大数据环境下道路场景高时空分辨率众包感知方法

Road crowd-sensing with high spatio-temporal resolution in big data era

道路场景作为人类发展演变中最重要、最复杂、最典型的载体之一,是道路基础设施与活动行为共同构成的综合体,链接和构建“人地关系”。道路场景感知从二维抽象简略到三维精细丰富、从静态过去式向动态现在时发展,成为智慧城市、智能交通、无人驾驶的关键技术支撑,是我国新型城镇化战略、交通强国战略的核心技术保障。本文立足于时空交通大数据,提出基于道路场景静态基础设施“形”和动态活动行为“流”的高时空分辨率道路场景感知方法。该方法从静态路网“形”角度,以“点-线-面-体”等要素为研究脉络,构建高精度道路地图众包感知的理论体系;在活动行为“流”感知上,突破传统的点模式分析局限,发展了道路网络空间活动流的时空建模与多尺度分析方法。本文揭示了静态基础设施“形”结构与动态活动行为“流”模式交互作用下的道路场景演化规律,形成以“形”控“流”、借“流”定“形”、“形”“流”叠置的高时空精度道路场景众包感知理论体系,为智慧城市、智能交通的发展提供核心技术与数据支撑。

As one of the most important, complex, and typical carriers in the evolution of human development, the road scene is a complex of road infrastructure and activity behavior, linking and constructing the “man-land relationship”. Road scene perception has developed from two-dimensional abstraction to three-dimensional refinement, from static past tense to dynamic present tense. It has become the key technical support for smart cities, intelligent transportation, and autonomous driving, and is the core technical guarantee for China’s new urbanization strategy and strong transportation strategy. Based on spatio-temporal traffic data, this paper proposes a new method for road crowd-sensing with high spatio-temporal resolution based on static infrastructure “form” and dynamic activity behavior “flow”. From the perspective of static road network “form”, the method takes “point-line-surface-body” elements as the research context, and constructs a theoretical system of high-precision road map crowd-sensing. In terms of activity behavior “flow”, we break through the limitations of traditional point pattern analysis and develop a spatio-temporal modeling and multi-scale analysis method for spatial activity flow. In this paper, we reveal the evolution pattern of road scenes under the interaction of static infrastructure “form” structure and dynamic activity behavior “flow” pattern. Furthermore, we develop a road crowd-sensing theoretical system in which “form” controls “flow”, “flow” determines “form”, and “form” overlaps “flow”, to provide core technology and data support for the development of smart cities and intelligent transportation.