Study On The Quasi-distributed Fiber Bragg Gratings Demodulation System For Temperature And Depth Measurement Under Water

Temperature and depth are the basic physical parameters of water environment. Because of the large amounts of the measurement points and serious condition in underwater environment, the traditional electrical and mechanical sensors are unable to meet the need of the long-term monitoring of the underwater application. Therefore, the optical fiber Bragg grating (FBG) is proposed to measure the temperature and depth underwater in this thesis. FBG is a kind of wavelength modulation sensor. The advantages of FBG sensor such as strong anti-interference, easily multiplexing make it very suitable for the temperature and depth detection in large underwater application. Based on the study of demodulation and multiplexing technologies at home and abroad, FBG demodulation system inculding hardware and software is designed in this thesis. FBG center wavelength demodulation algorithm and overlapping spectrum separation algorithm are also realized. Moreover, the experiments are also carried out to verify the the performances of the system.The core component of the designed system is tunable Fabry-Perot(F-P) filter, which is used to scan the FBG reflection spectrum. The demodulate system consists of optical circuit, electric circuit and softeware. The optical circuit includes broadband optical source, tunable Fabry-Perot filter, multi-wavelength reference etalon, photoelectric detector,3dB coupler and so on. According to the system requirements, the parameters of the optical components are carefully analyzed and selected to set up the optical circuit. In the optical circuit, the optical signal coming from the broadband optical source is transformed to narrow-band one with special center wavelength after going through the tunable F-P filter. The narrow signal is divided into two beams by a 3dB-coupler. Then two beams enter into the FBG sensing channel and reference channel respectively.The electric circuit can be divided into several parts. The core processor part mainly implements the work of digital signal processing and system controlling by the DSP F28335. The F-P filter's drive voltage part includes the peripheral circuit and the amplification circuit of D/A chip LTC1655, which produce saw-tooth voltage to change the transmitted central wavelength of the filter. The photoelectric conversion part selects the InGaAs photodiode as receiver of the reflection spectrum for both the sensing channel and reference channel, which converts the received optical signal to voltage signal. The AD converter part converts the voltage signal to digital one which can be processed by the DSP. The communication part adopts the RS232 serial port to realize the communication function between PC and DSP.The software part includes the program based on the core processor DSP F28335 and demodulation system of upper computer. In CCS development environment for DSP platform, several programs are developed as follows. First, the DSP can realize the periodic driving voltage output by configurating the necessary parameters of SPI. Then the AD converter program is completed to transform the analog signals of both FBG sensors and multi-wavelength reference etalon into digital signals. Finally, the serial port program is designed to help the DSP communicate with the upper computer. The demodulation system of upper computer is programed by Lab VIEW language. The LabVIEW program uses the VISA module to communicate with DSP. The demodualtion system in PC has such functions as configuring the parameters of the FBG sensors, displaying the diagrams and values of the temperature and the depth underwater in real-time and storing the datas of FBG wavelength.FBG center wavelength demodulation algorithm and overlapping spectrum separation algorithm are studied. First, four typical peak search algorithms are compared on the Matlab. After camparing the accury and stability of the four methods, the Gaussian polynomial curve fitting is finally elected. Considering the requirements of real-time in quasi distributed demodulation system, the collection digital datas are reduced after digital filter. So the core of the center wavelength demodulation algorithm includes digital filter and Gaussian polynomial curve fitting which is already developped on the DSP platform and runs successfully. In order to detect the peaks in the overlapping spectrum in wavelength division multiplexing(WDM) system, an algorithm called maximum spectrum of continuous wavelet transform(MSCWT) is first used in separating the overlapped spectrum of the FBG. The influences of the interval gap between two FBGs wavelength,3dB-bandwidth and optical power of the reflected spectrum are discussed. The proposed algorithm makes it possible to reduce the gap of the adjacent FBGs in the serial sensor array. In this way, the number of sensors in WDM-FBG network can be greatly increased under the same source light bandwidth.In order to verify the function of the designed system, a series of tests are conducted for each module such as the driving voltage, photoelectric detection, AD data acquisition and the multi-wavelength references. After finishing the tests of each module, a complete FBG demodulation system is set up to measure the temperature and depth underwater. The relationship between the center wavelength and temperature or depth is calibrated. The static and dynamic experiments of FBG sensors are completed through the demodulation system. According to the above experiments, it shows that the demodulation system can detect the center wavelength of FBG sensor with the stability of 5 pm. Besides, the Lab VIEW program can display the results of FBG sensors of both center wavelength and measured parameters of temperature and depth. The measurement range of the temperature and depth underwater is decided by the different packages of FBG sensors. In addition, the overlapping spectrum acquisition system is also built. The actual overlapping spectra are processed to verify the feasibility of the MSCWT algorithm. The simulation and experimental results show that the MSCWT algorithm has the abilty to separate the overlapping FBG sensor signal. The optimal separation accuracy of the actual overlapping spectrum is 1.2pm.In general, this thesis achieves the principal goal. The quasi distributed demodulation system based on DSP is finished and tested, which has the accuracy of 5 pm for single independent spectrum. In order to increase the number of FBG sensors, MSCWT algorithm is developed to demodulate the central wavelengths from the overlapped spectrum in the serial FBG system.