| 基于TS-NN模型的道路交通车流量预测 |
| |
| 引用本文： | 张扬永.基于TS-NN模型的道路交通车流量预测[J].福建工程学院学报,2021,0(6):560-567. |
| |
| 作者姓名： | 张扬永 |
| |
| 作者单位： | 中共福建省委党校 |
| |
| 摘 要： | 针对现有的智能交通系统预测方法，基于道路交通的关键参数车流量预测，提出了一种基于深度学习的时间序列交通流预测方法，进一步提升道路交通车流量预测准确率。在对道路交通数据集进行清洗后，使用时间序列和神经网络的结合算法TS-NN 进行车流量预测，实验表明，在城市路段的预测中，TS-NN 相对时间序列模型ARIMA、神经网络模型LSTM 准确率分别提升了1.62%和2.13%?在高速公路数据集上测试上，TS-NN 有更加明显的改进，相对ARIMA、LSTM 分别提升了20.87%和3.53%，在一定程度上，TS-NN 算法确实有助于改进智能交通系统核心算法。
|
| 关 键 词： | 时间序列 神经网络 道路交通 车流量预测 |
Prediction of road traffic flow based on the TS-NN model |
| |
| ZHANG Yangyong.Prediction of road traffic flow based on the TS-NN model[J].Journal of Fujian University of Technology,2021,0(6):560-567. |
| |
| Authors: | ZHANG Yangyong |
| |
| Affiliation: | Fujian Provincial Committee Party School of CPC, Fujian Academy of Governance |
| |
| Abstract: | For the existing intelligent traffic system prediction method, a time-series traffic flow prediction method based on deep learning, starting from the key parameters of traffic flow prediction, was proposed to further improve the accuracy of road traffic flow prediction rate. Firstly, the road traffic data set was cleaned, and then TS-NN, the fusion algorithm of time series and neural network, was used for traffic flow prediction. Experimental results show that the prediction accuracy of the TS-NN algorithm in urban sections is improved by 1.62% and 2.13% respectively, compared with that of ARIMA and LSTM. Besides, there are more obvious improvements in the test of the expressway dataset, which are 20.87% and 3.53% higher than ARIMA and LSTM respectively. Therefore, the TS-NN algorithm does contribute to the improvement of its core algorithm to a certain extent. |
| |
| Keywords: | time series neural network road traffic traffic flow prediction |
|
| 点击此处可从《福建工程学院学报》浏览原始摘要信息 |
|
点击此处可从《福建工程学院学报》下载全文 |
|
