Abstract: In this paper, the standard sparrow search algorithm SSA is added into Tent chaotic mapping and dynamic step factor to obtain the optimized sparrow search algorithm ISSA,so as to improve the diversity of the population and regulate the global search ability and local exploitation ability of the population. Then,the internal state function of the savings pool of the standard ESN is replaced by the hyperbolic tangent function to obtain the NESN. Finally,the sparsity SD and spectral radius SR of savings pool in nonlinear echo state network were optimized by using the optimized sparrow search algorithm ISSA,and the ISSA-NESN prediction model is built. The example analysis shows that the mean absolute percentage error（MAPE）of ISSANESN is 15.84%,and the root mean square error（RMSE）is 0.12,which is better than other comparison models.