A Signal Processing System Design For Interferometric Fiber-Optic Sensor

In recent years, sensor system is becoming more sensitive, adaptive and intelligent, especially, distributed and network sensor system perform a very important role. Fiber-optic sensor, as a new member of the sensor family, is now greatly popular because of its excellent performance. Comparing to other fiber sensor system, the interferometric fiber-optic sensor has the best sensitivity. This paper mainly concerns on a kind of fiber-optic sensor system which is based on the structure of Mach-Zehnder interferometer, and a signal processing algorithm is designed for this sensor system. The algorithm includes locating and event detecting function, which uses AR-model power spectrum estimation method. This algorithm, which implements real time alarm and accurate locating, makes this fiber-optic sensor valuable in communication cable and oil pipe protecting. The results of this paper provide a guide to design a long distance distributed fiber-optic sensor. This design has a potential application in industry and military field.This fiber-optic sensor system consists of four parts, which are sensor line, T/R module, data acquisition module and signal processing module. Signal processing module includes DSP card and PC. DSP performs event detection algorithm and PC implements locating algorithm. The experimental results imply that event detection can obviously reduce the error alarm rate of this system, and the locating error can be 100 meters by setting the sample rate properly.