. This paper focuses on the problem of short-term sales prediction of city gas. Five prediction models, namely AutoRNN, AutoTSMixerx, AutoNHITS, AutoMLP and AutoTCN, are selected. Combined with external influencing factors such as temperature and holidays, prediction models are constructed and empirical analysis is carried out. The study quantitatively evaluated the model performance through the root Mean square error (RMSE) and Mean Absolute Percentage Error (MAPE) indicators. The results showed that AutoTSMixerx and AutoMLP performed excellently. The research results provide a standardized solution for the short-term prediction of city gas sales and offer assistance for the efficient operation of the natural gas industry chain and the intelligent upgrade of the energy system.