Scholarlink

<p>In response to the problems of complex circuits, high costs, poor anti-interference ability, and limited accuracy in traditional phase measurement methods, this paper proposes a gradient descent phase measurement method based on the minimum mean square error criterion. This method combines the minimum mean square error criterion to construct a gradient descent phase measurement model and improves the accuracy and anti-interference ability of phase measurement while reducing the number of iterations by using an adaptive step-size factor, phase edge processing, and setting stop conditions. The study further establishes a numerical simulation analysis model, which is based on six sets of simulation experiment data for in-depth analysis. Simulation results show that, at SNR = 30dB, the proposed method achieves a measurement accuracy of approximately 0.1 degrees, outperforming most traditional methods and meeting the accuracy requirements for most phase measurement applications. This method effectively solves the problems of step-size factor selection and phase edge processing, has strong noise resistance, and is suitable for different phase measurements, providing strong technical support for phase measurement in engineering applications.</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

Research on phase measurement method based on gradient descent with the minimum mean square error criterion

  • 期刊论文

  • 发表日期：

    2026年3月14日

查看全文
  • Xin Zhang,
  • Haitao Liu,
  • Xingming Fan

摘要

In response to the problems of complex circuits, high costs, poor anti-interference ability, and limited accuracy in traditional phase measurement methods, this paper proposes a gradient descent phase measurement method based on the minimum mean square error criterion. This method combines the minimum mean square error criterion to construct a gradient descent phase measurement model and improves the accuracy and anti-interference ability of phase measurement while reducing the number of iterations by using an adaptive step-size factor, phase edge processing, and setting stop conditions. The study further establishes a numerical simulation analysis model, which is based on six sets of simulation experiment data for in-depth analysis. Simulation results show that, at SNR = 30dB, the proposed method achieves a measurement accuracy of approximately 0.1 degrees, outperforming most traditional methods and meeting the accuracy requirements for most phase measurement applications. This method effectively solves the problems of step-size factor selection and phase edge processing, has strong noise resistance, and is suitable for different phase measurements, providing strong technical support for phase measurement in engineering applications.