Co-combustion characteristics of textile dyeing sludge (TDS) and pomelo peel (PP) under O-2/N-2 and O-2/CO2 atmospheres were investigated using a thermogravimetric analysis (TGA) and artificial neural networks. 30% O-2/70% CO2 and air atmospheres led to a similar co-combustion performance. Increases in O-2 concentration and PP significantly improved the oxy-fuel co-combustion performance of TDS. Principal component analysis was applied to reduce the dimensionality of differential TGA curves and to identify the principal reactions. The interaction between TDS and PP occurred mainly at 490-600 degrees C, thus improving the process of residue co combustion. Radial basis function was found to have more reliable and robust predictions of TGA under different O-2/CO2 atmospheres than did Bayesian regularized network. Regardless of Flynn-Wall-Ozawa (FWO) and Kissinger-Akahira-Sunose (KAS) methods used, the lowest mean value of apparent activation energy (155.4 kJ.mol(-1) by FWO and 153.2 kJ.mol(-1) by KAS) was obtained under the 30% O-2/70% CO2 atmosphere.