Investigating the co-combustion characteristics of oily sludge and ginkgo leaves through thermogravimetric analysis coupled with an artificial neural network

The co-combustion characteristics of oily sludge and ginkgo leaves (GL) in an oxy-fuel atmosphere are investigated via thermogravimetric analysis coupled with an artificial neural network. The combustion characteristics of blends improve as the GL mass ratio increases. The interaction indices used to evaluate the interaction between the two solid combustibles present a complex nonlinear relationship in different stages. The Flynn-Wall-Ozawa and Kissinger-Akahira-Sunose methods are used to calculate the activation energy of the blends, which increases with an increase in the oxygen concentration, in different atmospheres. Compared with the radial basis function, the backpropagation neural network performs better in predicting the combustion curve of the blends.