Scholarlink

<p>Addressing the issue of flow measurement accuracy being easily affected by pipeline and environmental noise when combining Distributed Acoustic Sensing (DAS) technology with Flow-Induced Vibration (FIV) for flow monitoring. Environmental noise is mainly Gaussian noise. This paper proposes an improved Variable Step Size Affine Projection (VSS-AP) adaptive noise canceller algorithm. This algorithm introduces the autocorrelation of the error signal as the independent variable of the step size function, dynamically adjusting the step size factor, effectively weakening the impact of Gaussian noise, while significantly reducing steady-state error at the same time. Simulation results show that the improved algorithm has significant advantages in convergence performance and steady-state error. In actual flow measurement, mixed signals and reference noise signals are obtained through a double-layer pipeline, and processed using the adaptive noise canceller of the improved algorithm. Compared with the Least Mean Square (LMS), Fx-LMS, and Affine Projection (AP) algorithms, the denoising performance of the improved algorithm has increased by 24.9, 31.1, and 49.8% respectively, and the flow measurement error has decreased by 50.39, 40.06, and 39.07%. This paper provides an efficient and practical solution for oil well flow measurement, which has important engineering application value.</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

Adaptive Noise Canceller Based on Improved Affine Projection Algorithm

  • 期刊论文

  • 发表日期：

    2025年11月12日

查看全文
  • Lei Liang,
  • Zhengjie Han,
  • Xiaoling Tong,
  • Shu Dai,
  • Tianmignxuan Cao

摘要

Addressing the issue of flow measurement accuracy being easily affected by pipeline and environmental noise when combining Distributed Acoustic Sensing (DAS) technology with Flow-Induced Vibration (FIV) for flow monitoring. Environmental noise is mainly Gaussian noise. This paper proposes an improved Variable Step Size Affine Projection (VSS-AP) adaptive noise canceller algorithm. This algorithm introduces the autocorrelation of the error signal as the independent variable of the step size function, dynamically adjusting the step size factor, effectively weakening the impact of Gaussian noise, while significantly reducing steady-state error at the same time. Simulation results show that the improved algorithm has significant advantages in convergence performance and steady-state error. In actual flow measurement, mixed signals and reference noise signals are obtained through a double-layer pipeline, and processed using the adaptive noise canceller of the improved algorithm. Compared with the Least Mean Square (LMS), Fx-LMS, and Affine Projection (AP) algorithms, the denoising performance of the improved algorithm has increased by 24.9, 31.1, and 49.8% respectively, and the flow measurement error has decreased by 50.39, 40.06, and 39.07%. This paper provides an efficient and practical solution for oil well flow measurement, which has important engineering application value.