A GM-TCN based Hybrid Prediction Method for Ash Fouling Prediction in Coal-fired Power Plant Boiler

In response to the problem that ash deposition in boilers has a great impact on production efficiency as well as production safety, a grey level theory prediction model based on time convolution networks (GM-TCN) is proposed to predict the degree of ash in boiler heated areas by using a cleaning factor to assess the ash deposition condition on the heated surfaces. The accuracy of the prediction model is improved by utilizing the features of grey level theory that can predict the developmental changes in system behavior, combined with the high-precision approximation capability of the time-convolutional network. The modelling prediction of the ash level in the heated area of boilers during the production of coal-fired power stations is investigated. Experiments demonstrate that the model can predict the cleanliness factor with high accuracy, and the predicted values are closest to the true values compared to the other two models, verifying the accuracy and reliability of the model.