Modelado de sistemas de baja tensión en 4 hilos para estrategias de control mejoradas de convertidores conectados a red

Resumo

The evolution of the electricity system is leading to the emergence of new determinants in the traditional scenario. Thus, the presence of renewable sources, distributed generators, solutions for energy storage, intelligent systems and the proliferation of non-linear loads, represent a set of new challenges for the operation and safety of the system itself. The incorporation of these new players brings with it a change of paradigm: from a centralized system to a more distributed one. In this new scenario, generators are located closer to consumers, promoting systems with a high level of energy autonomy which, in the extreme case, can become independent from the public grid. This is how the concept of the microgrid arises, framed in the context of what we know today as smart grids. In these new networks, the incorporation of distributed resources interconnected by means of power converters plays a decisive role for the correct operation of the system. This work focuses on the operation of grid-connected power converters in the context of AC microgrids. Special attention has been paid to the problem of the equivalent impedance variability of the network upstream from the low voltage connection point of these converters, relating this issue to the impact it causes on the dynamic behaviour of the control system of these devices. The proposed solutions, based on impedance estimation techniques, allow the implementation of control systems that react to changes in the network and achieve real-time compensation of unbalanced conditions. To carry out this task, they only need the information of electrical variables measured at the connection point, making use of the flexibility of the converters to generate excitation signals superimposed on the fundamental component. In addition, the estimation of network parameters provides opportunities related to the improvement of system efficiency, so optimization solutions based on the operation of power converters at variable frequency have also been proposed. This thesis is organized in six main chapters to which a final chapter of conclusions is added. The first two serve to detail motivation, research opportunities and the study of the current state of the art. This provides an overview of the problem to be studied and the context of this work, as well as a description of the main challenges faced by grid-connected power converters. The contributions of this thesis, developed over the next four chapters, can be summarised as follows: 1) The development of a new methodology for network modelling, which allows to express in a compact way the equivalent impedance seen from the connection point of the distributed resource. This methodology includes four-wire systems and can be applied under both balanced and unbalanced conditions. The proposed model allows to determine, through a system of representation in stationary coordinates, the temporal evolution of the voltages/currents before the connection/disconnection of different local loads; 2) The implementation of a reliable technique to determine the equivalent impedance of a system at the fundamental frequency based on the injection of high frequency signals. This technique uses the converters themselves as a source of excitation; 3) A control strategy, based on the estimation of the impedance at the connection point, capable of compensating the negative and zero sequence components without the need for the additional sensors that are generally used to measure these components; 4) The extension of the methodology of impedance modelling in passive networks, understood as those without additional power converters, to networks with different converters operating simultaneously in parallel. This model has been used to propose an alternative impedance estimation technique, based on pulse injection and the application of a Recursive Least Squares (RLS) estimation algorithm; 5) A system efficiency optimization solution that, based on network impedance estimation, makes use of the operation of power converters at variable switching frequency. All contributions have been validated, both in simulation and experimentally, in the power systems laboratory of the LEMUR research group at the University of Oviedo.