The dual-active-bridge (DAB) converter serves as a crucial galvanic isolating solution to provide dc grid-forming for dc elements in low-voltage direct-current (LVdc) systems. Key performance metrics such as efficiency, current stress, power density, and cost of DAB converter are chiefly subject to the optimal design of magnetic components and modulation strategies. However, existing DAB converter designs yield compromised solutions that optimize a limited subset of these metrics. This article develops a comprehensive analytical framework to characterize DAB converter operation across three key dimensions: 1) zero-voltage switching (ZVS) range; 2) power rating utilization; and 3) reactive power. To achieve a well-balanced design, a holistic optimization methodology is proposed, integrating multiobjective particle swarm optimization (MOPSO) with triple phase-shift control. By optimally selecting the transformer turns ratio and product of switching frequency and series inductance, the proposed MOPSO approach can collectively or selectively improve these performance aspects, enabling tailored DAB converter designs to meet diverse performance objectives. Experimental validation on a 1-kW DAB converter prototype demonstrates enhanced ZVS capability, improved utilization of converter rating, reduced reactive power, and achieves a peak efficiency over 95.9%.

Triple-active-bridge (TAB) converters are key multiport hubs in DC systems, but constant-power loads (CPLs) and cross-port coupling can induce oscillations and erode stability margins. This article presents a stability-oriented and optimization-based impedance-reshaping controller design for TAB converters that retains the standard PI structure while systematically tuning the gains to meet interface stability requirements. The design directly targets the port output impedance and reshapes it in both magnitude and phase over a specified frequency band by minimizing the ℓ ₂ norm objective. This band-limited magnitude–phase shaping focuses on the impedance-interaction region responsible for CPL-driven instability, enlarging the Nyquist stability margin and improving tolerance to interface uncertainty. Compared with conventional approaches, the proposed method requires no additional control path and no prior knowledge of CPLs’ control strategies or parameters, facilitating practical deployment in multiconverter systems. Simulation and hardware experiments validate the approach, demonstrating elimination of oscillations, enlarged stability margins, and robustness of TAB converter systems.