基于深度学习的物流车队管理方法与应用研究

With the development of the economic environment and the evolution of mobile communications, the intelligent transportation system has been more and more popular for obtaining fleet management services for the logistics industries and bus carriers. A logistics company may have hundreds or thousands of trucks to provide freight services. However, the fuel cost of these industries is the most important challenge of fleet management services. The research collects practical data as training data and testing data from Fujian Provincial Department of Transportation and explores six scientific problems which include: (1) filtering out interference factors in mobile networks; (2) analyzing the time series data of mobile network signals; (3) analyzing the space-time correlation of travel time; (4) analyzing the time series data of travel time; (5) estimating fuel consumption; (6) planning routes. The main innovations　include: (1) proposing a mobile positioning method with higher accuracy; (2) proposing a spatial-temporal travel time prediction method; (3) proposing a fuel consumption estimation method with lower cost and a route planning method with lower fuel consumption. The research results of this　topic will improve and apply deep learning techniques in　various fields, such as intelligent logistics, intelligent transportation system, digital city, etc.

随着经济环境的发展和移动通信的发展，智能交通系统在物流行业和公交运营商获得车队管理服务方面越来越受欢迎。物流公司可能有成百上千辆卡车提供货运服务。然而，这些行业的燃油成本是车队管理服务面临的最重要挑战。因此，本课题以福建省交通运输厅真实数据作为训练数据和测试数据，并探索6个科学问题，包括：(1) 移动网络干扰因子过滤问题；(2) 移动网络讯号时间序列问题；(3) 旅行时间之时空关联问题；(4) 旅行时间之时间序列问题； (5) 油耗估计问题；(6) 路线规划问题。主要创新点包括：（1）拟提出高精准度移动网络定位方法；（2）拟提出融合时间和空间信息旅行时间预测方法；（3）拟提出低成本的油耗估计方法和低油耗路线规划方法。本课题的研究成果将改进现有深度学习技术在智能物流、智能交通系统、数字城市等领域的应用。

本课题以福建省交通运输厅真实数据作为训练数据和测试数据，并探索6个科学问题，包括：(1) 移动网络干扰因子过滤问题；(2) 移动网络讯号时间序列问题；(3) 旅行时间之时空关联问题；(4) 旅行时间之时间序列问题； (5) 油耗估计问题；(6) 路线规划问题。主要创新点包括：（1）拟提出高精准度移动网络定位方法；（2）拟提出融合时间和空间信息旅行时间预测方法；（3）拟提出低成本的油耗估计方法和低油耗路线规划方法。