基于抗干扰性和准确性协同提升的ACNS门槛阈值与路面平整度的匹配理论

Note that the vehicle Automatic Crash Notification System (ACNS) triggered by crash acceleration signal has a high false alarm rate and the theoretical basis for setting the discrimination threshold in the accident detection algorithm is absent. From the perspective of synchronously improving anti-interference ability and accuracy, the matching relationship between discrimination threshold and pavement roughness is studied. First, based on the real vehicle test and the simulation, the correlation model of discrimination threshold and International Roughness Index (IRI) is constructed; velocity characteristics of discrimination threshold are analyzed. Next, based on the vehicle dynamics model and data analysis, the correspondence between body acceleration signal and IRI is quantified; the correlation model of body acceleration power spectral density and IRI is formulated; the sensitivity analysis about the effect of the velocity on this model is performed. Next, the types and characteristics of the information influenced by discrimination threshold are discussed; the change regulation of information offset with discrimination threshold is investigated and the relationship between them is modeled; a self-correcting method of accident information based on discrimination threshold is proposed. Finally, the theory for matching discrimination threshold for pavement roughness is presented and verified by the experiments. The research results can change the existing methods for determining discrimination threshold which rely on experiences as well as are isolated and static, provide a theoretical basis for the selection of the discrimination threshold, provide a support for the detection algorithm and lay the foundation for intelligent ACNS.

针对以碰撞加速度信号触发的车辆事故自动呼救系统（ACNS）误报警率高且其事故检测算法中的门槛阈值的设定缺乏理论依据的难题，从协同提升抗干扰性和准确性的角度，研究门槛阈值与路面平整度的匹配关系。为此，基于实车试验及仿真数据，构建门槛阈值与国际平整度指数IRI的关联模型，分析门槛阈值的速度特性；基于车辆动力学模型，结合数据分析的手段，量化车身加速度信号与IRI的对应关系，构建车身加速度功率谱密度与IRI的关联模型，分析该模型的速度敏感性；界定受门槛阈值设定影响的事故信息的类型及特点，探究信息的偏移量随门槛阈值变化的规律，构建两者的关联模型，研究事故信息的自校正方法；最终形成门槛阈值与路面平整度的匹配理论，通过试验对所建立的理论进行验证。研究将改变现有的从经验出发孤立、静止的门槛阈值确定方法，为门槛阈值的选取提供理论依据，为事故检测算法可靠性的相关研究提供支持，为实现智能化的ACNS奠定基础。

本项目从协同提升车辆事故自动呼救系统（Automatic Crash Notification System, ACNS）抗干扰性和准确性的角度，研究门槛阈值与路面平整度的匹配关系。为此，构建了三维虚拟道路-车辆仿真模型，建立了门槛阈值与国际平整度指数（International Roughness Index, IRI）的关联模型；通过车辆动力学模型和道路-车辆仿真模型的结合使用，构建了车身加速度功率谱密度与IRI的关联模型，分析了该模型的车速敏感性；通过台车碰撞仿真模型，建立了速度变化量误差与门槛阈值的关联模型，并提出速度变化量的校正方法；基于移动窗积分算法，结合之前所建立的关联模型，提出具有自匹配门槛阈值功能的ACNS事故检测算法；对ACNS车载终端进行设计，通过台车和实车碰撞试验，就改进算法对于速度变化量信息准确性的影响进行验证。本项目研究所提出的门槛阈值与路面平整度的匹配模型，为门槛阈值选取提供理论依据，突破现有相关研究中静止的门槛阈值确定方法，为在车-路协同背景下提高ACNS抗干扰性和准确性的相关研究奠定技术基础。