In this study, an artificial neural network (ANN) was trained to model dynamic behavior of pressure fluctuations measured in a circulating fluidized bed with a riser having an inner diameter of 0.10 m and a height of 10 m, The ability of the neural network model to approximate the dynamic behavior is examined by comparing time-averaged characteristics, power spectra, and chaotic features of time series measured and generated by the ANN, It is found that dynamic behavior of the original time series is captured well by the ANN, and that the ability of the ANN for generation improves with the number of iterations.