Research On Temperature Error Measurement And Compensation Method Of Fiber Optic Gyroscope

As a kind of angular rate sensor applied in inertial navigation system, fiber optic gyroscope (FOG) possesses advantages such as no rotating components, short starting time,low cost, long life time, large dynamic range and etc. So it has been widely used in military as well as civil fields. As the core device of the FOG, the fiber coil is extremely sensitive to temperature variation of environment. Thus, the test and compensation work on temperature error of the fiber optic gyroscope (FOG) is quite important. This paper mainly study on two aspects: test and compensation on the bias and scale factor error of the FOG which caused by temperature variation, improve its temperature performance to meet the demands of more requirements in application. The main work of this paper is as follow:At first, the development history and current temperature compensation methods of the FOG both at home and abroad was expounded. Then the basic theory and the error mechanism of the FOG was explained. This part mainly analyzed the internal and external causes of bias and scale factor error, especially explained the influence caused by temperature change.The research contents was divided into two part: the test and compensation on bias error of the FOG and the test and compensation on scale factor error of the FOGIn the research on bias part, firstly, the common test method was introduced, aiming at its shortcoming which cannot show the relation between the output of the FOG and the changing rate of the temperature,a method using variable changing rates of temperature was presented and used in the experiment. Through analysis and calculation, a model was set to explain the relation between the standard deviation and changing rates of temperature during heating period, and the model was used as an evaluation index to describe the temperature error of the FOG and temperature changing rate. Then common compensation method (least square method-LSM) was introduced and used to compensate the output of the FOG To contrast which is better between linear and nonlinear compensation methods, a BP neural network interface was compiled to compensate the output of the FOG. The compensation results showed that BP neural network compensation method is better than the least square method.In the research on scale factor part, firstly, the common test method was introduced,aiming at its shortcoming that is too tedious and cannot show the continuous output change when the temperature keep changing, a full temperature scale factor test method was presented. Then the method was used to test the FOG,after processing the data, the relevant evaluation index was determined. After that, the scale factor of the FOG was compensated by least square method, contraposing its deficiency that has constant order and cannot realize on-line compensation, the fraction-tap least mean square (FLMS) method was presented and used in the experiment. The result showed that, the FLMS method is better than the LSM method, and it can evaluate the performance of the FOG better.