We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Prediction of Pyrolysis Oil Yield from Plastic Waste using Artificial Neural Network (ANN)
Summary
Researchers trained an artificial neural network on pyrolysis experiments covering 0.5–5 kg of plastic waste, using temperature, pressure, particle size, flow rate, residence time, and heating rate as inputs, achieving an R² above 0.998 for predicting oil yield — demonstrating that neural models can reliably optimize plastic-to-fuel conversion.
Plastic pollution has various negative effects, including significant ecological disruptions and the degradation of landscapes. An alternative to managing plastic waste (PW) is pyrolysis. In this study, the application of artificial neural network (ANN) for the pyrolysis of PW was investigated. The amount of pyrolysis oil yield (POY) from 0.5 kg to 5 kg of PW were determined. The Levenberg-Marquardt back propagation training algorithm (trainlm) was used to train the ANN using MATLAB R2022b ANN Toolbox. The POY is the output, and the input sets were temperature (T), pressure (P), particle sizes (PS), flow rate (FR), residence time (RT), heating rate (HR), and PW. The usefulness of the model for POY was demonstrated by the study effort, which verified the performance of the developed ANN models. The best validation performance's mean squared error (MSE) was 0.0060125. The ANN regression plots with an R2 of more than 0.998 also demonstrates excellent results. The suitability of the ANN model for precise estimation of POY from PW is shown by the high values of the coefficient of determination (R²). The study demonstrates that ANN is a versatile and effective technique for deciphering intricate pyrolysis data and maximizing the output of high value oil by optimizing operating conditions.