Ethyl levulinate(EL)has been considered as a new liquid biofuel candidate.One-pot production of EL from wasted biomass was investigated in this study.Furfural residues(FRs),the abundant wasted biomass from furfural production,were taken as substrates for EL production via one-pot process.To optimize the process,two optimal design methods,response surface methodology(RSM)and artificial neural network(ANN)were employed to emulate the process.Box-Behnken experimental design with four variables including temperature,catalyst loading,mass ratio of liquid to FRs and reaction time was used to construct the models.The results showed that the simulation of ANN was apparently better compared to RSM,and the maximum EL yield of 18.68%identified by genetic algorithm(GA)based on ANN model was higher than that of 15.43%indentified by RSM model.Thus,a higher theoretical EL yield of 51.2%can be obtaind under the optimum conditions.This result suggested that wasted FRs can be used as good raw materials for one-pot production of EL.