Feature calibration-enabled fine spatial resolution temperature demodulation with RDTS system

The low-pass characteristics of traditional Raman distributed temperature sensing systems smooth out abrupt temperature changes within the small temperature variation region (TVR), thereby increasing temperature demodulation errors. Typically, this problem can be addressed by establishing a linear correlation linking the characteristic of the descending edge of the thermal curve and temperature. However, the selection of the slope segment and the sampling interval of the descending edge of the thermal curve has not yet been systematically studied, and temperature demodulation is currently based on a single known TVR length, which limits its scope of application. To address this issue, we verify through simulations and experiments that the 67%-34% slope segment derived by dividing the descending slope of the thermal curve into three equal parts, which can more accurately characterize temperature changes for different TVR lengths when used as a temperature calibration feature. Furthermore, the temperature-descending slope curves exhibit a good linear relationship across different TVR lengths. Experimental outcomes indicate that this method enables temperature measurement of 71.09 °C with an error of ±0.24 °C in a 0.4 m TVR. In addition, we further investigate the interference resistance and temperature demodulation accuracy under different calibration characteristics, providing clearer guidance for the application of thermal curve characteristics within engineering temperature retrieval.