“The high alkali metal content in Zhundong coal can easily cause contamination of the boiler heating surface. Traditional methods have problems such as difficulty in simplifying the mechanism and lack of label data. Therefore, a machine learning based modeling method for monitoring heating surface ash pollution is proposed. Taking a 1000 MW power plant boiler as the research object, a high-precision dynamic simulation model including a control system is established, and new feature parameters are constructed on this basis. A pollution quantification characterization modeling method based on autoencoder and long short-term memory neural network is proposed, which effectively avoids the influence of load changes on thermal parameters and provides new ideas for monitoring heating surface ash pollution in boilers.”