Abstract: The potassium dichromate reflux method, as specified in the water quality monitoring standard (HJ 828-2017), serves as a conventional approach for chemical oxygen demand (COD) determination. However, its application in fundamental chemistry laboratory education has been constrained due to inherent safety hazards, high operational costs, and significant pollutant discharge. This study presents an innovative approach employing a color-sensitive camera to capture solution reaction images, with subsequent RGB data extraction through Python’s OpenCV library. By integrating machine learning-based cluster analysis, we achieved automated monitoring of both the heating reflux and titration processes. The incorporation of digital twin visualization technology into the potassium dichromate reflux method for COD measurement represents a novel advancement, with interactive operations significantly enhancing the effectiveness of simulated experimental instruction.

Key words: Digital twin, Potassium dichromate reflux method, Chemical oxygen demand, Machine learning, Automatic monitoring