Novel Photovoltaic Modules and Wind Turbine Performance Modeling Techniques Based on Artificial Intelligence and Deep-Learning

This work introduces novel modeling techniques for analyzing the quality of Vertical-Axis Wind Turbines (VAWT), photovoltaic (PV) panels, and Horizontal-Axis Wind Turbines (HAWT). Derived from real data sets from various production manuals, three methodologies are used to analyze and simulate the created models. The first strategy is depend on Artificial Neural Network (ANN), the second on curve fitting, and the third on a Convolution Neural Network (CNN). Based solely on-demand power, the created systems can estimate the running efficiency characteristics of VAWT, PV, and HAWT. As a result, it may readily support the developer in selecting an appropriate unit before the setup process. For a power range of 5.0W to 355W per panel module, the PV system will estimate the open circuit voltage, short circuit current, current and voltage at max output, module performance, and system price. Furthermore, wind turbine algorithms can forecast the rated speed, cut-in wind speed, speed of rotor, diameter of the rotor, turbine cost, and hub-height ranges ranging from 0.11kW to 100kW for VAWT and 0.50kW to 8000kW for HAWT. The findings reveal that the CNN technique outperforms the ANN and curve fitting methods in terms of data matching. Matlab/Simulink is used to perform the models.