A machine learning approach for predicting the performance of oxygen carriers in chemical looping oxidative coupling of methane
Abstract
A key focus of chemical looping oxidative coupling of methane is the screening of high-quality oxygen carriers. However, existing screening methods suffer from long material design cycles and high costs. Herein, we propose a novel method for predicting the reaction performance of oxygen carriers using machine learning models. Six different models are trained with the help of a dataset consisting of 300 groups. The Artificial Neural Network model is found to have the best accuracy and generalization ability. Using the ANN model, the reaction performance of eight new oxygen carriers is predicted. Among them, Na–LaMnO3 shows the highest C2 selectivity. We believe that machine learning provides a convenient and low-cost tool for screening oxygen carriers.