AI-Driven Smart Cleaning Optimization Strategy for Dust-Accumulated PV Module

Dust accumulation on solar photovoltaic modules is a major issue in dusty arid environments where particles are volatile and settle on the surface of the panels thus reducing solar irradiation. This soiling effect can lead to considerable power loss in peak dry months, resulting in poor system performance and overall energy yield. In this paper an intelligent data driven methodology is presented to ascertain for a 5 kW photovoltaic system operating in dry environments the optimal cleaning interval. The soiling damage is quantified by employing an exponential degradation model supported through data of energy production and soiling loss in real time. Thereafter a the Grey Wolf Optimiser algorithm is used to ascertain the optimal cleaning regime in respect of minimum cost. The study showed that a point could be reached where the cost of cleaning would be balanced against energy recovered by cleaning, this point being at a frequency of approximately 64 days interval. The model is scalable and adaptive leading to a methodology presented which optimises solar photovolatic cleaning regimes and increases energy yield in dusty environments.