Prediction of mass loss for sewage sludge-peanut shell blends in thermogravimetric experiments using artificial neural networks

Based on mass loss data of sewage sludge (SS) and peanut shell (PS) during combustion and pyrolysis obtained by thermogravimetric analysis (TGA), artificial neural networks (ANN) models were established to predict the experimental results in this paper. The effects of mixing ratio and temperature on the combustion and pyrolysis of SS-PS were studied. Mixing ratio and experimental temperature were taken as input parameters of ANN model, and sample mass was taken as output parameter. The optimal ANN models (ANN12 for combustion and ANN11 for pyrolysis) were determined, which can well predict the thermogravimetric curves of combustion and pyrolysis with the correlation coefficient of 0.99999. The predicted data are in good agreement with the experimental data of the ANN model, which proves the reliability and accuracy of the application of ANN for the thermogravimetric experiments of SS and PS. The established optimal ANN model can be used to predict the weight loss of SS-PS heat treatment without experiment.