Efficiency-Oriented Multiobjective Modulation Optimization for the Dual Active Bridge Converter Using Fuzzy Logic-Aided Strategy

As a key component in modern power systems, dual active bridge (DAB) bidirectional DC-DC converters should maintain high efficiency across the full power range. Existing optimization strategies typically focus on single-objective approaches—such as minimizing peak current, RMS current, or backflow power—to improve efficiency. However, optimizing a single efficiency-related parameter does not necessarily maintain optimal efficiency across the entire power range. While minimizing power losses directly depends on an accurate power loss model, which requires recalibration whenever system components change. To address these limitations, this study formulates an efficiency-oriented multi-objective optimization problem (MOP) by incorporating multiple efficiency-related factors as objectives while considering soft-switching and mode boundaries as constraints. An offline two-step strategy is proposed to solve this MOP. First, the multi-objective evolutionary algorithm by decomposition (MOEA/D) is employed to generate a Pareto front surface. Second, a fuzzy inference system (FIS)-aided approach is introduced to identify the optimal efficiency point on the Pareto front. The optimal phase shift parameters for different voltage gains and transmission power levels are then stored in a lookup table for real-time implementation. The proposed multi-objective optimization strategy is validated through both power loss analysis and experiments, demonstrating efficiency improvements across the entire power range. Compared to other methods, the maximum efficiency improvement is observed at 80% of the maximum power point, increasing from 90.9% to 93.1% when the input voltage is 200 V, and the output voltage is 180 V.