医疗系统中的计算机时间稳定性分析  

Analysis of Computer Long-term Clock Error in Health Care System

作　　者：孙朋[1] 王洋 

机构地区：[1]中国人民解放军第三O九医院政治部,北京100091 [2]河南省军区自动化站,河南郑州450000

出　　处：《中国医疗设备》2014年第4期21-23,50,共4页China Medical Devices

基　　金：国家自然科学基金(41174025)

摘　　要：以智能医疗和远程医疗为代表的现代医疗系统离不开精密的时间计量,计算机内部时间可以方便地为其提供时间基准,故需对其长期钟差进行分析。本文首先介绍了时间基准及其钟差的基本原理,分析了实验获取的90天钟差数据;然后利用二次多项式滤除钟差中的长期趋势项,再利用二阶正弦和函数对钟差的周期项进行了拟合分析;最后分别利用快速傅立叶变换(FFT)和功率谱法对钟差的周期项进行频谱分析。分析结果均表明,长期钟差中含有两个长度约为52天和28天的周期,根据长期趋势项和周期项对其长期钟差进行建模并改正,使计算机时间可更好地应用于医疗系统。Modern health care system represented by intelligent medical and telemedicine treatment can not be separated from the precise time measurement. The computer time can easily provide a time basis, so the long-term clock error of it needs to be analyzed. In this paper, the basic principles of time basis and clock error were introduced, and a 90-day clock error data is brielfy analyzed. The quadratic polynomial was used to iflter the long-term trend of clock error, and the second-order sum of sine functions was used to ift the periodic item of clock error. Finally, the fast Fourier transform （FFT） and power spectral method were used to analyze the clock error respectively. Analysis results indicate that long-term clock difference contains two cycles, 52 days and 28 days, and when corrected according to the long-term and cycle trend, computer time can be better applied in the health care system.

关 键 词：计算机时间 钟差 医疗系统 快速傅立叶变换 功率谱 

分 类 号：TP311.53[自动化与计算机技术—计算机软件与理论]