MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641119676 A) filed by Bannari Amman Institute Of Technology on October 06, 2026, for An Intelligent Photovoltaic Cleaning Management System Using Sparse Sensing And Predictive Control.
Inventors include Rithanya R B; Sivasankari K N; Lakshitha S; Sabariananthan P; and Sannasi Chakravarthy S R.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: AN INTELLIGENT PHOTOVOLTAIC CLEANING MANAGEMENT SYSTEM USING SPARSE SENSING AND PREDICTIVE CONTROL The present invention discloses an intelligent photovoltaic cleaning management system and method using sparse multi-source sensing from drone-based sensing unit, ground sensing units and Supervisory Control and Data Acquisition (SCADA) interface for estimation of soiling condition of a photovoltaic plant. The collected data are fused using a weighted rule to produce a zone-wise soiling index (101, 203). Test-patch cleaning is applied on selected reference panels and electrical comparison before and after cleaning is applied to establish a self-calibrating soiling-to-energy model (102). The system can thus predict the expected energy gain from cleaning any zone before any action is dispatched. An adaptive sparse-sensing soiling mapper (204) reconstructs a full soiling map, produces an uncertainty value per zone and triggers additional sensing only where needed (106). For each zone, a priority score is calculated (109) and tested (110) to determine whether cleaning is justified considering predicted energy gain, cleaning cost, soiling type, and reconstruction confidence. High-priority zones are assigned for drone cleaning, robot contact cleaning, or a mixed mode (208). The cleaned SCADA data are used to update model parameters in a closed feedback loop (114). The invention minimizes unnecessary sensing and cleaning and maximizes energy recovery.
Disclaimer: Curated by HT Syndication.
