The invention relates to a road traffic speed prediction method fused with a multi-feature neural network, and the method comprises the steps: carrying out the preprocessing of feature data, training a multi-feature input MF-Informer traffic speed prediction model through a data set, and carrying out the input standardization of a multi-source traffic speed prediction model through the preprocessing of the feature data; training a multi-feature input MF-Informer traffic speed prediction model by using the data set to realize prediction; the MF-Informer traffic speed prediction model comprises a position information embedding module, an encoder module and a decoder module. The model is continuously optimized through reverse gradient propagation of a prediction result output by a full connection layer and a real value calculation error, finally traffic speed prediction is performed on a prediction time period by using the trained model, output data is obtained after reverse normalization is performed on the output prediction result, and traffic speed prediction model evaluation is completed. According to the invention, multi-dimensional feature fusion is realized through combination of the floating car data, the meteorological data and the air pollution data, and the accuracy of traffic speed prediction is effectively improved.
本发明涉及一种融合多特征神经网络的道路交通速度预测方法,包括:特征数据预处理和利用数据集训练多特征输入的MF‑Informer交通速度预测模型,通过特征数据预处理,对多源交通速度预测模型的输入规范化;利用数据集训练多特征输入的MF‑Informer交通速度预测模型实现预测;所述MF‑Informer交通速度预测模型,包括:位置信息嵌入模块、编码器模块和解码器模块。经全连接层输出预测结果与真实值计算误差反向梯度传播不断对模型进行优化,最后使用训练好的模型对预测时段进行交通速度预测,输出预测结果进行反归一化后得到输出数据,完成交通速度预测模型评估。本发明通过浮动车数据与气象数据、空气污染数据相结合实现多维度特征融合,有效提高了交通速度预测的准确性。
Road traffic speed prediction method fused with multi-feature neural network
融合多特征神经网络的道路交通速度预测方法
2023-09-12
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
ROAD TRAFFIC SPEED PREDICTION METHOD FUSING MULTI-FEATURE NEURAL NETWORK
European Patent Office | 2024
|Transform-fused hypergraph neural network-based high-speed station traffic flow prediction method
European Patent Office | 2024
|Urban road network traffic speed prediction method and related equipment
European Patent Office | 2025
|Road Traffic Accident Prediction Based on BP Neural Network
Springer Verlag | 2022
|Sparse Data Traffic Speed Prediction on a Road Network With Varying Speed Levels
Transportation Research Record | 2023
|