• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>COTA international conference of transportation professionals >Application of Long Short-Term Memory Neural Network for Multi-Step Travel Time Forecasting on Urban Expressways
【24h】

Application of Long Short-Term Memory Neural Network for Multi-Step Travel Time Forecasting on Urban Expressways

机译：长短期记忆神经网络在城市高速公路多步行程时间预测中的应用

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Deep learning theory, as a powerful set of techniques for learning in neural networks, has achieved numerous successes in the domains of artificial intelligence (AI). Based on huge amounts of probe vehicle data, this study attempts to apply the novel deep learning theory to the multi-step travel time forecasting on urban expressways. Serving as state-of-the-art neural network architecture, long short-term memory neural networks (LSTMs) are employed due to their superiority to time series methods. LSTMs enable capture of the underlying structure of data from a pool of historical datasets and excavate the intrinsic traffic features without strong assumptions on their temporal evolution. The proposed approach is investigated on Ring 2, a 33 km urban expressway of Beijing, China. The results demonstrate the advantage of the proposed method, as well as its feasibility and effectiveness compared with other prevailing parametric and nonparametric algorithms.
机译：深度学习理论作为一种强大的神经网络学习技术，已经在人工智能（AI）领域取得了许多成功。基于大量的探测车辆数据，本研究尝试将新颖的深度学习理论应用于城市高速公路的多步行程时间预测。作为长期使用的神经网络体系结构，由于长时记忆神经网络（LSTM）优于时间序列方法，因此采用了长期短期记忆神经网络（LSTM）。 LSTM可以从一组历史数据集中捕获数据的底层结构，并挖掘内部流量特征，而无需对其时间演变进行强有力的假设。在中国北京33公里的城市高速公路2环上对提出的方法进行了研究。结果证明了该方法的优势，以及与其他流行的参数和非参数算法相比的可行性和有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号