大数据驱动的新能源汽车多维度安全预警建模方法研究

新能源汽车安全已是一个非常严峻的现实问题,直接影响到产业的未来与发展。综合阐述了基于大数据驱动的安全预警技术建模的方法学问题。从多个层次、多个维度讨论了安全预警的技术方法,重点分析提出了基于系统稳定性、相关性、一致性等累计风险模型与事故特征匹配追踪两大类模型。介绍稳态概率计算方法、累计风险计算方法、风险曲线识别方法。介绍了基于中心距、能量一致性的累计风险模型与应用情况,最后介绍一类全新的模式匹配识别模型:事故特征匹配追踪模型。总结提出了从微观定义安全、宏观预警安全、智能完善安全、数据保障安全一个比较完整的安全预警体系与方法。

Research on Multi-dimensional Safety Early Warning Modeling Method Based on a New Energy Vehicle Driven by Big Data

Electric vehicle (EV) safety has been a very serious practical problem, directly affects the future and development of the EV industry. This paper comprehensively expounds the methodological problems of the security early warning technology modeling based on big data. This paper discusses the technical methods of safety early warning from many levels and dimensions. It puts forward two kinds of models based on system stability, correlation, consistency and so on. The steady-state probability calculation method, cumulative risk calculation method and risk curve identification method are introduced. This paper introduces the cumulative risk model and application based on center distance and energy consistency. Finally, there is a new pattern matching recognition model: accident feature matching tracking model. This paper will summarize and put forward a relatively complete security early warning system. The method from micro definition security, macro early warning security, intelligent perfect security and data guarantee security.