摘要
目的将傅立叶变换去卷积方法用于峰纯度检查,并在此基础上建立峰纯度检查的概率模型,给出有准确率的判断结果。方法以计算机模拟大量各种重叠色谱峰,对每个重叠色谱峰计算能将其准确判断为重叠峰时的去卷积函数的最小标准偏差,将其记录下来做为观测值。在积累了大量观测值后,计算经验累积分布曲线。实际做出判断时,从该曲线上即可得到判断的准确率。结果成功建立了色谱峰纯度检查的概率模型,并应用于实践,取得较好效果。结论峰纯度检查的概率模型可以提高判断结果的可靠性,适合实际分析工作需要。
AIM： To apply Fourier de-convolution method in the purity test of chromatographic peak and develop a novel probability model for evaluation of the accuracy of purity judgment. METHODS： A large amount of overlapped peaks were simulated in computer. For each overlapped peak, the minimum standard deviation of deconvoluting function, just under which the overlapped peak could be judged correctly, was found. These standard deviations were recorded as observation values. The cumulative distribution curves were calculated when the observation values were enough. RESULTS： The probability model for purity test was built successfully and was applied in our experiment. The result of judgment with accuracy was satisfactory. CONCLUSION： The probability model for purity test can improve the confidence of purity judgment result. It has great applicable value in practical analysis.
出处
《中国临床药理学与治疗学》
CAS
CSCD
2006年第8期947-951,共5页
Chinese Journal of Clinical Pharmacology and Therapeutics
关键词
色谱峰纯度
傅立叶去卷积
累积分布曲线
计算机模拟
chromatographic peak purity
Fourier de-convolution
cumulative distribution curves
computer simulation
