The Research On Preprocessed Data Of Disciplined Clocks In GPS

As an important measurement of physical quantity in modern science and technology, time frequency is one of the important foundations and plays as a key role in the developing of science and technology. Especially with the rapid advancement of science, the high precise time frequency has been widely used from basic science research to engineering technology, such as science exploration of deep space and sea, electrical power system, navigation and location and so on. Satellite positioning and timing system can provide high precise time signal with long-distance transmission, which has the capability of timing, positioning and velocity measurement with high precision. The realization of high precise time synchronization is subject to more precise time frequency. As the most widely used positioning and timing system, GPS can provide the required high precise time frequency, and disciplined clock is a great realization means.Disciplined clock uses GPS standard time signal to lock down the local crystal oscillator and makes it output high precise frequency signal. This technology not only keeps better stability for local oscillator in a short time, but also makes the local controlled crystal oscillator effectively replicate the GPS standard time signal with long-time stability and precision. The conventional disciplined clock employs simulation or partial digital system which is high-cost, easy to be affected by environment and difficult to build noise model and debug. Based on these existing defects, this text puts forward and implements a disciplined clock system using the technologies of FPGA and ARM and put emphasis on estimation and preprocess of the clock correction in disciplined clocks in order to enhance the of time synchronization precision.The clock correction mainly includes systemic change and random change. The systemic change can be expressed by the initial time deviation, the frequency deviation and the aging while the random change can be expressed by individual power law noise spectrum in five kinds. This paper has studied and compared the results of Vondrak algorithm and Kalman filtering algorithm in the data preprocess of disciplined lock. After comparing the data preprocess algorithms, choosing the better one and stimulating Matlab, Vondrak filter technique has fused the crystal oscillator noise model with phase-locked loop structure. Eventually at the performance evaluation, the 1pps output of the optimized and filtering controlled GPS disciplined clock system has been measured as well as its frequency accuracy and stability and the result indicates that GPS disciplined clock can improve the accuracy and the stability of the local crystal oscillator to some extent.This paper has studied about the design optimization of the hardware in the disciplined lock system and the data preprocess of clock correction and it also takes the constraint condition of practical engineering into consideration. Therefore, the research result of this paper can provide some theoretical and practical engineering reference for disciplined clock technology.