In order to solve the thermal error compensation problem of electric spindle in a machining center,a thermal error compensation model with high prediction accuracy and strong robustness was established.The experimental platform was set up,and the temperature field and thermal error data of electric spindle were collected by using American Lion rotary error analyzer.Sparrow search algorithm (SSA) principle and specific optimization process were introduced.Using SSA to optimize the weights and thresholds of the BP neural network,the SSA-BP neural network prediction model was established.Compared with the previously established BP neural network prediction model,the prediction effect after optimization is better,which provides a new idea for modeling the thermal error of the electric spindle.