Noise Analysis And Optimization Of Distributed Raman Fiber Temperature Sensor

With the development of fiber technology, distributed fiber sensors have attracted much attention due to its superior performance and huge potential in recent years. In distributed fiber sensors, fiber is both the transmission medium and the sensing medium. It can transmit the optical signal for long distance, and the spatial distribution of temperature, strain or humidity along the fiber could be sensed by detecting the changes of characteristic parameters of optical signal. The distributed Raman fiber temperature sensor (RDTS) system is based on photoelectronic technology, computer technology and weak signal detection technology, and it could be applied in hazardous conditions owing to its anti-electromagnetic interference, good insulativity and high sensitivity. The Raman backscattering is extremely weak and its useful information is buried in the system noise. The signal collection and processing are of great difficulty. Thus it is of particular importance for analyzing the source and properties of system noise and adopting pointed measures to improve the signal-to-noise ratio and optimize the system performance.In this thesis, the theoretical analysis and experiment research of RDTS system have been made. The thesis’s main contents and innovation points are as follows:1. The optical time domain reflectometry (OTDR) technique and typical distributed fiber sensors are introduced. The Raman scattering is explained from traditional electromagnetic theory and quantum theory respectively, and the methods of temperature demodulation are analyzed.2. The construction of RDTS system and the selection principles of its main devices are introduced, the technical characteristics (such as precision, temperature resolution, spatial resolution etc.) and its influence factors are analyzed, and the signal-to-noise ratio and system stability are studied deeply.3. The necessity of denoising is analyzed, and the various kinds of noise existed in the system is discussed in detail. The noise mainly includes the photoelectric detector noise (such as dark current noise, shot noise, thermal noise, additional noise etc.) and the Rayleigh noise. The reason of Rayleigh noise is first analyzed and its impact on temperature measuring precision is theoretically simulated.4. A novel algorithm of eliminating the Rayleigh noise impact is proposed. Several temperature experiments are presented to validate the effectiveness of the proposed algorithm, and the experiment results indicates that the temperature measuring error could be reduced from about7%to less than2%. Based on the improvement of eliminating Rayleigh noise, the wavelet transform modulus maxima filter method is adopted to further eliminate the noise fluctuation and optimize the signal-to-noise ratio.