Problem Definition

Efficient power conversion for high-power applications is one of the biggest issue for decades. Multilevel power converters are used for efficient high-power conversion and one of the preferred choices for industry and academia. Currently these multilevel converters are used in many applications such as pumps, fans, grinding mills and extruders. Basic multilevel converters have three topologies: Neutral Point Clamped (NPC), flying capacitor (FLC), and cascade H-bridge. Because of its wide industrial application, researchers all over the world are working to improve its efficiency, reliability, cost, power density and simplicity of multilevel converters. There is no laboratory scale (3-5 kW) prototype convertors available in the market that fit all of the required applications. We need a plug and play convertor to fit the three power conversion applications. These converters will help out client to develop new power converter topologies and test new control schemes such as model predictive control. In this project, we will be working on multilevel converters using Neutral point clamped topology with model predictive control. We will implement a prototype of multilevel neural-point clamped (NPC) converter for high power applications and test it using model predictive control. Our prototype power level is at 5kW. We are using predictive control because with the help of predictive control we can handle system non-linearity and predictive control algorithm will provide better switching as compare to previous technique. In past decades, there is a great revolution in processor technologies. Inexpensive microprocessor had been introduced with very fast processing speed and can be used to control high frequency power converter switches very efficiently and will be very flexible as compare to previous approaches. So, in this project we will be controlling switches using control technology name as model predictive control