A Low-cost Full-spectrum Water Quality Real-time Monitoring System

This article focuses on a real-time water quality monitoring system based on ultraviolet-visible spectroscopy. The system is equipped with a self-cleaning system to ensure the robustness of long-term use and measurement. The system adopts multi-sensor information fusion technology, which can simultaneously realize multi-parameter measurement of Chemical Oxygen Demand (COD), NO ^3- and turbidity, and can carry out remote data transmission through wireless network to realize intelligent real-time assessment and early warning of water quality. Focus on establishing a full-spectrum inversion model for the three pollution factors of COD, nitrate nitrogen and turbidity, and use partial least squares (PLS) to establish the model. Finally, the linear correlation coefficient between the predicted value of COD and the actual value can reach more than 0.99. The NO ^3- concentration and turbidity can reach above 0.999.