基于滤波宽度动态调整的递推中值与外延算法  

Recursive Median and Extension Algorithm Based on Filter Width Dynamic Adjustment

作　　者：高宇 回朝妍 张辉 贾梦涵 马仡男 佟美琪 GAO Yu;HUI Zhao-yan;ZHANG Hui;JIA Meng-han;MA Yi-nan;TONG Mei-qi(School of Energy and Environmental Engineering,University of Science and Technology Beijing,Beijing 100083,China)

机构地区：[1]北京科技大学能源与环境工程学院,北京100083

出　　处：《仪表技术与传感器》2022年第1期104-108,120,共6页Instrument Technique and Sensor

摘　　要：为了抑制噪声对数据的干扰,提高数据采集准确度,提出一种能够动态调整滤波宽度的递推中值算法,对原始数据进行预处理。将递推算法与中值算法相结合,使曲线变化更加光滑,并且剔除了大的脉冲干扰;对递推队列长度进行动态调整,保留了原始数据的特征信息;利用最小二乘法对算法进行优化,使数据在保持基底特征情况下外延,减弱滤波宽度造成的信息丢失。利用该算法对实际检测的数据进行处理,结果表明:该算法能够取得较好的滤波效果,并且保留原始数据的重要特征。In order to suppress the influence of noise on data and improve the accuracy of data acquisition,a recursive median algorithm which can dynamically adjust the filtering width was proposed to preprocess the original data as digital filtering.The recursive algorithm was combined with the median algorithm to make the curve change smoother and eliminate the large pulse interference.The length of recursive queue was dynamically adjusted to retain the characteristic information of the original data.In order to extend data according to the characteristics of the base,the least square method was used to optimize the algorithm so that the information loss caused by the filtering width can be reduced.The algorithm was used to process the actual detected data.The results show that the algorithm can achieve better filtering effect and retain the important characteristics of the original data.

关 键 词：曲线滤波 递推中值 动态调整 队列长度 基底外延 最小二乘法 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]