The invention relates to the technical field of traffic flow analysis, and discloses a flow trend prediction method for regional dynamic traffic monitoring based on a graph convolutional neural network, and the method specifically comprises the following steps: S1, obtaining the position data of a gate; s2, accessing the traffic flow data of the checkpoint; s3, calculating traffic flow data; s4, constructing a weighted directed graph, taking the road intersection, the fixed gate equipment and the cruise unmanned aerial vehicle equipment as nodes, taking a middle road of a line segment between any two points of any intersection, the fixed gate equipment and the movable gate equipment as an edge, and taking the average passing speed of the normalized current gate identification vehicle as the weight of the corresponding road as the edge, so as to obtain a weighted directed graph of the road intersection, the fixed gate equipment and the cruise unmanned aerial vehicle equipment; circulating the operation to form a weighted directed graph; s5, extracting the feature representation of the graph, converting the graph into an algebraic form by using a spectrogram method, and performing Fourier transform on the algebraic representation of the graph to obtain the feature representation of the graph; s6, updating a checkpoint traffic flow prediction result table; and S7, reporting a bayonet traffic flow prediction result.
本发明涉及交通流量分析技术领域,公开了一种基于图卷积神经网络的区域动态交通监控的流量趋势预测方法,具体预测方法包括以下步骤:S1:获取卡口位置数据;S2:接入卡口车流量数据;S3:计算车流量数据;S4:构建带权有向图,以道路路口、固定卡口设备和巡航无人机设备为节点,以任意的路口、固定卡口设备和可移动卡口设备任意两点之间的线段中间道路为边,以归一化的当前卡口识别车辆的平均经过速度作为对应道路为边的权重,并循环操作,形成带权有向图;S5:抽取图的特征表示,使用谱图方法将图转化为代数形式,对图代数表示进行傅里叶变换得到图的特征表示;S6:更新卡口车流量预测结果表;S7:上报卡口车流量预测结果。
Flow trend prediction method for regional dynamic traffic monitoring based on graph convolutional neural network
一种基于图卷积神经网络的区域动态交通监控的流量趋势预测方法
2023-05-02
Patent
Electronic Resource
Chinese
European Patent Office | 2024
|Traffic flow prediction method based on graph convolutional network
European Patent Office | 2025
|Multi-scale traffic flow prediction method based on graph convolutional neural network
European Patent Office | 2021
|Traffic flow prediction method based on multimode dynamic memory graph convolutional network
European Patent Office | 2024
|Traffic flow prediction method based on adaptive dynamic fusion graph convolutional network
European Patent Office | 2024
|