Analysis of the power system protection using modern heuristic optimization techniques

Author(s)

Wellington Maycon Santos Bernardes

• Document type: Doctoral Thesis
• Press: São Carlos.
• Institution: Universidade de São Paulo (USP). Escola de Engenharia de São Carlos (EESC/SBD)
• Defense date: 2018-05-18

Examining board members

Eduardo Nobuhiro Asada; Denis Vinicius Coury; Mauro Augusto da Rosa; Silvio Aparecido de Souza; José Carlos de Melo Vieira Júnior

• Advisor: Eduardo Nobuhiro Asada

Abstract

The study of power system protection represents a highly relevant topic providing continuity of service and safety of operation. Today, the coordination of directional overcurrent relays (DORs) is performed using mathematical formulations that basically take into account the operation time of the devices and the coordination time interval (CTI). In this thesis, coordination and selectivity between DORs have been performed considering the simultaneous optimization of the instantaneous and time overcurrent unit (both phase and ground), contingencies in coupled mutually circuits and automatic determination of the curves. Some issues are also discussed such as criteria for short-circuit calculation and topological treatment for interconnected circuits. Initially, the studies were considered as being a case of Monobjective Optimization (weighted sum) by minimizing the sum of operation time of primary relays when occur close-in and line-end faults and also the sum of the instantaneous overcurrent unit. In sequence are proposed two approaches involving multiobjective aspect. The first minimizes the operating time of all protection devices, while maximizing a coordination index (here, CTI is non-fixed). The second, besides minimizing the operating time, the number of settings allowed to alter is limited by operator, if the coordination of all elements involved is not possible in practice. The settings of DORs have been found by using meta-heuristic algorithms (derived from Particle Swarm Optimization and Non-dominated Sorting Genetic Algorithm-II). Modern or intelligent methods, conceived from artificial intelligence, have evolved rapidly and obtained excellent solutions with the acceptable reliability for engineering applications. The test has been carried out on a transmission network from a Brazilian utility. Finally, the results were well satisfactory because using the instantaneous unit and multiple curves substantially reduced the sum of operating time of the protective devices, contributing to decrease workload of protection engineers with safety and rich technical information. In addition, the multiobjective strategies help the operator in the decision making since each solution satisfies specific constraints coming from used equipment or contingency states of the existing network. (AU)

• FAPESP's process: 12/25292-1 - Analysis of the power system protection using modern heuristic optimization techniques
• Grantee: Wellington Maycon Santos Bernardes
• Support Opportunities: Scholarships in Brazil - Doctorate