In the context of the digital era, the use of intelligent algorithms to improve production efficiency and decision support is receiving attention from many production enterprises. In this paper, an AttLR-LSTM-based time series model of production data is proposed based on time series data features, combined with LSTM network and attention mechanism to realize accurate prediction of key indexes of production devices, and an intelligent recommender system based on collaborative filtering algorithm is designed to improve the decision support capability and efficiency in the production process. In the comparison experiments with different machine learning models, the prediction effect of this paper’s method is improved by 88.91% and 60.92% compared with the LSTM with long-term time series data as input and the LSTM with short-term time series data as input, which fully proves the validity and stability of this paper’s method in the prediction of the operating state of production devices. Meanwhile, the real-time business indicator recommendation system designed in this paper not only has high satisfaction, but also receives unanimous praise for its accuracy and confidence.
