Abstract Water quality involves complex relationships between numerous parameters due to environmental, demographical, and anthropogenic adaptabilities. The presence of computational techniques provides solutions to the limitations of empirical models and derives the precise relation between parameters. The rate of reaeration is an indicator of the health of the water stream that requires accurate measurement. This study employs an adaptive neuro-fuzzy inference system (ANFIS) machine learning technique to predict the reaeration coefficient under varying hydrodynamic and water quality conditions. The five-year dataset prepared on Yamuna River, Delhi, India, to reflect the characteristics of the highly urbanized river, receiving high organic load from multiple point sources. The ANFIS models are developed using a combination of different input parameters. The data is classified using grid partitioning and subtractive clustering methods and trained on the Gaussian membership function type. The performance evaluation of developed ANFIS models is performed through correlation coefficient and root mean square error (RMSE). The results suggested that a robust rule base and flexible approach of ANFIS produces a highly accurate model that could be used to develop strategic schfemes of waste load carrying capacity assessment and water resource management plans.
展开▼
