The invention provides a traffic flow prediction method based on group normalization and gridding cooperation. The method comprises the steps: obtaining traffic flow data, and carrying out the preprocessing of the traffic flow data; dividing the preprocessed data into g groups, wherein the number of samples in each group is m, and an m*F vector is formed; performing vector format conversion on each group of sample data, namely converting the m*F vector into an M*T*F vector, and performing group normalization operation, wherein T is a time interval, M is the number of samples in the T time interval, F is a feature number; dividing the normalized samples of each sample group into training samples and test samples, wherein the sample numbers are respectively N and n; putting the obtained training samples into a grid loop network; and finally, obtaining an optimal traffic flow prediction model through linear judgment. The method has the main contribution that the defect that an LSTM and aGRU need to wait for the output of a previous moment at the next moment is overcome in the loop network, the model is simplified, and computing resources are saved.
本发明提供一种基于组归一化和网格化协同的交通流预测方法,通过获取交通流量数据,将交通流量数据进行预处理;将预处理后的数据,分成g组,每组样本数为m,形成m×F的向量;对每组样本数据进行向量格式转换,即将m×F的向量转换成M×T×F向量,其中T为时间间隔,M是T时间间隔内的样本数,F是特征数,并进行组归一化操作;将每组样本组归一化后的样本分为训练样本和测试样本,样本数分别为N和n;将得到的训练样本代入网格循环网络;最后经过线性判决得到最优的交通流量预测模型。本发明主要贡献在于循环网络中解决了LSTM和GRU后一时刻需要等待前一时刻输出的弊端,简化了模型,节约了计算资源。
Traffic flow prediction method based on group normalization and gridding cooperation
基于组归一化和网格化协同的交通流预测方法
2020-05-19
Patent
Electronic Resource
Chinese
Traffic flow time sequence prediction method based on double-domain normalization
European Patent Office | 2024
|Urban traffic accident early warning method based on space-time gridding data
European Patent Office | 2020
|Traffic flow prediction system and method based on vehicle-road cooperation
European Patent Office | 2015
|Research of Airspace Gridding Method and Its Application in Air Traffic Management
DOAJ | 2021
|Gridding of satellite observations
AIAA | 1964
|