Get started
Work

A Time-Difference Processing Algorithm for RTD-fluxgate Sensors Based on LOF and Weighted Averaging

Year: 2026
Type: article
Abstract: Residence Time Difference (RTD) fluxgate sensors exhibit advantages such as strong anti-interference capability and miniaturization. However, random core noise induced by the excitation magnetic field and environmental disturbances can cause abrupt anomalies in the time-difference signal, thereby degrading the measurement accuracy and stability of the sensor. To address the limitations of conventional outlier detection methods, such as the Laida criterion and the Mahalanobis distance with covariance, which suffer from insufficient detection accuracy and limited processing performance, this paper proposes an outlier detection and correction method that integrates the Local Outlier Factor (LOF) algorithm with weighted averaging. The proposed method employs the LOF algorithm to identify anomalous samples and introduces a standard-deviation-based adaptive threshold determination strategy for outlier detection. Subsequently, a weighted averaging scheme is applied to assign appropriate weights to anomalous data, thereby enabling effective correction of outliers. Experimental results show that the proposed method reduces time-difference fluctuation by approximately 80.7%, with only a 1% reduction in the mean and 95% retention of the original variance. Compared with the Mahalanobis distance with covariance method and the Laida criterion, the proposed method improves the average signal-to-noise ratio (SNR) by approximately 1.83 and 3.15 times, respectively. These results demonstrate that the proposed method effectively detects and corrects abnormal jumps, while preserving key signal characteristics and improving measurement stability, showing good engineering potential.
DOI: 10.1109/jsen.2026.3742940
Institution: Beihua University
Language: English
Open Access status: hybrid
Estimated APC paid: $2,645
OpenAlex ID: W7224425239
FWCI:
Cites:
Cited by:
Related to: 10