| It is necessary for aeronautics, petroleum artesian well, robot control, GPS navigation systems to calculate position and deduce orientation of objectives. Gyro is an important part in these systems, which affects the precision directly. Fiber optic gyro (FOG) is a new technology developed in recent 20 years. Compared with mechanical gyro, it has no rotating highly components and it is insensitive to acceleration. Compared with laser gyro, it has no both high power supply and mechanical vibrating elements. At present the global market occupational rate has been gained up to 50 percent by FOG, which holds small volume and lasts long, weighs lowly and consumes lowly. Further more, it has the prosperous prospect of being applied in the guiding and navigation system from low precision to moderate, high precision. In this paper, digital signal process is carried out to improve its performance oriented to FOG's characteristic. Adaptive filter theory, method and advantages are expatiated based on the analysis of FOG principle, random signal process and statistical mathematics. Then the main algorithms of adaptive filter – LMS (Least Mean Square) and RLS (Recursive Least Squares) are analyzed in detail. In the end, software and hardware design for VG941 FOG are put forward.For hardware design, selecting appropriate chip and designing schematic, functions and notices of all parts are described in detail, especially for DSP and FPGA technologies. For software design, analyzing and preprocessing dynamic data of FOG firstly, then FOG data is emulated via LMS and RLS adaptive filtering. Results indicate adaptive filtering is effectual for FOG, and error reduces obviously. By comparison, LMS algorithm is chosen because of its simple and effect. Program by C language and assemble language bases on DSP. Result achieves the intension on aspects of real-time and precision. At last, this paper discusses some problems in course of system debugging and points out the next step of the work.
|
PDF Full Text Request