AI BASED APPROACH FOR POWER PREDICTION AND PERFORMANCE EVALUATION OF THE SOLAR PV MODULES WITH VARYING DUST ACCUMULATION LEVELS

Abstract

Solar energy has shown to be the undisputed leader among renewable energy sources since it is clean and environmentally benign. A photovoltaic module's output power and longevity are controlled by a number of parameters, including solar insolation, clouds, cell temperature and other shading effects such as dust deposition, meteorological conditions, geographical location, module orientation, and so on. Energy demand and concerns over greenhouse gases have made the integration of solar PV into the grid imperative. Solar power forecasting models must have a high prediction accuracy to address the intermittent nature of solar irradiation. Solar PV power is significantly affected by the dust deposited on the PV panel surface. The impact of dust collection on the operation of the 5 kW photovoltaic system set up on the rooftop of the UEE laboratory at Delhi Technological University is initially investigated in this work. The performance of the 5 kW photovoltaic system is evaluated for 62 days, with the panels left naturally uncleaned for the first 31 days and then cleaned on a regular basis for the next 31 days. The performance ratio, capacity factor, system energy yield, and reference energy yield are all calculated. The pragmatic 5 kW system's performance analysis results were afterwards compared to the PVsyst software results. Later on, the amount of dust deposited on PV panel as one of the input parameters to predict solar PV power and solar irradiation is also studied. Multivariate analysis of three deep learning techniques that is LSTM (Long short-term memory), 1D CNN (Convolution Neural Network) and BilSTM (Bidirectional Long short-term memory) to predict the solar PV power and solar irradiation with varying dust accumulated levels for the 335-watt PV module set up on the rooftop of the lab at Delhi Technological University is presented. An artificial dust scenario is created by continually incrementing the dust level by 1.258 mg/cm2 .