TL;DR
This paper introduces a novel demodulation algorithm for interferometric fiber optic sensors that significantly improves accuracy and stability. By combining a Phase Delay Compensation (PDC) module with an Extended Kalman Filter (EKF), the algorithm effectively eliminates nonlinear distortions caused by variations in modulation depth and carrier phase delay, achieving state-of-the-art performance.
What Problem Does This Paper Address?
Interferometric fiber optic sensors are widely used due to their simple structure and high performance. However, their demodulation process, particularly using the Phase-Generated Carrier (PGC) method, is susceptible to nonlinear distortions. These distortions arise from two main sources: fluctuations in the "C value" (modulation depth) and an initial "carrier phase delay" (θ). The C value is influenced by factors like the fiber's refractive index and optical path differences, which can change with environmental conditions. The carrier phase delay is often introduced by signal transmission and optoelectronic conversion. These issues degrade the accuracy and linearity of the demodulated signal, limiting the sensor's overall performance. Existing methods often address one of these issues but struggle when both are present simultaneously, leading to inaccurate readings and nonlinear errors.
Methods
The core innovation of this paper is the development of a PGC-PDC-EKF joint algorithm. This approach integrates two key components:
- Phase Delay Compensation (PDC) Module: This module is designed to accurately extract and compensate for the carrier phase delay (θ). It achieves this by introducing a compensation phase into the carrier signals and then minimizing an objective function that represents the discrepancy between the filtered signals. This allows for precise θ estimation and correction, ensuring proper synchronization between the reference carrier and the interference signal.
- Extended Kalman Filter (EKF) Module: This module tackles the nonlinear distortions caused by fluctuations in the C value. It works by normalizing the amplitude of filtered signals and then employing an EKF to suppress harmonic distortions. This effectively stabilizes the demodulation process even when the C value drifts.
The combined PGC-PDC-EKF algorithm first uses the PDC module to handle the phase delay and then the EKF module to address C value variations, leading to a more robust and accurate demodulation.
Datasets
Not discussed in the paper. The study relies on simulated data and experimental measurements from a constructed optical fiber sensing system.
Key Insights and Ideas
- Synergistic Compensation: The paper demonstrates that addressing both carrier phase delay (θ) and modulation depth (C value) variations simultaneously is crucial for high-stability demodulation. The proposed PDC-EKF approach shows that a sequential compensation strategy (PDC first, then EKF) is highly effective.
- PDC Insensitivity to C Value: A key finding is that the PDC module's phase delay calculation is largely unaffected by changes in the C value, as shown in Figure 3. This independence simplifies the overall system design and improves robustness.
- EKF for Harmonic Suppression: The EKF's ability to suppress high-frequency harmonic components is vital for mitigating the nonlinear distortions introduced by C value fluctuations, leading to cleaner demodulated signals.
- Elimination of Phase Singularities: The proposed PGC-PDC-EKF scheme, by compensating for θ, effectively eliminates the phase singularities that plague other PGC algorithms at specific phase delay values (e.g., π/4, π/2, 3π/4).
Key Results
The proposed PGC-PDC-EKF algorithm demonstrates significant performance improvements:
- Phase Delay Compensation: The PDC module accurately extracts the phase delay (θ) with a resolution of 0.01745 rad.
- Signal Quality:
- SINAD (Signal-to-Noise and Distortion Ratio): Achieved a high of 54.01 dB, which is 18.03 dB higher than the basic PGC-Arctan algorithm on average.
- THD (Total Harmonic Distortion): Reached a low of -62.28 dB, which is 26.04 dB lower than the PGC-Arctan algorithm.
- RAE (Relative Amplitude Error): Achieved a very low value of 0.0374%, indicating excellent agreement with the original signal.
- Linearity: The demodulation system exhibits excellent linearity, exceeding 99.99%.
- Robustness: The algorithm shows high stability and immunity to C value drift and phase delay effects, outperforming other compared algorithms in various performance metrics.
Limitations and Future Work
- Computational Complexity: While the PDC module is noted for its fast calculation, the EKF component, though effective, can introduce computational overhead. The paper mentions that a step size of π/180 rad for phase delay calculation offers a good balance between resolution and computation time, suggesting that optimization of computational resources is a consideration.
- Future Work: The paper concludes by highlighting the proposed scheme's "great potential for application in interferometric optical fiber sensors," implying that further real-world deployment and testing in various sensor applications would be a natural next step. The authors also suggest that the algorithm's stability and low harmonic distortion make it suitable for demanding sensing applications.
Summary updated