Load flow solution for islanded microgrids using conventional and non-conventional methods

Abstract

Integration of distribution generators and uncertainties of environmental make the solution of power flow problem in islanded microgrid complex and non-trial. Several methods have been proposed to overcome these problems using conventional methods. However, these schemes are complicated and not easy to implement due to the complexity of equations. In this project, simple algorithms are proposed for load flow solutions in islanded MGs using meta-heuristic methods. Particle swarm optimization (PSO) and differential evolution (DE) are used to evaluate the frequency, Distributed generators active and reactive powers, voltage magnitudes and angles using droop control concepts with cascaded optimizations. For more simplification and to minimize the simulation time, a single optimization algorithm is proposed using the frequency as a variable in the fitness function. the proposed methods are applied to 7-bus and 38-bus test systems. The results are compared with PSIM and PSCAD which validate the effectiveness of the developed algorithms. The mismatches found by the evolutionary techniques are much lower than those found in the conventional solutions especially for a single optimization. The initial values selection’s complexity is alleviated in PSO and DE algorithms by selecting the upper and lower limits of optimization parameters. herefore, the conventional methods in the load flow problem are initial guess dependent, time- consuming and more complex whereas the meta-heuristic techniques are simple to implement, have a few parameters to adjust and produce better results than the Newton Raphson method.