Analysis and Prediction Transient Population in Expressway Service Area based Long Short-Term Memory

The expressway service areas are widely built in the expressway system, as an important place for regulating road traffic flow, alleviating traffic pressure, providing short breaks and adjustments. However, the current intelligent transportation analysis and prediction are mainly concentrated on urban roads, and relatively few expressways. This paper comprehensively analyzes the operating data of the expressway service area, including ingredients such as the number of cars, passenger consultation data, weather conditions and water consumption, to predict the transient population in the expressway service area based on Long Short-Term Memory(LSTM). The experimental results show that the Mean Absolute Percentage Error(MAPE) of the transient population prediction is 10.25%, which further verifies the effectiveness of the model.