MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641111933 A) filed by National Institute Of Technology-Warangal on September 17, 2026, for Cnn-Qlstm-Attention Based System For Smart Grid Electric Load Forecasting.
Inventors include Anil Kumar Maddali; and Chintham Venkaiah.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: TITLE: CNN-QLSTM-ATTENTION BASED SYSTEM FOR SMART GRID ELECTRIC LOAD FORECASTING The present invention provides an artificial- intelligence-based electric load forecasting system for a smart grid. Historical electric load data received from a smart meter network or grid database are cleaned, normalized and segmented into fixed-length sequential input windows. One-dimensional convolutional layers extract localized temporal feature representations, which are supplied to a Quantum Long Short-Term Memory network having recurrent gating operations incorporating variational quantum circuit transformations for nonlinear temporal dependency modelling. An attention mechanism determines attention scores for temporal hidden-state representations and generates a weighted temporal representation. A forecast generation layer generates future electric load values over a forecast horizon, and an energy-management interface communicates the forecast values to a smart-grid energy-management platform or demand-response platform. The architecture supports multistep short-term load forecasting while combining local temporal feature extraction, quantum-inspired sequential modelling and adaptive temporal weighting in a connected smart-grid forecasting framework. Figure 1 illustrates the CNN-QLSTM-Attention hybrid architecture.
Disclaimer: Curated by HT Syndication.
