MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641100653 A) filed by Gandham Yedukondalu; and Koneru Lakshmaiah Education Foundation on August 20, 2026, for A Machine Learning-Based System And Method For Predictive Optimization Of Wire Electrical Discharge Machining Of Aa6061–tungsten Carbide Metal Matrix Composites.
Inventors include Manojkumar Subrao Kate; Priyaranjan Samal; Kotthapalli Karthik; and G Yedukondalu.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The present invention relates to a machine learning-based system and method for predictive optimization of Wire Electrical Discharge Machining (WEDM) of AA6061–Tungsten Carbide metal matrix composites. The system comprises an Experimental Database, Data Acquisition Module, Data Preprocessing Module, Feature Engineering Module, Machine Learning Prediction Engine, Multi-Objective Optimization Engine, Decision Support Module, Continuous Learning Module, Knowledge Database, and Human–Machine Interface. The system acquires WEDM process parameters and corresponding machining responses, preprocesses the acquired data, identifies significant process variables, and employs a machine learning prediction model to predict machining responses for specified WEDM process conditions. The predicted machining responses are supplied to the Multi-Objective Optimization Engine to determine optimum or near-optimum WEDM process parameter combinations based on multiple machining objectives and process constraints. A Decision Support Module evaluates the optimized solutions and selects a preferred machining condition according to user-defined priorities. The selected machining condition is experimentally implemented on the AA6061–Tungsten Carbide metal matrix composite, and the experimentally measured machining responses are compared with corresponding predicted responses to determine prediction performance and prediction error. The validated experimental data are stored in the Knowledge Database and utilized by the Continuous Learning Module to retrain or fine-tune the machine learning prediction model for subsequent prediction and optimization cycles. The invention thereby provides an integrated data-driven framework for predictive machining, multi-objective parameter optimization, experimental validation, and continuous improvement of WEDM process recommendations for AA6061– Tungsten Carbide metal matrix composites.
Disclaimer: Curated by HT Syndication.
