Accurate prediction of the ionospheric Total Electron Content (TEC) is of great significance for improving the accuracy of satellite navigation and positioning. To this end, a TEC short-term forecasting model that combines the Whale Optimisation Algorithm (WOA) with Long-Short Term Memory Networks (LSTM) is proposed in this study; The optimal fitness of WOA algorithm is obtained by LSTM model training, and the Optimal parameters of LSTM model are obtained by WOA algorithm optimization. Finally, combined with the TEC grid dot data provided by the Center for Orbit Determination in Europe (CODE), the proposed model is verified; the test results show that : in the geomagnetic calm state, the combined model is flat. The average correlation coefficient increased by 2.8%, 6.2% and 14.8% respectively compared with the LSTM model at low, medium and high latitudes; the average correlation coefficient of the combined model under geomagnetic active state increased by 6.6%, 9.2% and 7.9% respectively compared with the LSTM model in low, medium and high latitudes. And the prediction effect of the model is related to geomagnetic active state, season, solar activity level, etc. Under different geomagnetic active state, season and different solar activity level, the prediction effect of the combined model is better than that of a single LSTM model, which provides a reference for the practical application of the ionospheric TEC prediction model.