VMD-based adaptive ultrasonic flowmeter echo signal denoising algorithm

Flow rate measurement of drilling return fluid plays a crucial role in the oil and gas industry, with significant impact on operational safety assurance, efficiency enhancement, and optimized resource management. However, the complex characteristics of drilling return fluids present persistent challenges to the long-term reliability and accuracy of conventional intrusive flow measurement methods. As a non-invasive approach, flow measurement based on ultrasound Doppler frequency shift demonstrate substantial application potential in this field. In order to cope with the severe noise interference of ultrasonic echo signals caused by the harsh environment of oil fields, this paper proposes a novel ultrasonic denoising algorithm integrating Ivy Optimization Algorithm-optimized Variational Mode Decomposition (VMD) with improved wavelet soft-threshold denoising. Specifically, the proposed algorithm initially decomposes ultrasonic echo signals into multiple Intrinsic Mode Functions (IMFs) through VMD optimized by the Ivy Optimization Algorithm. Subsequently, to prevent the elimination of useful signal components during the improved wavelet soft-threshold denoising process for these IMFs, an energy spectral density-based classification algorithm is designed to categorize all IMFs into signal-dominant IMFs and noise-dominant IMFs. The noise-dominant IMFs are then subjected to improved wavelet soft-threshold denoising. Finally, the reconstructed signal is obtained by combining signal-dominant IMFs with denoised noise-dominant IMFs. Simulation experiments demonstrate that the proposed method achieves superior denoising performance and robustness under various signal-to-noise ratio (SNR) conditions. Compared with traditional wavelet denoising, it improves SNR by 120.1 % and outperforms recent advanced algorithms by at least 6.3 %. Field tests reveal that the measurement error of ultrasonic flow meters using this algorithm is reduced by 67 % compared with conventional counterparts. These results confirm that the proposed algorithm effectively denoises ultrasonic flowmeter echo signals, thereby significantly enhancing measurement accuracy. This advancement provides a reliable technical solution for precise flow rate monitoring in complex drilling environments.