Related Simulation Prediction Research Of Calorific Value For Combustible Contents Of Household Waste And RDF

The combustible composition of household waste has a higher calorific value, and recycling the energy of that fully has great significance to sustainable utilization of the energy.This article chose Artificial Neural Network to do data mining, which based on the test data determined of household waste from residential quarter. The research chose the contents of PE/PP, paper, rubber, moisture and dry hydrogen as the input parameters and choose the low calorific value of the combustible contents as the output parameter, to establish a prediction model of low calorific based on back propagation neural network(BP), radical basis function neural network (RBF) and adaptive neural fuzzy inference system (ANFIS).The results showed that the prediction accuracy of BP neural network model was 93.36%, and RBF model was 96.87%, and ANFIS model was 91.06%. The comparison of three kinds of model showed that:each model can be used to predict calorific value. ANFIS model has a very high model imitative effect, but it has a higher validation error, which its average prediction accuracy is low. BP model has a higher prediction accuracy and can predict the calorific value of combustible contents better, but its results are lower than RBF model. The RBF model added a linear control based on BP model, so that it could makes a great improvement of the prediction accuracy. Therefore the the RBF model is more suitable for the prediction of the low calorific value of combustible composition for MSW with a more satisfactory result.When the contents of PE/PP, paper, rubber, moisture and dry hydrogen are 31.66%、 59.94%、0.03%、31.37%、8.74%, the calorific value was predicted about 16352.61kJ/kg by RBF model, which met the calorific value of RDF. Therefore, it has the theoretical basis by using combustible composition of household waste to produce RDF.