Developing an Innovative Platform for Modeling Active Road User Interactions and Safety: Integration of Computer Vision, Agent-based, and Machine Learning Models

开发用于对主动道路用户交互和安全进行建模的创新平台：计算机视觉、基于代理和机器学习模型的集成

• 批准号: RGPIN-2019-06688
• 负责人: Hussein, Mohamed
• 金额: $ 2.26万
• 依托单位: McMaster University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=749362
• 关键词: Developing; Innovative; Platform; Modeling; Active

Promoting active modes of transportation, such as walking and biking, is receiving more emphasis in transportation engineering as they represent vital contributors to transportation sustainability and have several benefits to public health and the environment. Nevertheless, active travellers are the most vulnerable road users, as they are at a higher risk of being killed or severely injured due to road collisions. There is an urgent need to develop a solid understanding of active road user interaction behaviour and adopt innovative techniques to enhance their safety. As such, an innovative platform for analyzing and modeling active road user interactions and enhancing their safety is proposed in this research program. The proposed platform involves three interrelated focus areas. Focus Area I involves the acquisition of a variety of active road user datasets (mainly, video data collected at locations of interest and GPS data) and applying innovative analytical tools, including machine learning techniques and computer vision, to analyze the collected data. The analysis will investigate active road user interactions with each other, interactions with motorized vehicles (conflicts), and road user violation behaviour (jaywalking and temporal violations), and explore the implications of these behaviours on road safety. In Focus Area II, a novel agent-based micro-simulation model for modeling detailed road user interactions will be developed. The model relies on artificial intelligence principles and utilizes the agent-based modeling approach to simulate active road user interactions. The model will provide a powerful tool for addressing a variety of applications related to road user safety and assessing the level of service of active transportation facilities. The proposed model will enable the analysis of different design alternatives before making any changes to existing conditions and investigating scenarios that might be difficult to explore using actual data. Focus Area III involves developing a comprehensive framework for analyzing active road user safety through the advanced statistical modeling of collision and traffic conflict data on both the macro and the micro levels. In addition, the research will establish the statistical relationship between road user collisions and conflicts. The proposed research will create powerful analytical tools to support the design of safe transportation facilities for active road users of all ages and abilities. The proposed platform will aid transportation engineers and planners to promote and accommodate active transportation activities, which contributes to public health, safety, and reducing GHG emission. It will also enable to consider road user safety in the early planning stages. The adoption of the proposed platform is expected to save thousands of lives, either directly by improving road safety or indirectly by improving air quality and increasing the physical activity level of the population.

促进积极的运输方式，例如步行和骑自行车，在交通工程上更加重视，因为它们代表了运输可持续性的重要贡献者，并为公共卫生和环境带来了一些好处。然而，活跃的旅行者是最脆弱的道路使用者，因为由于道路碰撞，他们的风险更高。迫切需要对积极的道路用户互动行为有深入的了解，并采用创新技术来提高其安全性。因此，在本研究计划中提出了一个创新的平台，用于分析和建模主动道路用户互动并增强其安全性。提出的平台涉及三个相互关联的焦点区域。焦点区域I涉及获取各种活跃的道路用户数据集（主要是在感兴趣的位置和GPS数据的位置收集的视频数据），并应用创新的分析工具，包括机器学习技术和计算机视觉，以分析收集的数据。该分析将调查彼此之间的积极道路用户互动，与电动车辆（冲突）以及道路用户违规行为（Jaywalking和临时违规）的互动，并探讨这些行为对道路安全的影响。在焦点区域II中，将开发一种基于新颖的基于代理的微模拟模型，用于建模详细的道路用户交互。该模型依靠人工智能原则，并利用基于代理的建模方法模拟主动的道路用户交互。该模型将为解决与道路用户安全相关的各种应用程序提供强大的工具，并评估主动运输设施的服务水平。拟议的模型将在对现有条件进行任何更改并调查可能难以使用实际数据探索的方案之前对不同的设计替代方案进行分析。重点区域III涉及开发一个综合框架，通过对宏观和微观水平的碰撞和交通冲突数据进行高级统计建模来分析主动道路用户安全。此外，研究将建立道路用户碰撞与冲突之间的统计关系。拟议的研究将创建强大的分析工具，以支持各个年龄段和能力的活跃道路使用者的安全运输设施的设计。拟议的平台将帮助运输工程师和规划人员促进和接受积极的运输活动，这有助于公共卫生，安全和减少温室气体排放。它还可以在早期计划阶段考虑道路用户安全。预计通过提高道路安全性或间接提高空气质量并提高人口的体育锻炼水平，预计采用拟议的平台将挽救数千人的生命。