Modeling, analysis, and control of multi-port DC-DC converters

Multi-port DC-DC converters are an emerging technology because they have the potential to combine a variety of energy sources and optimize corresponding management strategies. Their utilization can help reduce the cost and size of the conventional energy sources. In this research, an integrated approach for the modeling, analysis, and control of N-port converters is discussed. The non-linear time-varying nature of the switching transients and the pulse width-modulation process associated with the N-port converters make this task extremely challenging. To overcome these challenges, the mainstay of the modeling approach, which is the cyclic average current (CAC) representation, is first derived for the N-port converter. A generalized circuit configuration of the transformer that is more suitable for the analysis is then established. The CAC is used to obtain the equivalent large-signal and small-signal models. Recognizing the importance of power conditioners, an optimization procedure to design filters while preserving the converter dynamical performance and stability is presented. A control method along with a sequential quadratic programming algorithm are finally developed to offer the flexibility to meet any desired control and power optimization objectives. The results of this research show that using the CAC representation significantly reduces the modeling and analysis complexity of N-port converters. Even though they are lossless, the derived models are able to accurately capture the dynamics of the converter. The controller performance is proven to be robust in disturbance rejection and reference tracking.