基於偏微分方程的模擬，機器學習工具及其混合技術在交通流量預測中的比較研究;A comparative study of PDE based simulation, machine learning tool, and their hybrid technique for traffic flow prediction

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/77841

Title:	基於偏微分方程的模擬，機器學習工具及其混合技術在交通流量預測中的比較研究;A comparative study of PDE based simulation, machine learning tool, and their hybrid technique for traffic flow prediction
Authors:	吳柏汎;Wu, Bo-Fan
Contributors:	數學系
Keywords:	機器學習;交通流模型;數值模擬;machine learning;traffic flow model;numerical simulation
Date:	2018-07-25
Issue Date:	2018-08-31 14:59:06 (UTC+8)
Publisher:	國立中央大學
Abstract:	現今以深度學習方式預測交通流量、旅行時間的相關研究已經非常純熟，但是結合交通基本理論的相關討論研究比較缺乏，所以此篇論文探討的是結合交通流理論模型、深度學習、非線性雙曲型守恆定律，預測高速公路下交通流量預測，深度學習、數值模擬、結合數值模擬跟深度學習模型的比較研究。;Using deep learning model to predict traffic flow nowadays is a very popular method for research, but most of the traffic flow research only put data into deep learning model without traffic flow fundamental theorem. We combine the numerical simulation which is partial differential equation model with deep learning model which is recurrent neural networks model and predict the traffic flow. In partial difference equation, there is some theorem of traffic flow, and we have assumption by the theorem. We use a machine learning tool, partial differential equation based numerical simulation, and their hybrid technique for traffic flow prediction.
Appears in Collections:	[Graduate Institute of Mathematics] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	264	View/Open

社群 sharing

Loading...