Scholarlink

<p>With the widespread application of sealed electronic components in aerospace, automotive electronics, and communication equipment, traditional component signal processing methods are difficult to accurately identify subtle changes in complex nonlinear and non-stationary signals. Therefore, a signal processing and recognition method based on dynamic time warping optimization algorithm is proposed. The improved Sakoe Chiba global constraint is introduced in the standard dynamic time warping framework to limit the path search range, and multiple lower bound functions are combined to achieve fast distance filtering and upper bound pruning, improving matching efficiency, reducing redundant calculations, and significantly improving the accuracy and efficiency of signal recognition. The experimental results showed that the proposed dynamic time warping optimization algorithm performed well on multiple public datasets, not only outperforming traditional methods in recognition accuracy, but also significantly reducing computation time. The average running time on the four datasets was 5.7s, 4.8s, 6.5s, and 10.5s, respectively. In addition, the root mean square errors were 4.4378, 2.2757, 5.1274, and 3.3756 on four datasets, the mean absolute percentage errors were 1.2456, 2.4763, 2.4123, and 1.4537, respectively, demonstrating strong robustness and generalization ability, and providing effective technical support for efficient processing and accurate identification of sealed electronic component signals.</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

Signal processing and recognition technology of sealed electronic components: based on DTW optimization algorithm

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月14日

查看全文
  • Yuchao Yang,
  • Xu Yang

摘要

With the widespread application of sealed electronic components in aerospace, automotive electronics, and communication equipment, traditional component signal processing methods are difficult to accurately identify subtle changes in complex nonlinear and non-stationary signals. Therefore, a signal processing and recognition method based on dynamic time warping optimization algorithm is proposed. The improved Sakoe Chiba global constraint is introduced in the standard dynamic time warping framework to limit the path search range, and multiple lower bound functions are combined to achieve fast distance filtering and upper bound pruning, improving matching efficiency, reducing redundant calculations, and significantly improving the accuracy and efficiency of signal recognition. The experimental results showed that the proposed dynamic time warping optimization algorithm performed well on multiple public datasets, not only outperforming traditional methods in recognition accuracy, but also significantly reducing computation time. The average running time on the four datasets was 5.7s, 4.8s, 6.5s, and 10.5s, respectively. In addition, the root mean square errors were 4.4378, 2.2757, 5.1274, and 3.3756 on four datasets, the mean absolute percentage errors were 1.2456, 2.4763, 2.4123, and 1.4537, respectively, demonstrating strong robustness and generalization ability, and providing effective technical support for efficient processing and accurate identification of sealed electronic component signals.