Design of Efficient Storage Unit and EP-ANFIS Controller for On-grid and Off-grid Connected PV-WT System

Abstract

The controllers developed so far for the on-grid and off-grid operation is based on frequency regulation at grid and have yield poor switching by inducing oscillation. Hence to solve this problem in this paper, the switching between the on-grid and off-grid are made by the Emperor penguin based Adaptive fuzzy neuro inference system (EP-ANFIS) controller, which works based on the energy supplied to the load. To serve the transient load condition the hybrid storage unit is modelled by social-ski driver algorithm (SSD) that will schedule the energy supplement between the grid and generator. The proposed controller model provides stable operation during the switching process that is without the transient oscillation the switching is smooth. The error in data selection for the ANFIS is reduced by the Emperor penguin optimization (EPO) as 0.1 for the test data. The proposed work is implemented in Matlab/Simulink platform. The results are compared with the existing works in terms of voltage, current, power, and THD. When examining for transient load condition the settling time for the controller to the steady state is 0.78 s which is comparatively low with PID, PI, ANFIS, and ANFIS-PSO controllers. The THD is reduced to 9.11% from the existing methods by maintain the fundamental frequency of 50 Hz.

Multiple Grid-Connected Microgrids with Distributed Generators Energy Sources Voltage Control in Radial Distribution Network Using ANFIS to Enhance Energy Management

“…For connected PV-WT systems that operate on both ON and OFF the grid, an efficient storage unit, and Emperor Penguinbased Adaptive Neuro-Fuzzy Inference System (EP-ANFIS) controller were designed in [24] to regulate the voltage obtained from RES, the results obtained were presented in Figures 5 and 6 of [24] for off-grid mode and on-grid mode respectively. Considering the proposed technique of the paper [24], during the off mode of their system, the voltage, current, and power profile have some ripples. However, during the on-grid, smooth signals were recorded.…”

“…In article [22], the authors present a Resilience-Oriented Bidirectional ANFIS Framework for connected Microgrid Management systems with Renewable Energy Sources such as Photovoltaic, Wind Turbines, Batteries, and Intelligent Load Control, while electrical networks that are integrated with storage power sources, voltage consistency in the integrated networks with optimal battery charge management controllerbased hybrid GA-ANFS for PV-wind microgrid was presented in this paper [23]. In [24], the authors implemented a switching mechanism using the Emperor penguin based on an adaptive fuzzy neuro inference system (EP-ANFIS) controller to ensure efficient storage unit control in a Microgrid with PV-WT System, and Smart battery controller using ANFIS [25]. Extracting the optimum power from a photovoltaic (PV), wind plant, and battery bank using ANFIS with fractional order PID (FO-PID) controller for a smart DC-microgrid was introduced in [26].…”

“…The 1000A current was supplied and maintained throughout the entire network but the bus voltage and current distributed energy sources (PV, WT, and BESS) are 340VAC and 100A for photovoltaic plants as depicted in figure 20, 340VAC and around 600DCA of the battery energy storage system as in figure21before connected through the Voltage source converter (VSC) controller where the VSC now control these parameters to meet grid network parameters (voltage and current) for proper synchronization while the bus voltage and current of the wind turbine was 340VAC and 3000A respectively as shown in figure22. Using the proposed controller, further evaluation of voltage and current supply to the above-mentioned loads were realized that, for critical loads considered a 340VAC and 1000A were supplied to that load as shown in figure23, and for the noncritical loads, a voltage of a 340VAC and a current around 800A were supplied as shown in figure24. In all these simulation scenarios, the time of 0.1 seconds was used in setting the scope time for clarity of graphs.…”