Multi-objective based Hybrid Artificial Intelligence Controlled Parallel Inverter in Islanded and Grid Connected Operations

Abstract

Background: The Integration of non-conventional energy systems (NCES), like solar, wind, etc., into the grid with power electronic devices is adapted to meet the demand. Parallel connection of inverters (PCI) is an efficient method to boost power handling capacity, reliability, and system efficiency. However, the main drawback is the unequal power sharing among the inverters while using the conventional droop control technique (CDC). In addition, the circulating currents (CC) flow between these PCI, leading to common mode voltage (CMV), current waveform distortion, and reduction in the system's overall performance. Objectives: This work consists of a photovoltaic system (PVS) and battery energy storage system (BES) as the distributed generation (DG) unit to voltage source inverter (VSI) 1 and 2. The multi- objectives of the suggested work are (a) the equal power/load sharing among two inverters, (b) effective minimization of CC and the CMV, (c) maintaining constant DC-link (DCL) voltage during different solar irradiation and constant temperature, and (d) the reduction in total harmonic distortion (THD) of load current. Methods: A novel approach related to the enhanced droop control (EDC) method with an adaptive neuro-fuzzy hybrid controller (ANFHC) was suggested here to overcome the above issues. The performance analysis of the suggested technique was done in the Matlab/ Simulink platform with islanded and grid-connected modes for different loads. Results: A comparative analysis with the available methods like the proportional integral controller (PIC) and sliding mode controller (SMC) was carried out to exhibit the viability of the developed control technique. Conclusion: This study focused on the operation and control of a PCI with ANFHC in islanded and grid-connected modes to address issues such as uniform power sharing and reduction of CC and CMV.