Yin, Zhijian

PROJECT TITLE: Capacitor Sizing and Optimization for Highly Reliable MMC Systems

PhD period: 2016.11.01 – 2019.10.31.
Section: Power Electronic Systems
Research Programme: Efficient and Reliable Power Electronics
Supervisor: Huai Wang
Co-Supervisor: Yongheng Yang
Contact Information

Funding: China Scholarship Council; the Department of Energy Technology.


The modular multilevel converters (MMC) have been popularly used for high-power energy conversions over the past few years. As the voltage stress on the main switches can be significantly reduced, they are preferable in applications with high input voltages, such as renewable energy sources like wind turbines, PV systems. For improving the performance of MMC, intensive research has been devoted to advanced topology and control methods. However, few pieces of literature concentrate on the design of capacitors for MMC, especially from the reliability point of view. As a result, the existing design and control methods cannot guarantee to achieve a reliable performance in MMCs. Therefore, it is challenging to develop new solutions to evaluate the reliability of the entire MMC system.

Thus, this Ph.D. project aims to find an optimized design for the capacitors in MMC system. The research may contain the following aspects at the present stage:

● Sizing and configuration of capacitors of MMC

 As dc-link capacitors highly relate to the cost, size and failure rate of MMC system,this study will consider dynamic performance and worst operation conditions to size the required minimum capacitance and physical configuration of capacitor banks, which relates to the MMC control strategy as well.        

● Modeling and optimization of capacitor banks in MMC

During this stage, some mathematical models including electro-thermal-lifetime modeling and capacitor bank optimization will be set up corresponding to testing results. By using these models, it will be possible to predict the lifetime of the new converters implemented with different design and control methods. Also, these models can help researchers make a design optimization for the modular multilevel converters and keep a perfect balance between the life-cycle and total cost.

● Condition monitoring of capacitors based on predictive algorithm and wireless network

Real-time operation status of each component in MMC systems is of high concern. By developing predictive algorithms and wireless sensor networks, it is possible to gather all the information about the components, such as degeneration degree, temperature, etc. This will be beneficial for the maintenance and reliable operation of the converters. Furthermore, this information will provide component manufacturers with suggestions on improving the performance of their products. It is expected to build up a prototype to demonstrate a faster and easier condition monitoring method for power electronic customers in this project.


Publications in journals and conference papers may be found at VBN.