摘要
为了精准预测交通流量,充分提取交通流中复杂的线性和非线性特征及其依赖关系,提出了融合多维时空特征的CLABEK模型。其中,由Conv-LSTM、BiLSTM和Dense神经网络分别提取时空特征、周期特征和额外特征(节假日、天气状况以及温度等),并通过将上述模型融合从而全面获取交通流的非线性特征;由卡尔曼滤波提取交通流的线性特征。在公开数据集上的对比实验证明,CLABEK模型在短期交通流预测任务上表现出最好的预测效果。
In order to accurately predict the traffic flow and fully extract the complex linear and nonlinear features and their dependence in the traffic flow,a CLABEK model integrating multi-dimensional spatial-temporal features was proposed.Among them,the temporal and spatial features,periodic features and additional features(holidays,weather conditions and temperature)were extracted by Conv-LSTM,BiLSTM and Dense neural networks respectively,and the nonlinear features of traffic flow were comprehensively obtained by combining the above models.Then the linear features of traffic flow were extracted by Kalman filtering.The comparative experiments on public data sets show that the CLABEK model performs the best in short-term traffic flow prediction task.
作者
宋瑞蓉
王斌君
仝鑫
刘文懋
SONG Rui-rong;WANG Bin-jun;TONG Xin;LIU Wen-mao(College of Police Information Engineering and Network Security, People’s Public Security University of China, Beijing 100038, China;Nsfocus Information Technology Co.,Ltd., Beijing 100089, China)
出处
《科学技术与工程》
北大核心
2021年第31期13439-13446,共8页
Science Technology and Engineering
基金
国家社会科学基金重点项目(20AZD114)
公安部科技强警基础工作专项(2018GABJC03)
CCF-绿盟科技“鲲鹏”科研基金(CCF-NSFOCUS 2020011)。
关键词
交通流量
时空融合
神经网络
卡尔曼滤波
traffic flow
integration of time and space
neural network
Kalman filtering
