Scholarlink

<p>Despite the success of deep learning models for the state-of-health estimation of lithium-ion batteries in battery management systems, their susceptibility to adversarial attacks raises concerns about security risks, leading to misdiagnosis and unnecessary maintenance. However, their practical robustness against adversarial perturbations remains largely underexplored, despite its importance for real-world deployment. Thus, we develop a novel adversarial attack to unveil potential risks associated with battery management systems. Our approach assesses model robustness in practical scenarios where model information is inaccessible by selectively exploiting frequency information that benefits the capture of unique characteristics for the state of health. In particular, we generate a <i>malicious</i> but <i>imperceptible</i> example by manipulating low-frequency components of a signal while preserving their relationships. Through a series of experiments, we demonstrate the effectiveness of our approach in uncovering and understanding potential risks associated with deep learning models in battery management systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Generating imperceptible adversarial examples via low-frequency aware transfer attacks on battery management systems

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月17日

查看全文
  • Kyeongseo Min,
  • Yeseo Joo,
  • Jiho Hong,
  • Hoki Kim,
  • Sangho Lee,
  • Youngdoo Son

摘要

Despite the success of deep learning models for the state-of-health estimation of lithium-ion batteries in battery management systems, their susceptibility to adversarial attacks raises concerns about security risks, leading to misdiagnosis and unnecessary maintenance. However, their practical robustness against adversarial perturbations remains largely underexplored, despite its importance for real-world deployment. Thus, we develop a novel adversarial attack to unveil potential risks associated with battery management systems. Our approach assesses model robustness in practical scenarios where model information is inaccessible by selectively exploiting frequency information that benefits the capture of unique characteristics for the state of health. In particular, we generate a malicious but imperceptible example by manipulating low-frequency components of a signal while preserving their relationships. Through a series of experiments, we demonstrate the effectiveness of our approach in uncovering and understanding potential risks associated with deep learning models in battery management systems.