At present

the commonly used traffic prediction algorithms are mainly mobile average method and exponential smoothing method. Although the traffic trend prediction can achieve a certain accuracy

the traditional traffic prediction algorithm has some defects for short-term traffic cyclical changes

such as tidal effects. Therefore

it is necessary to use BP neural network algorithm for more accurate and intelligent traffic prediction

and provide the algorithm basis for tidal carrier scheduling. This paper will study the application of BP neural network in traffic prediction and evaluate the prediction performance of the model.