A multi-objective weighted constraint axial segmentation structure and collaborative optimization method for suppressing abnormal temperature rise in bridge arm reactors

Abstract

As a critical component of flexible DC transmission systems, the thermal stability of bridge arm reactors directly determines system reliability. Under AC-DC composite conditions, nonuniform current sharing among encapsulations intensifies local Joule heating and creates hotspots, limiting performance and service life. Conventional design approaches typically decouple structural design from parameter optimization, impeding effective thermal management. This study proposes a multi-objective, weighted-constraint method that jointly optimizes axial segmentation structure and encapsulation parameters. A constraint framework integrating current density, resistive voltage-drop, and temperature rise is established, with parameters optimized via gradient descent. By coupling parameter optimization with axial segmentation design, the proposed approach further introduces zonal weighting to regulate the upper and lower encapsulations according to their distinct heat dissipation characteristics. The effectiveness of the strategy is verified through finite-element simulations and prototype experiments. Results indicate that, with equal constraint weights, hotspot temperature drops by approximately 11% relative to a traditional design. After a second-stage refinement using the zonal weighting strategy, the hotspot temperature is further reduced by approximately 12.5%, with significantly improved temperature uniformity. Overall, the proposed method effectively suppresses hotspots and balances thermal distribution, offering a practical route for temperature-rise-constrained optimal design of bridge arm reactors.