The invention discloses a light-weight traffic flow prediction method based on deep learning, and the method comprises the steps: building a plurality of space-time attention mechanisms to excavate the space-time dependence of traffic flow, carrying out the processing of output data through a convolution module, carrying out the light-weight processing of a model through employing a pruning and distillation technology, removing the redundant parameters in a prediction model, and carrying out the prediction of the light-weight traffic flow. And a lightweight prediction model with relatively high accuracy is obtained. The method solves the problem that a microcontroller equipped for traffic infrastructures in a real road network environment is small in memory and is insufficient to deploy the most advanced traffic flow prediction model, realizes balance between traffic flow prediction precision and model calculation cost, is high in prediction precision and small in calculation memory, and is suitable for large-scale popularization and application. And a theoretical basis is laid and technical support is provided for high-precision traffic flow prediction in the microcontroller.
本发明公开了一种基于深度学习的轻量化交通流预测方法,通过构建多个时空注意力机制挖掘交通流的时空依赖性,并利用卷积模块对输出数据进行处理,最后使用剪枝和蒸馏技术对模型进行轻量化处理,去除预测模型中的冗余参数,得到具有较高准确率的轻量预测模型。本发明解决了真实路网环境下交通基础设施配备的微控制器内存较小,不足以部署最先进的交通流预测模型的问题,实现了交通流预测精度和模型计算成本之间的平衡,不仅预测精度高且计算占用内存小,为实现在微控制器中进行精度较高的交通流预测奠定理论基础并提供了技术支持。
Lightweight traffic flow prediction method based on deep learning
一种基于深度学习的轻量化交通流预测方法
2024-07-26
Patent
Electronic Resource
Chinese
IPC: | G06Q Datenverarbeitungssysteme oder -verfahren, besonders angepasst an verwaltungstechnische, geschäftliche, finanzielle oder betriebswirtschaftliche Zwecke, sowie an geschäftsbezogene Überwachungs- oder Voraussagezwecke , DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Traffic flow prediction method based on hybrid deep learning
European Patent Office | 2023
|Urban traffic flow prediction method based on deep learning
European Patent Office | 2025
|Traffic Flow Prediction Model Based on Deep Learning
British Library Conference Proceedings | 2022
|Short-term traffic flow prediction method based on deep learning
European Patent Office | 2025
|