Bed temperature is a critically important process parameter influencing the efficient

safe and environmental operation of circulating fluidized bed incinerators(CFBIs).By taking use of functions like the nonlinear dynamics

self-learning ability and generalization ability of the adaptive neuro-fuzzy inference system(ANFIS)

aprecise model was developed for forecasting the bed temperature of a CFBI

which can incinerate 650 tons municipal solid waste(MSW)per day.Through analyzing the operation mechanism of the CBFIs

the input variables of the model was chosen.Then

the subtractive clustering algorithm was employed to determine the initial membership functions(MFs)by partitioning the archive data and extracting a set of fuzzy rules.Moreover

prediction accuracy of the established model was compared with that of the multi linear regression(MLR)model

the typical back propagation neuro network(BPNN)model and the support vector machine(SVM)model.The result shows that

the proposed model has a higher prediction accuracy

and this conclusion can provide guidance for the control of CFBIs operation.