跳到主要导航 跳到搜索 跳到主要内容

Event camera-based high-efficiency transient sparking fault detection in Hall thrusters

  • Yibo Zhang
  • , Jing Jin
  • , Wenzhe Zheng
  • , Yihao Zhou
  • , Zhucheng Tan
  • , Yan Shen*
  • , Chenyang Shi*
  • *此作品的通讯作者
  • Beihang University
  • Tianmushan Laboratory
  • Sun Yat-Sen University

科研成果: 期刊稿件文章同行评审

摘要

This paper presents a novel approach to detecting sparking phenomena in Hall thrusters using an event camera. The method addresses challenges associated with propulsion system reliability in the aerospace industry. Hall thrusters, commonly used in commercial satellites and deep space exploration, require reliable operation. The sparking phenomenon, one of the key faults of Hall thrusters, disrupts the normal behavior of the plume and poses multiple risks to thruster operation and mission success. Therefore, detecting sparking is essential. Compared with traditional diagnostic methods, visual sensing achieves finer spatial characterization but frame-based cameras remain limited in dynamic perception and on-orbit practicality. Event cameras, with microsecond-level time resolution, low power consumption, and a wide dynamic range, offer great potential for transient sparking detection. This paper is the first to utilize event cameras for detecting transient sparking in Hall thrusters. A novel detection method based on event rate bursts and ambient diffusion optical flow estimation is proposed. When plume fluctuations cause the event rate to exceed a defined threshold, optical flow computation is triggered for spark verification. Ground experiments show that the method can efficiently detect sparks with an average throughput of 334.77 kHz, achieve 95.7% detection accuracy, and continuously record the spark process. Comparative results with high-speed cameras confirm the superior performance of the event camera. The reliability and scalability of the method are also examined. These advances lay a significant foundation for future on-orbit fault detection and monitoring of Hall thrusters.

源语言英语
文章编号101085
期刊Journal of Industrial Information Integration
51
DOI
出版状态已出版 - 5月 2026

指纹

探究 'Event camera-based high-efficiency transient sparking fault detection in Hall thrusters' 的科研主题。它们共同构成独一无二的指纹。

引用此