pp. 193-205
S&M3900 Research Paper of Special Issue https://doi.org/10.18494/SAM5257 Published: January 31, 2025 Mamba-based Multibranch State Space Iterative Fusion Algorithm for Multisource Power Grid Survey Data [PDF] Aiyuan Zhang, Jingguo Lv, Jiyong Zhang, Xiaohu Sun, Chunhui Zhao, Changjiang Yang, and Junjie Sun (Received July 26, 2024; Accepted January 20, 2025) Keywords: multisource remote sensing data fusion, power grid surveying, Mamba, iterative attention mechanism
The effective integration of multisource survey data for power grids benefits designers by providing comprehensive and accurate analyses of the terrain and landforms surrounding the survey area. In this study, inspired by the Mamba concept, we propose an iterative attentional feature fusion Mamba (iAFF-FMA) framework that constructs a multibranch state space for iterative fusion, reducing differences between data modalities and enhancing feature interaction within the same modality. Experiments conducted with actual engineering data from ultra-high-voltage direct current (UHVDC) transmission lines demonstrate the iAFF-FMA framework’s superiority over six common fusion methods. This offers a novel technical approach to the integration of power grid survey data.
Corresponding author: Jingguo Lv![]() ![]() This work is licensed under a Creative Commons Attribution 4.0 International License. Cite this article Aiyuan Zhang, Jingguo Lv, Jiyong Zhang, Xiaohu Sun, Chunhui Zhao, Changjiang Yang, and Junjie Sun, Mamba-based Multibranch State Space Iterative Fusion Algorithm for Multisource Power Grid Survey Data, Sens. Mater., Vol. 37, No. 1, 2025, p. 193-205. |