基于STM32的A/D采样软件滤波改进算法研究  

Research on Improved Algorithm of A/D Sample Software Filtering Based on STM32

作　　者：黄健[1] 张善文[1] 周端[2] 

机构地区：[1]西京学院电子信息工程系,陕西西安710123 [2]西安电子科技大学计算机学院,陕西西安710071

出　　处：《仪表技术与传感器》2016年第3期83-85,共3页Instrument Technique and Sensor

基　　金：国家自然科学基金项目(61473237)

摘　　要：在工程实践和实验中,经常要对经过A/D转换后的数据进行采集,但由于存在外界干扰和电路设计不合理等因素,将会导致数据不停的跳变,非常不稳定。为了改善这种情况,有效地去除干扰,提出一种改进的软件滤波算法。采用STM32作为高速处理器,对16个通道的A/D连续采样1 600次,每个通道采集100次,将采样到的数据通过内部的高速DMA通道传送到内存的数组中。首先对每个通道的100个数据进行冒泡排序,去掉排序后的前后各10个数据,用剩余的80个数据再取平均值。测试结果表明:这种改进的均值滤波算法,能够有效的去除跳变的干扰数据,使得处理后的数据变得稳定、可靠,提高了精度和稳定性。A / D conversion data are often sampled in engineering practice and experiment,but external interference and incorrect circuit design and some other factors will cause data hopping and instable. In order to solve this problem,a new software filtering algorithm was raised to remove interference. Using STM32 microprocessor to scan 16 external A / D channels and continuous sample data 1 600 times,each channel sample 100 times,through DMA these sample data were transferred to memory. One hundred data of each channel can be bubbled sort,10 data of minimal and 10 data of maximal after sort were removed,the left 80 data were averaged. The test results show that this improved algorithm can effectivly get rid of disturbance data,making sample data stable and realiable,thus improving the precision and stability.

关 键 词：软件滤波 冒泡排序 A/D采样 

分 类 号：TN919.6[电子电信—通信与信息系统]