Optimal power flow Essay

to reduce the real power losses to find the optimal values of control variables in the power system, under the condition that the equality condition and inequalities constraints were respected.

The ORPD problem with multi-model optimization is a non-linear, multi-types and difficult function which has been solved using numerical methods.

Recently classical meta-heuristic methods are vastly used in the ORPD problem. Most of them used the evolution process and best interactions between the elements in the group to solve the global optimization searching problems. Among them we have genetic algorithm (GA)[1-3], evolutionary programming (EP) algorithm [4], particle swarm optimization (PSO)[5-6], differential evolution algorithm (DE)[7-8], gravitational search algorithm (GSA) [9], biogeography based algorithm (BBO)[10], and harmony search algorithm HSA[11]. Novel families inspired from the manner of insect and animal in reality such as gray wolf optimization algorithm (GWO) [12], ant lion optimization algorithm (ALO) [13] are used for solving ORPD problem. In [14] a teaching-learning based algorithm (TLBA) has been applied in ORPD problem.

Also, a combination between different meta-heuristic methods has been applied to improve the

quality of solution such as Hybrid shuffled frog leaping algorithm (HSFLA)[15], Nelder-Mead simplex based firefly algorithm [16], PSO with improved pseudo gradient IPGPSO [17] , PSO with the multi-verse algorithm (MVPSO) [18].

These algorithms are classified in global search methods which explores search feasible domain to find the global optimum or approach to the global optimum of a nonlinear, multi-dimensional function with the high quality of the solution. However, most of them are complex and there exist difficulties to choose the appropriate control parameters to achieve the goals of the algorithm.

In the present work, a novel algorithm called the Jaya algorithm is applied for solving the ORPD problem. First, this algorithm was developed and detailed by Rao[19], it explores the best element in the group, thus it is based on the concept that the solution obtained for a given problem should tends to the best solution and should diverges from the worst solution. Moreover, this algorithm uses a simple formulation for adjusting positions with only one random control parameter and selects the best agents in the next generations. It resembles the PSO in adjusting positions and DE in selecting the next generations.

How to cite this essay: