Implementation of the VNS-DEEPSO algorithm for the energy dispatch in smart distributed grid

• Main Authors: Pedro Julián García Guarín, Julián Cantor López, Camilo Cortés Guerrero, María Alejandra Guzmán Pardo, Sergio Rivera
• Format: Article
• Language: English
• Published: Universidad de la Costa 2019-06-01
• Series: Inge-Cuc
• Subjects: heuristic algorithms; renewable energy; optimization; intelligent network; electric vehicles
• Online Access: https://revistascientificas.cuc.edu.co/ingecuc/article/view/1984

Introduction− Traditional electric networks are mi-grating to new configurations of intelligent networks, which bring operational and planning challenges. In order to advance in these challenges, an optimization problem is proposed to solve in the programming of intelligent network elements.Objective− The optimization problem consists of man-aging the energy dispatch of an intelligent network to optimize the available resources, considering the uncer-tainty of renewable energies, planned trips of electric vehicles, cargo forecast and market prices.Methodology− It was proposed to use an assembly between two heuristic methods. The VNS algorithm (Variable Neighborhood Search) and the DEEPSO (Dif-ferential Evolutionary Particle Swarm).Results− The value obtained by the VNS-DEEPSO algorithm was 18.21, being 7 % better than the second algorithm classified in the competition. Conclusions− The VNS-DEEPSO algorithm was the winner among 9 metaheuristic algorithms that solved the problem. This problem has a greater increase in difficulty due to the uncertainty generated by weather conditions, load forecast, planned EV ́s trips, and mar-ket prices. According to the results, the VNS-DEEPSO algorithm proved to be the most efficient in minimizing operational costs and maximizing the revenues of the intelligent network.