论文部分内容阅读
研究了不同PCDD（全名Polychlorinateddioxin）同系物分子结构的表达及特征参数的选择，应用神经元网络方法对其分子结构与色谱保留值进行了关联。对49种PCDD同系物在DWS往上不同温度下保留时间进行了预测，结果95％以上的数据点相对误差小于10％，而80％以上的数据点相对误差小于5％。
The molecular structure of different PCDD (Polychlorinateddioxin) homologues was studied by selecting the molecular structure and selecting the characteristic parameters. The relationship between the molecular structure and the chromatographic retention value was analyzed by using the neural network method. The retention times of 49 PCDD homologues at different temperatures above DWS were predicted. The relative error of 95% or more data points was less than 10%, while the relative error of 80% or more data points was less than 5%.