The invention discloses an urban road network vehicle path reconstruction method based on a bidirectional long short-term memory network. The method mainly comprises the steps of firstly, cleaning AVI detection data, processing the AVI detection data into vehicle path information, and dividing extracted vehicle paths into complete paths and incomplete paths according to path connectivity; secondly, the vehicle path information is processed into character string data according to node numbers, a word library is established for the node numbers, semantic coding is carried out, and semantic coding strings are obtained; and finally, taking the processed semantic coding string as the input of the bidirectional long-short-term memory neural network, carrying out iterative updating, carrying out model convergence, extracting a trained neural network model, carrying out coding cycle on the incomplete path node number string, inputting the incomplete path node number string into the trained neural network model, and obtaining node numbers of all missing positions. And outputting the reconstructed complete vehicle path.
本发明公开了一种基于双向长短期记忆网络的城市路网车辆路径重构方法。主要步骤包括:首先,将AVI检测数据清洗并处理为车辆路径信息,对提取后的车辆路径按照路径连通性划分为完整路径和不完整路径;其次,对车辆路径信息按照节点编号处理为字符串数据,对节点编号建立词库,进行语义化编码,得到语义编码串;最后,将处理后的语义编码串作为双向长短期记忆神经网络的输入,迭代更新,模型收敛,提取训练后的神经网络模型,对不完整路径节点编号串进行编码循环输入训练后的神经网络模型,得到所有缺失位置的节点编号,输出重构后完整的车辆路径。
Urban road network vehicle path reconstruction method based on bidirectional long short-term memory network
一种基于双向长短期记忆网络的城市路网车辆路径重构方法
2023-10-24
Patent
Electronic Resource
Chinese
IPC: | G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / 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 |
Transportation Research Record | 2023
|Taylor & Francis Verlag | 2023
|European Patent Office | 2022
|European Patent Office | 2021
|