Μοντελοποίηση Υβριδικών Δικτύων AC-HVDC στα πλαίσια Εκτιμητών Κατάστασης Συστημάτων Ηλεκτρικής Ενέργειας

Abstract

Electricity is an essential part of the modern way of life. The continuous and effective operation of Power Systems is a crucial necessity in order to sustain power flow to the consumers. Those responsible for the operation of the aforementioned systems, rely on the system’s current state in order to control it effectively, make decisions and take actions. The most common way of obtaining the system’s state, is its estimation using measurements derived from different parts of the Power System. The state estimator algorithm formulates an estimation of all system variables, that is an estimation of all bus voltage phasors, as accurately as possible, by taking into consideration all available measurements. Up to this point, state estimation relied mainly on SCADA (Supervisory Control and Data Acquisition) measurements. This method considers Power Flow equations and therefore leads to a non-linear measurement model, which requires the usage of iterative arithmetic methods in order to compute the solution of the state estimation. Also, measurements obtained through the SCADA system are sampled with a sampling period of several seconds. Phasor Measurement Units (PMUs) comprise an advanced and widely used technology that promises to solve the problems that arise with the traditional state estimation method. PMUs are able to measure voltage and current phasors directly, at the bus they are placed at, which in turn leads to the formulation of linear measurement models. At the same time, PMU sampling rates are much higher than those of the SCADA system, reaching 60 measurements per second. Furthermore, traditional state estimation methods have been implemented mainly for classic AC Power Systems. However, with the growing use of HVDC links in modern networks, as well as other types of FACTS, arises the need for proper modelling of hybrid AC-HVDC networks, in order to be able to solve state estimation problems in such grids. This diploma thesis firstly presents the fundamentals of state estimation theory and then attempts to develop an appropriate network model and static state estimation algorithm for hybrid AC-HVDC grids, which depends entirely on PMU measurements. Afterwards, the subject of bad data detection is briefly discussed and an algorithm for bad data detection is developed, according to the Largest Normalized Residual method. Finally, various test cases are studied and the results of the static state estimation algorithm are presented for each case.