Research And Implementation Of Distributed Optical Fiber Temperature Sensing System

Temperature detection is an important part in the field of industrial control,especially in the power and coal chemical industry which is requiredto be able to detect in time.Usingthe traditional electric temperature sensor the multipoint distribution and long distance one-time measurement can not be realized,alsoin any other bad environment.However,distributed optical fiber temperature sensor has the unique characteristics of continuous distributed sensing and optical fiber's essential security.The obvious performance advantage makes it have a wide range of applications,and has become a hot research field of optical fiber sensing technology.Based on these the distributed Raman optical fiber temperature sensing system and its key technology have been studiesdeeply here.In addition,the interaction parameters of the systemhas beenanalyzed and calculated to improve the performance of the system.In order to decrease the temperature measurement errorcaused by weak Raman scattering signal and strong noise,a new filter algorithm has been designedto improve the temperature measurement accuracy of the system.Firstly,in this thesis,two kinds of optical fiber temperature measurement techniques have been introduced.Comparing with the quasi distributed optical fiber temperature measurement of Bragg gratings,the advantages of distributed optical fiber temperature sensing in temperature measurement are clearly defined.According to the relation between physical quantity and temperature in theoretical formula,the Raman scattering light is selected to calculate the temperature and Rayleigh scattering is used to determine the location on the theory of three types of scattering in optical fiber.The advantages and disadvantages of the three methods of demodulation temperature have been calculated and analyzed,and then the best method has been chosen to design the system.Secondly,in order to improve the technical performance of the distributed optical fiber temperature sensing system,the hardware parameter matching has beenconsidered and a reasonable and effective system hardware scheme has beenconstructed based on the operation characteristics of the module and the influence of the system performance parameters.Thenthe source of noise in the system hardware has been analyzed in detail,and the practicalsoftware filtering algorithm has beendesigned to improve the temperature accuracy of the system.Thirdly,a series of simulation and comparison experiments have been carried outby Matlab.The experiments haveshown that the modulus maxima wavelet filtering algorithm based on signal correlation and fusion adaptive decomposition scale has good results in reliability,real-time and accuracy.It can reduce the noise and greatly improve the signal to noise ratio.Finally,the software filtering algorithm has been applied to the designedtesting platform.The experiments including measurement accuracy,spatial resolution and response time have been completed.A lot of experimental results haveverified the effectiveness,reliability and practicability of the filtering algorithm and its improving methods in this thesis.