Modeling, control, and implementation of compact multilevel converters for power quality applications

Abstract

To ensure access to affordable, reliable, and modern energy for all, increasing the penetration of renewable energies into the electricity grid is of great importance. Consequently, power quality-related challenges should be considered prudently. Therefore, the focal goal of this thesis is to develop various high-power-density multilevel converter topologies and the required controllers for improving the power quality of electrical systems efficiently and reliably. First, a seven-level modified packed U-cell (MPUC7) based static synchronous compensator (MPUC7-STATCOM) with an autotuned finite control-set model predictive control (AFCSMPC) is introduced. In comparison with a seven-level cascaded H-bridge (CHB) based STATCOM not only, the proposed configuration has exceptionally reduced active/passive component count, but also MPUC7-STATCOM designed AFCS-MPC control method complexity is attenuated meaningfully. Boost-mode operation and low voltage rating of the components can be also mentioned as the merits of the MPUC7-STATCOM. Moreover, the weighting factors of the proposed AFCS-MPC are tunable automatically and effectively in real-time. Second, a novel configuration of electric spring based on the modified five-level packed u-cell (MPUC5) inverter for mitigating harmonics and voltage fluctuations at various points of a grid with unstable generated power from distributed renewable energy sources is introduced. Moreover, operation principles, design procedure, and configuration of the MPUC5-based electric spring (MPUC5-ES) are presented. A simple and yet efficient controller without any extra control loop for regulating DC bus voltages has been also designed and applied to MPUC5-ES. Third, a high-power-density multilevel capacitor-based electric spring (ES-1) configuration with boost-mode operation based on a seven-level modified packed U-cell converter (MPUC7- ES1) is presented. A novel control strategy based on finite control-set model predictive current control (FCS-MPCC) is also proposed for MPUC7-ES1 application. This algorithm is designed to predict the system behavior for all the conceivable switching vectors based on the discrete models of MPUC7-ES1 that is developed for the first time. Comparing to the conventional ES- 1 linear control methods, the proposed strategy has key merits including considering the dynamic models of ES-1 converter, not requiring a modulator, and lower switching frequency. Fourth, aiming at delivering power to sensitive loads with an enhanced level of reliability and quality, a compact multilevel battery-based electric spring (ES-2) topology, founded on the Packed E-Cell (PEC) inverter, and its respective artificial neural network (ANN) based control strategy are introduced. From the reliability point of view, the PEC-based ES-2 (PEC-ES2) has the capability of instant nine to five-level operation under its bidirectional switch faulty condition. Regarding the power quality, in comparison with the half or full-bridge ES-2 topologies, PEC-ES2 has halved voltage rating switches, lower harmonic content in its output current and voltage, considerably lower switching frequency, higher power applications, etc. The proposed intelligent ANN-based controller can also tune and stabilize both the grid voltage and responsive load setup power factor independently and instantly with improved dynamic performance. Steady-state and dynamic operations of the introduced compact multilevel electric spring and STATCOM topologies and the proposed control techniques are also illustrated through extensive simulation and experimental results.

Date	13 Dec 2021
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Ambrish Chandra (Supervisor)