Study on the Measurement and Prediction Model of Sludge Drying Characteristics
Sludge dewatering and drying are crucial steps in the resource utilization of sludge.There is currently a lack of a comprehensive drying model of sludge.The effects of drying temperature,relative humidity,sludge thickness,and drying time on drying sludge were analyzed in the experiment.The fitting effects of five commonly used thin-layer drying models on the sludge drying process were compared.A BP neural network model was established for predicting sludge drying.This was compared with the Midilli model,which has traditionally shown better fitting results for prediction accuracy.The results indicate that the drying of sludge is significantly affected by temperature and relative humidity when it is dried at low temperatures.The higher the relative humidity and the lower the temperature,the slower the rate at which the sludge dries.The Midilli model has a high coefficient of determination,and its chi-square and RMSE values are relatively low.It is the best-fitting model among the five commonly used thin-layer drying mod-els.The error compared to the experimental results is within 15％.The BP neural network sludge drying prediction model can predict the sludge drying process very well.The prediction results have less than 5％errors compared to the experimental results.The model has a higher predictive accuracy than that of the Midilli model.The BP neural network model for sludge drying prediction provides a new method to simulate the sludge drying process.
SludgeThin layer drying modelBP neural network model for sludge drying prediction
万方数据
维普
