Estratégia de simulação de transitórios eletromecânicos com emprego de unidades de simulação funcional e OpenDSS

Abstract

This master thesis presents the implementation of an implicit alternating method to solve a transient stability problem involving a large electrical power system. To achieve this, the system is split into three elements (or subsystems): the electrical transmission network, the dynamic electrical generators (as well as the associated dynamic equipment) and the distribution network. Considering the proposed dismemberment, each one of the subsystems can be modeled in different domains, such as frequency and time domains. Furthermore, it’s possible to assign appropriate computational tools and solution methods to each subsystem. As a result, the simulation is expected to offer greater modeling flexibility, potential for model reuse, and computational performance improvement. The first subsystem, corresponding to the transmission network and its loads (whether linear or non-linear), is represented by its equivalent positive sequence circuit using Python language, where the nodal solution method is implemented. For the second subsystem, corresponding to electrical machines, its dynamic nature must be considered and, therefore, the adopted computational tool is the Functional Mock-up Interface (FMI), which consists of a standardized and free interface that aims at coupling between mathematical models and simulators. The models, or libraries, that adopt the FMI interface are called Functional Mock-up Units (FMU), and can be classified as two different types: Co-Simulation (CS) and Model Exchange (ME). This study adopts the ME-type FMU, whose solution must be performed by an external solver based on the trapezoidal integration method. The third and final subsystem, corresponding to the distribution network, is represented through a package based on the OpenDSS, which consists of a simulator of electrical distribution systems and Distributed Energy Resources (RED). For Python-based communication between the transmission network and the other subsystems, synchronous machine models are implemented by the software OMEdit, encapsulated into FMUs by a Modelica compiler and imported into the Python environment by PyFMI library, while the distribution subsystem is imported by the OpenDSSDirect package. With the defined subsystems, the results obtained by the proposed simulation with an ME-type FMU match satisfactorily with those obtained by ANATEM, a program of time-domain dynamic simulation and analysis of electromechanical transients.