An airborne CH sensor with temperature compensation based on a miniature optical structure for natural gas pipeline leakage analysis.

Pipeline transport plays an important role in the exploration and transmission of natural gas. An airborne near-infrared (NIR) methane (CH) sensor has been developed to analyze CH leakage in the pipeline transport process based on the tunable diode laser absorption spectroscopy combined with wavelength modulation spectroscopy (WMS-TDLAS) technology relying on the high efficiency and convenience of unmanned aerial vehicles (UAVs). A laser diode in the form of an ultra-compact 8-pin package with a center wavelength of 1653.7 nm and an independently-designed miniature optical structure with an optical path length of 36 cm are adopted, which greatly reduces the size of the sensor. In order to improve the measurement accuracy, a temperature compensation link is designed to minimize the influence of temperature changes on the measured value. The original second harmonic signal is denoised by the variational mode decomposition algorithm optimized by Genghis Khan shark optimizer algorithm (GKSO-VMD), and the Bi-directional long short-term memory GKSO-BiLSTM algorithm is used for CH concentration inversion. The measured and real concentration values of the sensor are closely related, with a correlation coefficient (R) of 0.9996. The response time experiment demonstrates that the response time is approximately 6 s. The great stability of the sensor is confirmed by the long-term stability experiment. According to the Allan-Werle deviation, it is estimated that the sensor has a limit of detection (LoD) 613 ppb with an integration time of 230 s. Therefore, an effective analysis method is proposed for CH leakage analysis during natural gas pipeline transport.