Research Article
Design and Implementation of Traffic Flow Prediction Model Based on Short and Long Time Memory Network
@INPROCEEDINGS{10.1007/978-3-031-65126-7_10, author={Sheng Liu and Xinyue Li and Ting Cao and Shuxiao Chang}, title={Design and Implementation of Traffic Flow Prediction Model Based on Short and Long Time Memory Network}, proceedings={Quality, Reliability, Security and Robustness in Heterogeneous Systems. 19th EAI International Conference, QShine 2023, Shenzhen, China, October 8 -- 9, 2023, Proceedings, Part I}, proceedings_a={QSHINE}, year={2024}, month={8}, keywords={Neural network Integrated learning LSTM Short time traffic flow forecast}, doi={10.1007/978-3-031-65126-7_10} }- Sheng Liu
Xinyue Li
Ting Cao
Shuxiao Chang
Year: 2024
Design and Implementation of Traffic Flow Prediction Model Based on Short and Long Time Memory Network
QSHINE
Springer
DOI: 10.1007/978-3-031-65126-7_10
Abstract
Due to the randomness, fuzziness, time variability and uncertainty of traffic flow, it is difficult for traditional forecasting models based on time series or artificial neural networks to accurately reflect the actual traffic situation, etc. This paper takes the demand for short-term traffic flow forecasting of urban rail transit as the research object, and analyzes the implementation methods suitable for short-term traffic flow forecasting. LSTM neural network was used to construct the model for simulation experiment analysis. The results of data analysis show that the LSTM neural network model obtains the minimum average absolute percentage error MAPE value of 10.6% and the highest average accuracy of 89.4%, which has a good prediction effect and can improve the prediction work of short-term traffic flow.
