一种色谱谱峰识别算法的实现

摘要

色谱在化学分析中的应用越来越广泛，针对色谱信号数据处理要求，实现一套谱峰识别的算法。在数据预处理和提出的程序法确定合适阈值的基础上，该算法结合谱峰曲线及其一阶导数曲线的趋势对谱峰进行识别，算法经宽松变换后，用于区分化学谱图中谱峰的终点是否真正落回到基线上，并由此实现一种具有健壮性的重叠峰检测方法和及时准确的谱图基线校正方法。由于该算法的阈值参数可以固定，从而保证了重叠峰检测较少地受到人为的影响，进一步提高了化学谱图数据处理的自动化水平。%The application of chromatography is more and more popular in chemical analysis .In this article , we realise a set of algorithms for chromatographic peak recognition in light of the requirement on chromatogram signal processing .Based on data pre-processing and appropriate threshold determined with procedural method , the algorithms distinguish the peaks by combining the trend of chromatogram curve with the trend of its first derivative curve , after loose transformation , the algorithms are applied to differentiate whether the end points of peaks in chemical chromatogram have really fallen back to the baseline , and thus implement an overlapping peaks detection method with robustness and the timely and accurate method of chromatogram baseline correction .As the threshold parameter in the algorithm can be fixed , therefore it ensures the overlapping peaks detection less impacted by the human , and further improves its automation level in chemical chromatogram data processing .