The target image deviation caused by aero-optical effect has a great influence on the navigation,positioning and homing of aircraft,so it has an important practical v alue for the real-time compensation of aero-optical imaging deviation.In this paper,an improved sparro w search algorithm optimized BP neural network (ISS-BP) model is proposed to predict the aero-optical imaging deviation.In order to improve the ability of the prediction algorithm to search and jump out of the local optimal,the bird swarm algorithm (BSA) flight behavior in the s tandard sparrow search algorithm (SSA) is used to make the entrants approach the finders with a certain probability to shorten the running time of the algorithm and ensure the global convergence and the diversity of the population.Finally,the algorithm model in this paper,BP neural network model and sparrow search algori thm optimization BP neural network (SSA-BP) model are compared and three evaluation indexes are used to evaluate the three algorithm models.The results show th at the ISSA-BP model in this paper can predict the aero-optical imaging deviation accurately a nd timely,and the mean square error (MSE),mean absolute error (MAE) and determination coefficient (R²) of ISSA-BP model are 2.511E-11,1.969E-06and 0.99989,respecti vely.