Research On Performance Improvement Methods Of Fiber Raman Distributed Temperature Sensor System

Raman-based distributed temperature optical fiber sensor (RDTS) has the characteristics of large temperature detection range, powerful anti-electromagnetic interference capability, real-time detection of temperature distribution, simple structure and low cost. It has been widely used in the areas of temperature monitoring, such as electricity, gas pipelines, mines, atmosphere, aerospace and other fields of real-time temperature detection. Because intensity of spontaneous Raman scattering in a fiber is usually very low, and fiber loss is unavoidable, the Raman signal at the rear end of the fiber in a long-distance temperature sensing system is very weak. In view of the existing problems, the method of improving the performance of fiber Raman distributed temperature sensor system is researched, mainly including the following two aspects.Firstly, a fiber Raman temperature measurement system supported by Raman amplification is proposed in this work. The measurement system employs a backward Raman pump to amplify the laser pulse, which compensates loss of the laser pulse in a long distance transmission and consequently improves the signal-to-noise ratio of backward Raman scattering light. The results of numerical simulations show that resolution of temperature measurement of the proposed system is more than six times higher than that of a conventional system under the same condition. Alternatively, the distance of temperature measurement improves almost two times compared with that of a conventional system for the same temperature resolution.Secondly, the linear cumulative average algorithm, wavelet transform algorithm and empirical mode decomposition algorithm is used for the denoising of the temperature measurement system. In view of the shortcomings of long time and imperfect denoising effect of traditional empirical mode decomposition denoising method, an improved empirical mode decomposition denoising method combined with wavelet transform is proposed. The results show that the denoising effect of the improved denoising method on the system is improved, and the time has been shortened.