Solving optimal reactive power problem by improved variable mesh optimization algorithm

Received Apr 29, 2019 Revised Aug 12, 2019 Accepted Sep 9, 2019 In this work Improved Variable Mesh Optimization Algorithm (IVM) has been applied to solve the optimal reactive power problem. Projected Improved VMO algorithm has been modeled by hybridization of Variable mesh optimization algorithm with Clearing-Based Niche Formation Technique, Differential Evolution (DE) algorithm. Mesh formation and exploration has been enhanced by the hybridization. Amongst of niche development process, clearing is a renowned method in which general denominator is the formation of steady subpopulations (niches) at all local optima (peaks) in the exploration space. In Differential Evolution (DE) population is formed by common sampling within the stipulated smallest amount and maximum bounds. Subsequently DE travel into the iteration process where the progressions like, mutation, crossover, and selection, are followed. Proposed Improved Variable Mesh Optimization Algorithm (IVM) has been tested in standard IEEE 14,300 bus test system and simulation results show the projected algorithm reduced the real power loss extensively.


INTRODUCTION
Reactive power problem plays a key role in secure and economic operations of power system. Optimal reactive power problem has been solved by variety of types of methods [1][2][3][4][5][6]. Nevertheless, numerous scientific difficulties are found while solving problem due to an assortment of constraints. Evolutionary techniques [7][8][9][10][11][12][13][14] are applied to solve the reactive power problem, but the main problem is many algorithms get stuck in local optimal solution & failed to balance the Exploration & Exploitation during the search of global solution. In this work Improved Variable Mesh Optimization Algorithm (IVM) has been applied to solve the optimal reactive power problem. Projected Improved VMO algorithm has been modeled by hybridization of Variable mesh optimization algorithm with Clearing-Based Niche Formation Technique, Differential Evolution (DE) algorithm. Mesh formation and exploration has been enhanced by the hybridization. Amongst of niche development process, clearing is a renowned method in which general denominator is the formation of steady subpopulations (niches) at all local optima (peaks) in the exploration space. Each niche has a leading (master) individual, i.e. the one with the most excellent fitness. In Differential Evolution (DE) population is formed by common sampling within the stipulated smallest amount and maximum bounds. Subsequent to the launch of generating the population, DE travel into the iteration process where the progressions like, mutation, crossover, and selection, are followed. Proposed Improved Variable Mesh Optimization Algorithm (IVM) has been tested in standard IEEE 14,300 bus test system and simulation results show the projected algorithm reduced the real power loss extensively.

PROBLEM FORMULATION
Objective of the problem is to reduce the true power loss Voltage deviation given as follows Voltage deviation given by Constraint (Equality) P = P + P (4)

VARIABLE MESH OPTIMIZATION
Variable mesh optimization algorithm (VMO) engendered population is scattered as a mesh. Mesh is poised of Z nodes ( , , . . , ) that symbolize the solutions in the search space [15]. Every node is oblique as a vector of M floating point numbers = , , . . , , . . , which designate the solution. In exploration procedure two methodologies called development and narrowing are utilized. During the development, new-fangled nodes are created in the direction of local maximum, comprehensive end and the boundary nodes. Grounded on an elite approach, nodes are prearranged bequeath to their superiority in ascending order. Then clear out adaptive operator is then applied; every node is evaluated to its successor to eradicate those that do not outdo the threshold. Threshold value is computed by When end criterion is met, process will be stopped End

CLEARING-BASED NICHE FORMATION TECHNIQUE
Amongst of niche development process, clearing is a renowned method in which general denominator is the formation of steady subpopulations (niches) at all local optima (peaks) in the exploration space. Each niche has a leading (master) individual, i.e. the one with the most excellent fitness [16]. To a certain niche an individual fit in when its distance to the leading (master) individual is less than a given threshold called as clearing radius. This process share the possessions of a niche among a set of winners (individuals to be profited by clearing), whereas it sets to zero then the fitness of all erstwhile individuals will be in the same niche. Those restrained by the winner are deceitfully separated from the population. Subsequently reiterate this method for a definite number of iterations, then all winners will come into view.

DIFFERENTIAL EVOLUTION
In Differential Evolution (DE) population is formed by common sampling within the stipulated smallest amount and maximum bounds [17]. Subsequent to the launch of generating the population, DE travel into the iteration process where the progressions like, mutation, crossover, and selection are followed. "DE/best/1" "DE/current-to-best/1" "DE/rand/1" "DE/current-to-rand/1" DE/rand/2" Improved strategy of the binomial crossover described as follows Then IEEE 300 bus system [19] is used as test system to validate the performance of the Improved Variable Mesh Optimization Algorithm (IVM). Table 2 shows the comparison of real power loss obtained after optimization.

CONCLUSION
In this work Improved Variable Mesh Optimization Algorithm (IVM) has been successfully solved the optimal reactive power problem. Mesh formation and exploration has been enhanced by the hybridization. Amongst of niche development process, clearing is a renowned method in which general denominator is the formation of steady subpopulations (niches) at all local optima (peaks) in the exploration space. In Differential Evolution (DE) population is formed by common sampling within the stipulated smallest amount and maximum bounds. Proposed Improved Variable Mesh Optimization Algorithm (IVM) has been tested in standard IEEE 14,300 bus test system and simulation results show the projected algorithm reduced the real power loss extensively.