中文摘要：

采用蒽酮法测定强化生物除磷（EBPR）反应器污泥中糖原物质的含量，运用间隔偏最小二乘法(iPLS)对近红外光谱进行建模，将光谱划分为20个子区间，选出最佳波数区间为4775-4389cm-1，并用这个模型预测未知样品中糖原含量。结果表明：所建模型对糖原校正时的相关系数（rc）为0.9583，校正均方根误差（RMSECV）为0.3554；预测时的相关系数（rp）=0.9769，预测均方根误差（RMSEP）为0.4243。所建的模型对糖原含量的预测效果较好，模型预测精度高，为污泥中糖原的测定提供了一种快速可行的分析技术。

英文摘要：

Glycogen of the sludge from the Enhanced Biological Phosphorus Removal reactor (EBPR) was determined by anthrone method. Applying iPLS to model the near infrared spectra, the spectral range was divided into 20 inte rvals, and wave number range from 4775 t o 4389 cm-1 was selected as the best wavelength range, which was used to predict unknown samples’ glycogen content. Results showed that the correlation coefficient(rc)of this model was 0.9583, and root mean square error of cross v alidation (RMSECV) was 0.3554;the prediction correlation coefficient(rp) was 0.9769, and the root mean square error of prediction(RMSEP)was 0.4243. This model was better for forecasting the content of glycogen, and the accuracy of the model was high, thus providing a feasible analysis technique for the rapid determination of intracellular glycogen in sludge.