论文部分内容阅读
应用沉积物吸附双酚A(BPA)BP神经网络模型,模拟了松花江表层沉积物的不同泥水比、非残渣态组分(有机质、铁氧化物、锰氧化物)和BPA初始浓度对BPA吸附量的影响。所建BP神经网络模型相关系数R2为0.9665,校正集均方差(MSEc)、验证集均方差(MSEv)和预测集均方差(MSEp)分别为0.0068、0.0596和0.1285;利用遗传算法优化估算了基于BP神经网络模型的沉积物吸附BPA的最大吸附量,优化值与实验值的相对偏差为0.96%~8.21%。此外,利用BP神经网络模型预测了沉积物非残渣态组分(有机质、铁氧化物、锰氧化物)质量百分比及摩尔含量变化与BPA吸附量的关系,经分析可知,铁氧化物和有机质对沉积物吸附BPA起着促进作用,沉积物非残渣态组分吸附BPA的相对贡献(K)为KFe>KOMs>KMn,即沉积物中铁氧化物是BPA的主要吸附位,而Mn氧化物则对沉积物吸附BPA起着抑制作用。
BPA neural network model of sediment-adsorbed BPA was used to simulate the adsorption of BPA on the surface sediment of Songhua River under different slime-water ratios, non-residual components (organic matter, iron oxide, manganese oxide) and BPA initial concentration The impact of volume. The correlation coefficient R2 of the built BP neural network model was 0.9665, the mean square error of calibration (MSEc), mean square error of validation (MSEv) and mean square error of prediction (MSEp) were 0.0068,0.0596 and 0.1285 respectively. The maximum adsorption capacity of BPA on the sediment of BP neural network model was 0.96% -8.21%. In addition, BP neural network model was used to predict the relationship between the change of mass percentage and molar content of non-residue components (organic matter, iron oxide, manganese oxide) and the amount of BPA adsorbed on the sediments. It can be seen from the analysis that iron oxide and organic matter pairs The adsorption of BPA on the sediments promoted the adsorption of BPA on the non-residual components of the sediment (K) as KFe> KOMs> KMn, that is, the iron oxides in the sediments were the main adsorption sites of BPA while the Mn oxides Sediment adsorption of BPA plays a suppressive role.