Conversores multiníveis modulares: contribuições ao estudo e ao desenvolvimento de sistemas de controle distribuídos

Abstract

This doctoral thesis presents contributions to the study and development of distributed control systems applied to the Modular Multilevel Converter (MMC) operating with a high number of submodules (SMs), which is the most promising multilevel converter topology, especially in high voltage and high power applications. Initially, operating principle, main concepts and a power flow analysis are presented in order to justify the proposal. Based on the MMC power flow analysis, a control system is proposed to ensure the power balance between the dc voltage bus and the ac power grid, the power processing distribution between the arms and the individual capacitor voltage control, with the control of N–1 capacitor voltages. One of the main MMC implementation challenges is to ensure the suitable control of SM floating capacitors with an easily scalable control system with reduced communication complexity. Thus, a distributed control structure using sliding mode observer approach is proposed to reduce the data flowing among through between the SMs needed to implement the MMC control system. With this concept, the sliding-mode observer approach consists of: (i) an equivalent capacitor voltage observer in the continuous-time domain, which are discretized for processor implementation; and (ii) an equivalent capacitor voltage observer in the discrete-time domain. Regarding the equivalent observers, the design methodology for continuous-time and the results for the discretized observer are presented. The design methodology of the discrete-time equivalent capacitor voltage observer is shown in discrete-time to guarantee the steady-state and the transient performances and the results for the discrete observer are presented. The Lyapunov stability analysis is presented for each observer. Finally, it is worth to emphasize that simulation and experimental results are presented considering the observer and the control performances under different operating conditions.