Optimization Strategies for Triple-Phase-Shift Modulation in Dual-Active-Bridge Converters: A Comprehensive Review

Triple-phase-shift (TPS) modulation endows dual-active-bridge (DAB) DC-DC converters with three independent degrees of freedom, enabling simultaneous control of power flow, device stress, and soft-switching conditions. Exploiting this flexibility requires the development of carefully crafted duty-cycle schedules. These schedules must jointly minimize conduction and transformer losses, extend the zero-voltage-switching (ZVS) range, and accommodate wide input/output envelopes. This paper delivers a unified review of TPS duty-cycle optimization strategies. A systematic taxonomy is proposed, separating offline analytical and lookup-table solutions from online adaptive methods based on mathematical programming, metaheuristics and artificial-intelligence (AI) frameworks. Representative techniques are benchmarked against four key performance indices, RMS/peak current, efficiency, ZVS coverage, and reactive-power circulation, to expose inherent trade-offs among current stress, switching loss, and control complexity. Reported case studies indicate that segmented analytical schemes reduce worst-case RMS current while reinforcement-learning controllers sustain full-range ZVS achieving real-time efficiency improvements over static lookup tables. Emerging trends highlight hybrid AI-analytical controllers, physics-informed neural surrogates, and FPGA-accelerated digital twins that promise sub-100 $\mu $ s optimization latencies. Open challenges, such as holistic electro-thermal objectives and data-efficient training, are distilled into a roadmap for future research.