LMSRE-Based Adaptive PI Controller for Enhancing the Performance of an Autonomous Operation of Microgrids

Ahmed M Hussien,Ahmed Al-Durra,Hany M Hasanien,Rania A Turky

doi:10.1109/access.2021.3091496

Abstract

Microgrids take a large part in power networks thanks to their operational and economic benefits. This research introduces a novel implementation of an adaptive proportional plus integral (PI) controller to boost the autonomous microgrid operation efficiency. The least mean and square roots of the exponential algorithm are utilized in the adaptive PI control strategy. The multi-objective function for both sunflower optimization (SFO) and particle swarm optimization (PSO) algorithms is obtained by The Response Surface Methodology. The system is evaluated under different environments, which are stated as follows: 1) disconnect the system from the grid (islanding), 2) autonomous system exposure to load variability, and 3) autonomous system exposure to a symmetrical fault. The proposed practicality of the control plan is shown by the data of the simulation, which is extracted from PSCAD/EMTDC software. The strength of the suggested adaptive control is confirmed through matching its results with those obtained using the SFO and PSO based optimal PI controllers.

Highlights

This paper suggests a new adaptive technique titled as the least mean (LM) and square root of exponential (SRE) technique
It is important to remember that the voltage and frequency are maintained by the grid
This research focuses on enhancing the MG during the autonomous operation mode using the cascaded control scheme, which is discussed

Summary

INTRODUCTION

The MG contains multiple DGs and different loads It can work either grid linked or in an autonomous operation mode [3]. Many metaheuristic algorithms have been improved in order to overcome those difficulties, such as particle swarm optimization (PSO) [8], sunflower optimization algorithm (SFO) [9,10], hybrid GWO-PSO optimization technique [11], genetic algorithm (GA) [12], hybrid firefly and particle swarm optimization technique [13], Harris hawks optimization Method [14], marine predators algorithm [15], hierarchical model predictive control [16], Tabu search [17], quasi‐oppositional selfish herd optimization (QSHO) [18], Cuttlefish optimization algorithm (CFA) [19], and teaching-learning based optimization [20] Each of those techniques has its benefits and drawbacks [21]. Research gap and motivation The optimization approaches have many limitations, such as dynamic systems, high storage demands, and high microprocessor capacities, etc These drawbacks mean that new control procedures, including adaptation techniques, are needed. LMSRE has constants (μ and α) that combine rapid convergence and stability with minimal error

Contribution and paper organization The major contributions of this paper are

System Modelling

The RSM