Frequency Control in PV-Integrated Microgrid Using Hybrid Optimization-Based ANFIS and Deep Learning Network

Frequency variation in the network and severe impact of linked frequency-sensitive loads are the main concerns of integrating RESs with traditional power system networks. Because RESs are separated from the traditional grid by power electronic converters, they have zero or very little inertia, which is the primary source of frequency variation. If different electrical motors are connected with PV, the rotor speed frequency and the pulse of the PV panel are assorted. Therefore, this paper suggests an adaptive Neuro-fuzzy inference system (ANFIS) and Deep Neural Network (DNN) based controller for improving the performance of the power system. The developed approach's primary goal is to regulate the output waveform, hence reducing the error among the control and reference signals. Moreover, the performance model is further enhanced by optimizing the ANFIS controller using the novel hybrid optimization algorithm, named Honey Badger-based grey wolf Optimization (HB-GWO) Algorithm. The performance of the implemented scheme is done in MATLAB and the results over various controllers concerning switching frequency and time analysis. Accordingly, the optimal parameter values obtained using the HB-GWO Algorithm are, the initial step size is 0.31258, the decrease rate of step size is 0.47177 and the increase rate of step size is 1.3754.