MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641121799 A) filed by Meenakshi Sundararajan Engineering College on October 12, 2026, for Intelligent Predictive Multi-Mode Power Conversion System Using Ai-Assisted Adaptive Control.
Inventors include Sai Shashank S; M. S. Meenakshi; Dr. S. V. Saravanan; Dr. K. S. Saji; Bhasker K; C. Jerin Mahibha; Vinoth Kumar R; C. Durkadevi; Selvarani; and R. Muthukumar.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: The increasing demand for the higlh-efficiency and the smart power conversion systems in renewable energy monitoring systems, EV (Electric Vehicle), mobile electronics and automation in industries has mandated the development of adaptive DC-DC converters which has the capability to maintain the performance optimization under the dynamic operating conditions. These conventional DC-DC converters mainly comprises of fixed control strategies, which often expected to experience reduced output during light-load conditions, by increased switching losses, redundant mode transitions and degrading transient response due to continuous varying in load demands. To address these limitations, this paper proposes an Intelligent Predictive Multi- Mode Power Conversion System with AI-Assisted Control and it is implemented using a synchronous buck converter platform. This system has been proposed for the architecture integration with real-time sensing, predictive control and adaptive optimization and IoT-enabled monitoring within a integrated smart control architecture. The proposed system continuously captures voltage, current and temperature measurements within dedicated sensing unit and the process follows the data preprocessing, load estimation and efficiency determination for the convertor's operating condition. A predictive Efficiency Knee Point (EKP) Engine supported by AI-assisted load forecasting, analyzing the historical and real-time operating data to preempting the future load variations and to identify the optimum operating region before efficiency degradation occurs. Unlike this traditional reactive control techniques, this proposed system architecture introduces a Preparation- Validation Control Framework, in which the prediction operating conditions are prepared in advance and executed only after verifying using real-time sensor measurements and also by the adaptive hysteresis mechanism, further prevents the redundant mode transitions and also improves the converter stability during rapid charging load conditions. A Multi-Objective Optimization Engine which concurrently considers the conversion efficiency, output ripple voltage and thermal performance to determine the optimal operating strategy. Based on these conditions the controller dynamically selects the Pulse Width Modulation (PWM), Dynamic Frequency Pulse Width Modulation (DFPWM), and Pulse Frequency Modulations (PFM) to maximize the efficiency across the varying load conditions. An Adaptive Digital Constant Frequency (DCF) Controller which improves the switching frequency stability, while an AIAssisted Digital Transient Optimized Control (DTOC) enhances the response and the ESP32, PWM drives the synchronous converter. A cloud- based IoT platform enables real-time tracking.
Disclaimer: Curated by HT Syndication.
