Design, Analysis and Control of Power Converters for Solid-State Transformer Applications

Abstract

This thesis investigates the design, analysis and control aspects of the power converters for SST configurations. The focus of this thesis is set on the analysis of an SST topology consisting of a multi-cell AC/DC converter, input-series output-parallel (ISOP) isolated DC/DC converter and a DC/AC converter. For the first step, a single-phase multi-cell rectifier is proposed and analysed. The proposed converter is conceptualized to reduce the control complexity of the multi-cell converter. Moreover, this converter is capable of reducing the size of the filter components for a specific number of converter cells compared to the cascaded H-bridge (CHB) converter. Furthermore, another multi-cell converter solution with CHB converter is analysed for the development of a high power SST configuration. For the DC/DC stage, DAB has been investigated in this thesis. A specific guideline for the DAB converter design is proposed in order to achieve higher efficiency at full load. As an alternative solution, the application of phase-shifted full-bridge (PSFB) DC/DC converter is investigated in this work. The PSFB converter overcomes some of the disadvantages of the prior art and provides a useful alternative for SST applications. Experimental results are presented which confirm that the leakage inductance of the transformer does not constrain the control of power flow. Besides, the control aspects of a three-phase four-leg inverter are analysed. As an alternative approach to the conventional method, model predictive control is employed as the control method which allows simplicity in the implementation and better transient performance and inclusions of other control goals in the cost function. Moreover, the control and design issues of a three-phase two-level voltage source converter are also investigated. The predictive control approach integrated with discrete space vector modulation technique! is investigated and compared to conventional solutions. The experimental results confirm that the proposed approach can significantly improve the input current quality of the converter with an acceptable computational burden and solves one of the major drawbacks of the predictive control approach. Based on the design and analysis of the above-mentioned power converters, the hardware prototypes for the two SST topologies are then realized.