A hybrid artificial neural network as a software sensor for prediction of solid waste generation

Accurate prediction of quality and quantity of municipal solid waste is crucial for designing and programming municipal solid waste management system. But predicting the amount of generated waste is difficult task because various parameters affect it and its fluctuation is high. In this study with application of feed forward Artificial Neural Network (ANN), an appropriate model for predicting the weight of waste generation in Mashhad, was proposed. Since there are many variables in this research, entrance of effective variables can improve results, so Principal Component Analysis (PCA) technique that reduces the number of variables and entry effective variables in ANN used to model waste generation (PCA-ANN). The PCA is firstly employed to reduce input variables and after changing 13 original variables to 13 Principal Components (PCs), first 8 PCs were used as network inputs. Finally, these two models, ANN and PCA-ANN, were compared with each other and results showed network operation which was improved through preprocessing on input variables. In addition, PCA-ANN is the proposed model because it possesses faster training speed and more satisfactory predicting performance, also the simpler network architecture that the number of neurons in hidden layer decreased from 16 to 3 than the ANN model.

A hybrid artificial neural network as a software sensor for prediction of solid waste generation Keywords:

Waste Generation , Principal Component Analysis , Artificial Neural Network , Weekly Waste Prediction , Mashhad