The classical Echo state network (ESN)cannot fully exploit its advantages in some workwhich characterizestrong nonlinearity and high-order statistics.In order to overcome the shortcomings, this paper using the nonlinear readout nodesinstead of the traditional linear readout nodes in classical ESN, Radical Basis Function (RBF) neural network was introduced to read out the reservoir state.Benchmark experiments on Lorenz chaotic time series and NARMA modelidentification shown that the proposed new type of ESN canimprove the nonlinear characteristics andits performance exceeded the classical ESNin dealing with some higher degree of nonlinear system and model.