Up to now, massive power electronics converters have been derived to satisfy various demands in practical applications. However, most of these methods focus on discovering effective converters, ignoring the redundant switches in converters. This paper aims to delete converters with redundant switches efficiently. First, the identification principle is introduced based on the connection between legitimate switching modes and redundant switches. After that, the principle is transformed into the corresponding graph principle. Finally, graph theory and graph searching algorithms are used to develop the auto identification method based on the graph principle. The results demonstrate that all redundant switches may be effectively recognized with the use of computers when the suggested method is applied to single-inductor multi-port converters.
- Article type
- Year
With the widespread application of single-inductor multiple-output (SIMO) DC-DC converters, their dynamic response has much concern for domestic and foreign researchers. To explore the theoretical limit of the optimal dynamic response for the SIMO Buck converter, this paper first establishes a corresponding mathematical model according to the working principle of the converter, and then an improved differential evolution (DE) algorithm is employed to search out the solution. In addition, the suggested approach can investigate the best dynamic response with various goals, including the best self-regulation or cross-regulation. It is also possible to know the optimal dynamic response with different constraints, such as different dynamic response times and peak inductor currents. In order to completely recognize the dynamic performance of SIMO DC-DC converters, the theoretical limit exploration of optimal dynamic response based on the heuristic DE intelligent algorithm is useful. Ideally, this will direct the controller design in order to obtain an optimized dynamic response.
In the process of deriving topologies for power electronic converters, researchers often use topology equivalence or topology isomorphism to identify topologies with different structures but identical performance, thereby avoiding redundant studies. However, the connotations of topology equivalence and topology isomorphism differ significantly, which can easily lead to confusion. To address this issue, this paper aims to clarify the distinctions and connections between the two concepts and propose a method for accurately identifying topologies with identical performance. First, it is deduced that a necessary condition for topology isomorphism is that their adjacency matrices have the same determinant. Furthermore, it is derived that the necessary and sufficient condition for topology equivalence is that the components are identical and the fundamental loops correspond one-to-one. Subsequently, an analysis from the perspective of topology subgraphs reveals that topology isomorphism is a sufficient but not necessary condition for topology equivalence, while topology equivalence is a necessary and sufficient condition for identical performance. Based on this, the paper recommends prioritizing topology equivalence over topology isomorphism in practical applications to identify topologies with identical performance. The theoretical results are validated through case studies demonstrating their correctness and feasibility. Additionally, this paper proposes a method based on the depth-first search algorithm to automatically determine equivalent topologies, enabling the rapid and accurate identification of converter topologies with identical performance.
京公网安备11010802044758号