Quantitative Detection of Copper Ions in Water via Feature-Level Fusion of UV-Vis Absorption and Fluorescence Spectra with Optimized XGBoost

In response to the increasingly severe issue of heavy metal pollution in water, this paper proposes a method for the robust quantitative analysis of copper ions in purified water and real water samples based on the feature-level fusion of ultraviolet-visible absorption (UV-Vis) spectra and fluorescence spectra, combined with the Extreme Gradient Boosting (XGBoost) algorithm. Specifically, this study introduces a feature-level fusion strategy to overcome the limitations of single-spectrum detection, while the optimized XGBoost algorithm is employed to model the complex non-linear relationships that are difficult to capture using traditional linear regression methods. An optimization algorithm is introduced to fine-tune the model’s hyperparameters, thereby enhancing its predictive performance. Using the coefficient of determination (R2) and root mean square error (RMSE) as evaluation metrics, rapid and accurate detection of copper ions in water is achieved. Experimental results show that, for standard solutions, the optimized XGBoost model achieves a coefficient of determination of 0.9915 and a root mean square error of 2.6663 mg/L; for actual water samples, the optimized XGBoost model achieved a coefficient of determination of 0.9892 and RMSE of 1.2738 mg/L. This demonstrates the model’s strong generalization ability in overcoming the physical limitations of optical probes. This method effectively identifies and quantifies copper ions in water samples, demonstrating good accuracy and stability.