Comparative Analysis of optimization methods applied to the solution of the reconfiguration of electrical distribution systems

• Grant number: 18/18659-2
• Support type: Scholarships in Brazil - Master
• Effective date (Start): March 01, 2019
• Effective date (End): July 31, 2020
• Field of knowledge: Engineering - Electrical Engineering - Power Systems
• Principal researcher: John Fredy Franco Baquero
• Grantee: Christoffer Lucas Bezão Silveira
• Home Institution: Faculdade de Engenharia (FEIS). Universidade Estadual Paulista (UNESP). Campus de Ilha Solteira. Ilha Solteira , SP, Brazil

Abstract

The reconfiguration of distribution systems has been extensively studied in recent years, with the main objective the reduction of power losses. Different methods have been developed to solve the reconfiguration problem, which can be classified as heuristics, metaheuristics (e.g., genetic algorithms and tabu search), and mathematical optimization (e.g., linear programming and nonlinear programming). Although many works reported the computational effort required for each method, it is not always possible to make a direct comparison among different approaches due to the different conditions in which they were developed/tested (e.g., test systems, programming language, and computational equipment). This research project will carry out a comparative analysis among different heuristic, metaheuristic, and mathematical optimization methods applied to the solution of the reconfiguration of electrical distribution systems. That analysis will allow the classification of the methods' performance according to the value of the objective function and the computational effort. In particular, the performance of nonlinear, linear, and conic mathematical formulations for the reconfiguration problem will be evaluated through test and real distribution networks typically used in the specialized literature. Furthermore, from the study of those different mathematical approaches, the application of surrogate constraints, which colud accelerate the solution process, will be evaluated. The mathematical formulations will be implemented using the modeling language AMPL and solved using commercial solvers. (AU)