MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641099899 A) filed by Anurag University on August 19, 2026, for Ai-Integrated Mathematical Optimization System For Real-Time Decision Support.
Inventors include Dr. Dingari Manohar; Dr. Srinadhuni Hariprasad; Dr. Lavanya Srinathuni; Dr. V. Sumalatha; Mrs. B. Rama; Mrs. Rana Fathima; Mrs. M. N. V. Lalitha Devi; Mr. M Narsimulu; Mrs Andugula Shilpa; and Mr. Dasari Ravi Kumar.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: ABSTRACT OF THE INVENTION The present invention discloses an AI-integrated mathematical optimization system for real-time decision support. The system comprises a real-time data ingestion layer, an artificial intelligence predictive and learning layer, a mathematical optimization engine, a decision orchestration and explainability layer, and a feedback and continuous learning layer, all operating in a closed-loop architecture. Streaming data from heterogeneous sources is continuously cleaned, transformed, and enriched. Artificial intelligence models generate probabilistic forecasts and scenario sets that dynamically update the parameters, constraints, and objective coefficients of mathematical programs. The optimization engine solves these programs under strict latency constraints by employing warm-starting, decomposition techniques (including Benders decomposition and ADMM), surrogate neural- network models with error bounds, and parallel computing. When an exact optimum cannot be obtained within the allotted time, the best feasible solution is returned together with a certified optimality gap. Recommendations are accompanied by sensitivity analysis, multi-objective Pareto information, and natural-language explanations that identify binding constraints and key trade-offs. After decisions are implemented, actual outcomes are observed; performance metrics and concept-drift signals drive automatic or human- supervised updates to both predictive models and optimization formulations, enabling continuous improvement. The invention overcomes the static nature and computational latency of traditional optimization while providing the optimality guarantees, constraint satisfaction, and explainability that pure artificial-intelligence approaches typically lack. It is suitable for deployment in supply-chain operations, energy management, transportation, manufacturing, finance, and healthcare, and may be realized as a cloud-native, on- premise, or hybrid edge-cloud platform. By tightly coupling adaptive artificial intelligence with rigorous mathematical optimization in a real-time closed loop, the system delivers fast, reliable, transparent, and continuously improving decision support under uncertainty.
Disclaimer: Curated by HT Syndication.
