Penalty Functional Beamforming Algorithm Based on Quadtree Hierarchical Grid Search for Sound Source Identification

Le résumé fourni par la source

To enhance the sound source identification performance of functional beamforming (FB), this paper proposes a penalty functional beamforming algorithm based on a quadtree hierarchical grid search, denoted as H-PFB. The algorithm constructs a penalty matrix by normalizing the FB output and uses penalty coefficients to dynamically scale steering vectors, thereby iteratively optimizing the point spread function to improve spatial resolution. Furthermore, H-PFB employs a quadtree hierarchical grid search strategy to perform coarse-to-fine grid refinement, which not only enhances computational efficiency but also mitigates steering vector mismatches caused by sound source deviations from grid points. Simulation and experimental results demonstrate that the proposed algorithm achieves superior identification performance and improves computational efficiency by approximately four times compared to FB. In an experiment to identify sound sources from compressed gas leaks in pipelines, H-PFB yielded clear imaging results and precisely identified the leakage sources, verifying its effectiveness for practical engineering applications.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.