Abstract: With the rapid development of Internet of Things technology and computer technology, coal mining enterprises are making great strides towards the direction of intelligence and informatization.When extracting key information from the vast amount of operating state data of electromechanical equipment, the traditional single-model prediction methods cannot meet the requirements of intelligent coal mine management due to their low accuracy.In view of this, this paper analyzes various prediction methods of data mining and attempts to use the weight distribution strategy to organically combine the time series model and the BP neural network model to build a new comprehensive data prediction model, which effectively improves the prediction accuracy.And taking the real-time monitoring data of the shearer in Zhujadian Coal Mine as a sample for model verification, the results show that the combined prediction model can accurately predict and evaluate the operating conditions of the electromechanical equipment, providing a strong support for the improvement of the intelligent management level of the coal mine and helping coal mining enterprises to better achieve intelligent transformation and efficient production.