Global modeling method for ZVRT buck converters based on mixed logical inequalities

Purpose This paper aims to address the issues of excessive logical variables and complex constraint conditions arising in modeling multi-mode switching converters using mixed logical inequalities. Taking the zero-voltage resonant transition (ZVRT) buck converter as a representative example, a simplified modeling framework for multi-mode switching converters is proposed. Design/methodology/approach By using mixed logical inequalities, the proposed method characterizes the relationship between the constraint conditions of the converter’s equivalent operating modes and the logical variables. By using the continuity of the converter’s state variables during operation, when the equivalent constraints of a given mode consist of multiple sub-constraints, these can be unified into a single constraint by exploiting the continuity of the variables. In this way, the required number of logical variables introduced during modeling is reduced. Findings The ZVRT buck converter is adopted as a case study for validating the proposed modeling and analysis method. Simulation results show that the developed model can accurately reflect the actual operating process of the converter. The comparative simulation analyses among different models further confirm the feasibility and reliability of the proposed modeling method. In addition, a prototype is further built for experimental validation. The experimental results exhibit strong consistency with the simulation outcomes, providing effective evidence of the proposed model’s alignment with the actual system and thereby demonstrating the effectiveness of the modeling method. Originality/value A novel multi-mode modeling strategy is proposed for accurately describing the ZVRT buck converter, which effectively avoids the introduction of auxiliary logical variables when multiple sub-constraint conditions exist for the converter’s operating modes. Compared with the bi-switching modeling method, the number of logical variables is reduced by three, thereby simplifying the overall model structure and reducing modeling complexity.