Abstract The study bases on the pyrolysis characteristic, kinetic, and thermodynamic parameters and evolved gas analysis to quantity Chinese medicine residues (CMR) and uses artificial neural network (ANN) to reconstruct and jointly optimize pyrolysis. The main weightlessness interval of CMR is between 150 and 600 °C including organic matter decomposition. Four model-free methods and one model-fitting method were provided to find function mechanisms and kinetic parameters show it existing kinetic compensation through pyrolysis. TG-FTIR finds the gases and functional groups included CO2, C O, H2O, CH4, CO, C C, and C–O. And the main pyrolytic products were detected included esters, phenols and acids et al. 9-octadecenoic acid (z)-, methyl ester as one of the high quality products was in the highest proportion about 53.75%. The temperature-, heating rate-, and their non-linear dynamics of the multiple pyrolysis response surfaces were reconstructed and jointly optimized using an artificial neural network algorithm. Finally, the study is helpful for Chinese medicine residues high-value utilization.