•  
  •  
 

Keywords

power clock synchronization system;Kalman filter;random error elimination;outlier detection;error weighting correction

Abstract

The frequency source of the power clock synchronization system is affected by factors such as changes in the external environment or self‑aging, and there will be frequency drift problems, which can not maintain long‑term stability. A low‑cost crystal oscillator frequency taming calibration scheme based on Beidou pulse per‑second signal and improved Kalman filter algorithm is proposed. Firstly, take the Beidou second pulse signal as the time reference and use Kalman filter to eliminate the random error contained in the reference; Secondly, in view of the divergence of Kalman filtering prediction results and large errors, the RBP neural network is used to correct the filtering errors online in real time to improve the filtering performance; At the same time, aiming at the problem of jumping outliers in Beidou pulse per‑second, an outlier detection and weighted correction method is proposed based on innovation change rate. Experimental results based on simulation data show that the proposed method can effectively improve the frequency accuracy and stability of the frequency source of the power clock synchronization system, and has strong adaptability to jump outliers.

DOI

10.19781/j.issn.1673-9140.2023.02.026

First Page

232

Last Page

239

Download

Share

COinS