Novel Control of Wind-PV-Battery based Standalone Supply System with LSTM Controllers

Abstract

Integrated Wind - Photovoltaic (PV) based standalone electric power supply systems are widely used in many areas for various applications. These systems require a battery storage system to ensure a continuous power supply to loads, regardless of fluctuations in loads, wind speed, and solar irradiance. To ensure a reliable and stable power supply to consumers, several changes need to be made in these hybrid systems. Power quality plays a crucial role in standalone power systems, especially in hybrid energy sources based standalone power supply systems. The battery needs to charge when there is surplus power generation and discharge when there is a demand from the loads. To achieve this, a bidirectional DC to DC converter is used to connect the battery to the network, with a proper controlling mechanism. Additionally, maximum power point tracking devices with appropriate algorithms are incorporated for PV and wind turbines to optimize their utilization in all weather conditions. This paper considers multiple PV systems and wind turbines, each with proper arrangements of series and parallel combinations of PV modules, to determine the appropriate rating for the power supply system. Long short term memory (LSTM) based artificial neural network (ANN) controllers are implemented for various control units in this hybrid standalone power system. These proposed control techniques significantly improve power quality under various situations. The performance of the proposed method is evaluated using MATLAB/Simulink, and the results are presented in this paper.