A Novel Maxwell–EWMA Control Chart Integrated With Support Vector Regression For Biological Water Quality Monitoring

Abstract

Rapid industrialization in coastal regions has considerably enhanced the discharge of wastewater enduring several organic and inorganic pollutants into marine ecosystems. Such discharges can adversely distress seawater quality and generate serious environmental and social consequences. Therefore, drastic evaluation and monitoring of marine water quality are essential, particularly through oxygen-related indicators such as Biological Oxygen Demand (BOD), Chemical Oxygen Demand (COD), and Dissolved Oxygen (DO), which offer imperative indication about the level of organic pollution and the global health of aquatic systems. In this study, a hybrid  statistical monitoring approach is proposed for detecting changes in marine water quality with principal component analysis. The resulting principal component scores are subsequently incorporated into a  neutrosophic based Exponentially Weighted Moving Average  with Maxwell distributed quality characteristics (NVEWMA) for monitoring framework to improve the sensitivity of the control chart to variations in the underlying water quality process. The integrated PCA–NVEWMA framework is applied to a real-world marine water quality dataset to identifying abnormal variations. The findings demonstrate that the integration of PCA-based feature extraction with the NVEWMA control chart can serve as a valuable approach for continuous marine environmental monitoring and the early detection of undesirable changes in seawater quality.