摘要
机载电子电气设备检测容易受到外界干扰,出现参数辨识不精准,导致检测结果不精准的情况。针对该问题,在智能网联技术下设计机载电子电气设备检测系统。使用嵌入式检测适配器,对输出信号进行放大、隔离、变换处理。结合防错插设计方式,确保检测装置安全连接。利用光耦合器实现信号的驱动与变换,使用智能网联技术聚集特征向量构建密度函数,获取聚类同期运行的峰值参数。归一化处理智能网联聚类参数,计算实际与理论辨识参数之间的差值,以此为依据判断设备运行状态。以报文ID数据标识为特征,计算时间窗口为1的信息熵值,实现固定时间窗口下的设备检测。由实验结果可知,该系统在三种场景下与实际熵值的最大误差为1.2,说明使用该系统检测结果较精准。
Airborne electronic and electrical equipment detection is prone to external interference and inaccurate parameter identification,resulting in inaccurate detection results.Aiming at this problem,the airborne electronic and electrical equipment detection system is designed under the intelligent network technology.The embedded detection adapter is used to amplify,isolate and transform the output signal.Combined with the wrong insertion prevention design,the safety connection of the detection device is ensured.The optical coupler is used to drive and transform the signal,and the intelligent network technology is used to gather the feature vectors to build the density function,and obtain the peak parameters of the cluster running in the same period.Normalize the clustering parameters of the intelligent network,calculate the difference between the actual and theoretical identification parameters,and judge the equipment operation status based on this.Based on the message ID data identification,the information entropy value with a time window of 1 is calculated to realize the equipment detection under a fixed time window.According to the experimental results,the maximum error between the system and the actual entropy value in the three scenarios is 1.2,which indicates that the detection results of the system are accurate.
作者
孙小伟
张启浩
李琦
刘畅
张东海
SUN Xiaowei;ZHANG Qihao;LI Qi;LIU Chang;ZHANG Donghai(Kunming UHV Power Transmission Bureau of China Southern Power Grid Co.,Ltd.,Kunming 650214,China;Yunnan Zhaofu Technology Co.,Ltd.,Kunming 650217,China)
出处
《电子设计工程》
2024年第4期153-157,共5页
Electronic Design Engineering
关键词
智能网联技术
机载电子
电气设备
参数辨识
intelligent network technology
airborne electronic
electrical equipment
parameter iden-tification