摘要
作者为解决NaI(Tl)γ谱仪在测量过程中谱漂移对解谱的影响,提出了一种软件分时动态谱漂移修正方法,并用3组份点源进行了检验.结果表明,修正后用神经网络方法解谱的误差由4.6%减少到3.0%.
In the application of artificial neural networks for the analysis of gammaray spectra measured with a NaI(TI) spectrometer,the spectral drift problem is the main factor to effect analytical precision.The authors have developed a method based on software for correct spectral drift in the period of measurement.The whole measuring time for one spectrum has been divided to a series of time intervals,the spectrums been measured in every time intervals were saved to hard disk for each time interval,thus a series of spectrums were obtained.For every spectrum,correct its referenced peak to the peak of first spectrum,and added the all corrected spectrums to one spectrum,so obtaimed the drift corrected spectrum.A method for determonate peak center position was also present.The approved experiment with three kind of point radiation source shows that the analytical errors were reduced from 4.6% to 3.0% by this method.
出处
《四川大学学报（自然科学版）》
CAS
CSCD
北大核心
2003年第3期502-506,共5页
Journal of Sichuan University(Natural Science Edition)
基金
国家自然科学基金委员会与中国工程物理研究院联合基金(10176015)
关键词
NaI(T1)谱仪
谱漂移修正
谱分析
NaI(Tl) spectrum
spectrum-drifting modifing
spectrum analysis
