Research On Fiber Bragg Grating Pressure Sensor Of Temperature Soft Compensation Based On Neural Network

Pressure is an important parameter for the operating state and safety assessment of the pipeline.Sensor for measuring liquid and gas pressure has become one of the most commonly used sensors in industrial production, is widely used in measuring the pressure of water conservancy and hydropower, railway transportation, intelligent buildings, production automatic control, aerospace industry, military project, petrochemical, oil, electric power, shipbuilding, machine tools and many other industries and fields. Pressure sensor based on fiber Bragg grating sensing technology compared with traditional mechanical and electromagnetic pressure sensors has many advantages,such as strong ability to resist electromagnetic interference, safety explosion-proof, corrosion resistance, suitable for long distance transmission, easy to reuse and so on. In the field of petroleum chemical industry, water conservancy, machinery etc., Optical fiber grating pressure sensor has a wide application prospect in the aspects of the pressure's multipoint distribution measurement and remote online monitoring. To solve the problem of cross sensitivity of strain / temperature, is an important prerequisite for fiber Bragg grating pressure sensor applied to practical engineering.Based on pressure measurement system, this paper makes fiber Bragg grating pressure sensor and uses the BP neural network to realize the temperature and pressure decoupling, solves the problem of pressure / temperature cross sensitivity problem to meet the demand of temperature soft compensation. The main research work is as follows.(1)Analyze and summarize several structure of fiber Bragg grating pressure sensor. At the request of high sensitivity and low repeatability measurement error, diaphragm structure is selected as the design structure of the fiber Bragg grating pressure sensor in this paper.(2)Research the existing temperature compensation scheme in fiber Bragg grating pressure measurement. Based on the BP neural network simulating real MIMO system, pressure / temperature decoupling method based on BP neural network is used. The system is designed to get the operating temperature and pressure, solve the problem of pressure / temperature cross sensitivity.(3)Product fiber Bragg grating pressure sensor, and conduct pressure tests and temperature experiments on the sensor. Improve the packaging process, repeat the tests until the test results meet the requirements of the sensor.(4) Based on the finite element analysis of the pressure sensing components of the fiber Bragg grating sensor, and use of MATLAB to establish pressure / temperature decoupling BP neural network. Confirm that the soft temperature compensation method is effective.