Research On Model Predictive Control Of Multi-Module Dual Active Bridge Converters In Power Electronic Transformers

With the increasingly severe issues of energy security and environmental protection in China,the development of new energy technologies has received more widespread attention.In the field of new energy power conversion,power electronic transformers have many functions such as intelligent energy conversion,precise power control,reactive power regulation,fault isolation,so have broad application prospects.The dual active bridge converter,due to its advantages of bidirectional energy transmission,high-frequency electrical isolation,high power density,and easy implementation of soft switching,meets the control requirements of the intermediate stage DC/DC converter in power electronic transformers and has become an ideal topology for intermediate stage converters.In multi-module series-parallel connected dual active bridge converters applied to high-power scenarios,fast dynamic response,inter module power balance,high modularity,and high reliability are important performance requirements.However,existing control methods are difficult to balance these requirements.Applying model predictive control with high dynamic performance,multi-objective optimization ability,and no multi loop nested structure to multi-module dual active bridge converters can achieve comprehensive improvements in control performance.This paper conducts research work with these goals.Firstly,to address the problem of the existing continuous control set model predictive control of single-module dual active bridge converters being susceptible to sampling noise,this paper analyzes the sampling noise sensitivity from the optimal phase shift ratio calculation formula,and proposes a sampling noise suppression continuous control set model predictive control method with variable prediction discrete step.This method does not require additional filtering equipment or algorithms and is easy to implement.Through simulation and experimental verification,the results show that the proposed method can effectively suppress the increase in output voltage ripple and inductance current stress caused by sampling noise,improving the applicability of the dual active bridge converter model predictive control.Secondly,the sub modules of multi-module series-parallel connected dual active bridge converters may experience power imbalance due to inconsistent actual parameters.This paper analyzes the principles and compares the advantages and disadvantages of three classic power balancing strategies:voltage gradient control,dual loop decoupling control,and master-slave independent control.To address the problem of poor dynamic performance of master-slave independent control,this paper proposes a power balancing continuous control set model predictive control method for multi-module dual active bridge converters which combines high dynamic performance,high modularity,and high control accuracy.Through simulation verification,the results show that the power balance control effect of the proposed method is good,the dynamic response speed is greatly improved,and the shortcomings of traditional master-slave independent control have been overcome.Finally,in order to further improve the modularity and response speed of the system,this paper applies the moving discretized control set model predictive control,which is easy to achieve multi-objective optimization,to the multi-module series-parallel converter,and proposes a power balancing moving discretized control set model predictive control method for the multi-module dual active bridge converter.This method concentrates the two objectives of output voltage control and input voltage balance control in the same model predictive controller,which has higher dynamic performance,realizes the standardization of the sub module controller,and improves the scalability and reliability of the system.Through simulation verification,the results show that the proposed method achieves further improvement in response speed of output voltage control and power balance control and has a higher degree of modularity.