Impact of Visitor Flow on Baiyun Mountain Scenic Area and Transformer Model Prediction
Abstract：Mobile operator-provided mobile signaling data in the Baiyun Mountain Scenic Area region was used to reverse-engineer the visitor flow data. After cleaning and processing historical meteorological data and visitor flow data, SPSS software was used to analyze the relationship between 14 types of influencing factors, including time factors such as year, month, day, hour, and meteorological factors such as temperature, relative humidity, 1-hour cumulative rainfall, wind direction, and wind speed, and the visitor flow. The results showed that visitor flow is related to six time factors and eight meteorological factors, with more significant correlations with year, month, hour, temperature, relative humidity, and pressure. Based on the analysis results, a Transformer model for predicting visitor flow was constructed. When the forecast lead time was ≤4 hours, it achieved good prediction results, with an average absolute percentage error ranging from 3.12% to 6.53%.
Keywords：Meteorological conditionsVisitor flow predictionTransformer modelBaiyun Mountain Scenic Area
Publication Date：2025-02-19
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)