基于FFT卷积的计算机故障检测云平台  

Cloud platform for computer fault detection based on FFT convolution

作　　者：姚强 王晓红 Yao Qiang;Wang Xiaohong(Changzhou Vocational College of Technology,Changzhou 213164,China)

机构地区：[1]常州市高级职业技术学校

出　　处：《无线互联科技》2022年第16期148-150,共3页Wireless Internet Technology

基　　金：江苏省职教学会2021—2022年度职业教育研究课题,项目名称:云技术背景下培养中职学生创新能力的实践研究,项目编号:XHYBLX2021164。

摘　　要：文章利用大数据分析云计算技术,通过获取电路板上Debug口数据,对波形进行整形,将信号进行叠加处理;通过FPGA芯片进行FFT蝶形运算,获取基波分量与谐波分量,进入集成运放电路;通过激活函数f(x)=1/(1+e^(-x))分析,选取不同的放大倍数,获得稳定的输出电压,将电压输入ARM STM32F407芯片,12位A/D转换成数字信号,然后根据数字信号进行10位编码。当电路板出现故障时,10位编码各不相同,以此判定故障。This paper uses big data analysis and cloud computing technology to obtain the debug port data on the circuit board,shape the waveform,superimpose the signal,perform FFT butterfly operation through the FPGA chip,obtain the fundamental component and harmonic component,enter the integrated operational amplifier circuit,and select different amplification factors to obtain a stable output voltage through the analysis of the activation function f(x)=1/(1 + e^(-x)),The voltage is input into ARM STM32 F407 chip,12 bit A/D is converted into digital signal,and then 10 bit coding is carried out according to the digital signal.When the circuit board fails,the 10 bit codes are different,so as to determine the fault.

关 键 词：大数据分析 FFT卷积 蝶形运算 FPGA 激活函数 编码 

分 类 号：TP306[自动化与计算机技术—计算机系统结构]