In order to monitor the abnormal speed of the ship out of the lock, scientifically dispatch the ship out of the lock, and improve the efficiency of the ship out of the lock, the distribution law of the ship speed and the application of the prediction are studied. Based on the AIS statistical data, the relationship between the ship"s distance from the lock, the ship"s size and the ship"s speed is analyzed by histogram method, Based on the BP neural network based prediction model of ship speed is established based on 80% AIS data, and the prediction model is verified with 20% AIS data; Finally, a case was given to illustrate the application of the speed prediction model. It was showed that the average error of the model is 6.7%, and there is no significant difference between the predicted speed and the actual speed by one-way ANOVA. It can be seen that the ship speed prediction model based on BP neural network can be applied to the ship speed prediction.