MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641112615 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And Technology on September 19, 2026, for A Machine Learning-Based Dissolved Oxygen Forecasting System Using Differential Evolution Optimization For Smart Aquaculture.
Inventor includes V Sahiti Yellanki.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: ABSTRACT [0019] The present invention provides a sensor-driven predictive framework that combines Light Gradient Boosting Machine with Differential Evolution optimization to forecast dissolved oxygen levels in intensive aquaculture systems with high precision. Dissolved oxygen is a critical water quality parameter that directly affects the survival, health, and growth of aquatic organisms. Traditional estimation methods struggle to capture the dynamic and nonlinear interactions among environmental variables such as temperature, depth, pH, salinity measured as conductivity, and turbidity. The system is trained on a real-world dataset of 47,293 records that capture these physicochemical parameters together with corresponding dissolved oxygen values. Comprehensive preprocessing that includes imputation of missing values, Z-score-based outlier filtering, and Min-Max feature scaling ensures data integrity. Differential Evolution is applied to automatically tune key LightGBM hyperparameters, including the number of leaves, maximum depth, learning rate, and minimum data per leaf, thereby minimizing prediction error. The optimized model achieves 97.2 percent prediction accuracy, with a root mean square error of 0.493, a mean absolute error of 0.292, and an R-squared value of 97.29 percent, substantially outperforming the baseline LightGBM configuration. The resulting framework supports real-time monitoring through a web interface, enabling proactive water quality management, risk mitigation, and yield optimization in intensive aquaculture operations.
Disclaimer: Curated by HT Syndication.
