Wu Di, Liang Qing-Quan, Hu Bing-Xuan, Zhang Ze-Ting, Wang Xue-Feng, Jiang Jia-Jun, Yi Gao-Wei, Zeng Hong-Yao, Hu Jin-Yuan, Yu Yang, Zhang Zhen-Rong
School of Computer Electronic and Information, Guangxi University, Nanning 530004, China.
College of Sciences, National University of Defense Technology, Changsha 410073, China.
Sensors (Basel). 2025 Aug 14;25(16):5052. doi: 10.3390/s25165052.
This review systematically examines intelligent event perception in distributed acoustic sensing (DAS) systems. Beginning with the elucidation of the DAS principles, system architectures, and core performance metrics, it establishes a comprehensive theoretical framework for evaluation. This study subsequently delineates methodological innovations in both traditional machine learning and deep learning approaches for event perception, accompanied by performance optimization strategies. Particular emphasis was placed on advances in hybrid architectures and intelligent sensing strategies that achieve an optimal balance between computational efficiency and detection accuracy. Representative applications spanning traffic monitoring, perimeter security, infrastructure inspection, and seismic early warning systems demonstrate the cross-domain adaptability of the technology. Finally, this review addresses critical challenges, including data scarcity and environmental noise interference, while outlining future research directions. This work provides a systematic reference for advancing both the theoretical and applied aspects of DAS technology, while highlighting its transformative potential in the development of smart cities.
本综述系统地研究了分布式声学传感(DAS)系统中的智能事件感知。从阐明DAS原理、系统架构和核心性能指标开始,它建立了一个全面的评估理论框架。本研究随后描述了传统机器学习和深度学习方法在事件感知方面的方法创新,以及性能优化策略。特别强调了混合架构和智能传感策略的进展,这些策略在计算效率和检测精度之间实现了最佳平衡。涵盖交通监测、周边安全、基础设施检测和地震预警系统的代表性应用展示了该技术的跨领域适应性。最后,本综述解决了包括数据稀缺和环境噪声干扰在内的关键挑战,同时概述了未来的研究方向。这项工作为推进DAS技术的理论和应用方面提供了系统的参考,同时突出了其在智慧城市发展中的变革潜力。