Short-term Prediction of Passenger Flow in Jiuzhaigou Scenic Area Based on SCBANet Model
Aiming at the problems of poor feature extraction ability,large prediction error,and difficulty in capturing unconventional changes in the short-term prediction of scenic spots,a neural network model SCBANet is proposed,which combines spatiotemporal normalization,convolutional neural network,two-way long short-term memory network and attention mechanism.Firstly,the two modules of spatiotemporal normalization are used to refine the high-frequency components and local parts of the passenger flow data;secondly,the convolutional neural network is used to extract the features of the processed data;then the two-way long short-term memory network uses the extracted features to predict the tourist flow of scenic spots;and finally the attention mechanism is used to capture the influence of different time frequencies on the passenger flow of scenic spots in the past,so as to improve the accuracy of prediction and capture unconventional changes.The experimental results show that compared with other algorithms,the prediction error of the model proposed in this paper can be reduced by 97.63%,and the daily relative error of the prediction of the passenger flow of scenic spots in the coming week is less than 4%,so it is more suitable for the prediction of short-term passenger flow of scenic spots.
NETL
NSTL
万方数据
维普
