拉曼光纤随机激光结合无源传感单元可以实现超长距离的准分布式传感。然而，受限于光谱探测速度，该传感方案通常只适用于静态传感领域。针对该问题，将拉曼光纤随机激光与拍频光谱探测技术相结合，提出了一种新型的拉曼光纤随机激光长距离动态传感技术。首先，基于含时光谱稳态模型论证了光谱快速测量对长距离动态传感的适用性。随后，在原理性验证实验中通过处理本振光与光纤随机激光拍频后的时域信号，实现了对光纤随机激光光谱的快速测量，并突破了光波往返时间对传感带宽的限制。同时，利用去噪卷积神经网络对光谱的中心波长变化进行标定，大幅提高了扰动信号探测的信号质量，实现了对不同频率、不同波形的扰动信息的准确测量。该研究为进一步拓展光纤随机激光的应用领域提供了新的思路。

Raman random fiber laser combined with passive sensing units can realize quasi-distributed sensing over ultra-long distances. However, limited by the spectrum detection speed, this sensing scheme is usually only applicable to static sensing fields. To address this issue, a novel Raman random fiber laser long-distance dynamic sensing system is proposed by integrating Raman random fiber laser with beat frequency interrogation. Firstly, the suitability of rapid spectral measurement for long-distance dynamic sensing is demonstrated based on the time-dependent spectrum-balanced model. Then, in the proof-of-concept experiment, the spectrum of Raman random fiber laser can be measured rapidly by processing the temporal signals of the local oscillator light and the random fiber laser after beating, breaking through the limitation of the round-trip time of the light on the sensing bandwidth. Meanwhile, the center wavelength of the spectrum is calibrated by using a denoising convolutional neural network, which significantly improves the signal quality of disturbance signal detection and achieves accurate measurement of disturbance information with different frequencies and waveforms. This research provides new insights for further expanding the application fields of Raman random fiber laser.