Robustness- and Processing-Rate-Improved OFDR Based on Local Search and Kalman Prediction

Distributed optical fiber sensing based on optical frequency domain reflectometers is a promising technology to achieve high-density sensing information in harsh environments. Aiming at the major influence factor on robustness due to degradation of the similarity or correlation degree (SCD) between the reference and measurement spectra, in this letter, high-orders residual sampling error of nonlinearly swept laser source is found to cause random position deviations besides accumulated compensation based on thermal-/elastic-optic effect, and a local search and dynamic prediction method based on Kalman filter is proposed for position matching reference and measurement sensing gauges. Through this method, SCD can be improved when the accumulated compensation is invalid. Experiments indicate that a wide measurement range (highest temperature of ~ 450 °C and largest strain of
$\sim 10~000~\mu \varepsilon$
) under a spatial resolution of 5 mm, as well as high robustness and high accuracy, can be achieved along a 50-meter-long fiber under test. Besides, the computation amount can be reduced to 1/3.5 ~ 1/17.2. The advantages of this method include high robustness, wide measurement range, high processing rate, and excellent applicability with current OFDR systems.