知识与财富的链接
针对传统化学需氧量（chemical oxygen demand, COD）检测存在检测成本高、耗时、易造成二次污染,以及现有检测模型泛化性较差等不足,难以满足水环境实时监测需求的问题,本研究提出基于近红外光谱技术的COD快速无损定量预测模型。实验结果表明,本模型在污水COD光谱数据集上的预测性能,相较于传统机器学习算法,以及现有其他深度学习算法更优。测试的决定系数（R2）和均方根误差（RMSE）分别达到0.992 1和27.47 mg·L-1,模型卷积层的输出特征可解释强,能有效表征关键波长点。为实际水体COD含量快速检测提供一种新的方法。
The traditional chemical oxygen demand (COD) detection method is known for its high cost, time-consuming process, and potential for secondary pollution. Moreover, existing detection models often lack generalization, making it difficult to meet the demands of real-time water environment monitoring. In this study, we propose a rapid and non-destructive quantitative prediction model for COD based on near-infrared spectroscopy. Experimental results show that the prediction performance of this model on the sewage COD spectrum dataset is better than that of traditional machine learning algorithms and other existing deep learning algorithms. The model achieved a high coefficient of determination of 0.992 1 and a low root mean square error of 27.47 mg·L-1. The output features of the model''s convolutional layer are interpretable and effectively capture the key wavelength points. This research provides a new method for the rapid detection of COD in practical water samples.