Abstract: With the development of new energy vehicles, it has become an important means to acquire experimental analysis data by collecting controller area network (CAN) bus signal in vehicle benchmarking test. CAN bus protocol is a confidential document for the vehicle manufacturer, so it is necessary to research a method to reverse crack the CAN bus signal of the vehicle. The research on CAN signal analysis method based on similarity principle is to use mathematical algorithm to realize signal positioning in the bus report by covering the typical features of some signals under specific conditions and then covering the common features of different vehicles. This method saves the time of manual processing a large number of CAN bus data, filters out the ID of the signal quickly and effectively, and realizes the cracking of the CAN bus signal of the vehicle. Aiming at the reverse cracking of vehicle can bus signal, this paper proposes to quickly locate the position of the required signal in the bus message through the similarity principle and algorithm screening, so as to reverse analyze the required signal. The results show that this method can quickly crack most can signals in vehicle benchmarking, and improve the accuracy and efficiency of analysis.

Key words: Controller area network bus; Reverse cracking of signals; Similarity principle; Mathematical algorithm

摘要： 随着新能源汽车研发的不断深入，在车辆对标测试中，通过采集控制器局域网络（CAN） 总线信号来获取实验分析数据成为一种重要手段。CAN 总线协议对于车辆制造厂是一项保密 文件，因此，研究一种逆向解析车辆的 CAN 总线信号的方法是必要的。基于相似性原理的 CAN 信号解析方法研究，是通过某些信号在特定工况下所表现的典型特征，进而覆盖到不同 车辆共有特征，利用数学算法实现信号在总线报文中定位。这种方法节省了人工处理大量 CAN 总线数据的时间，比较快速有效地过滤出信号所在的 ID，实现对车辆的 CAN 总线信号的解 析。文章针对车辆 CAN 总线信号逆向解析，提出通过相似性原理，经过系统数学算法筛选， 快速定位所需信号在总线报文中的位置。研究结果表明，该方法适用于车辆对标测试中大部 分 CAN 信号，从而提高 CAN 信号解析的准确性和工作效率。

关键词: 控制器局域网络总线；信号逆向解析；相似性原理；数学算法