使用Kalman滤波器调整预测值的时间尺度算法
发布时间：2019-04-12 14:31
【摘要】：时间尺度算法用于综合钟组内所有的原子钟,建立一个频率稳定度更高的时间尺度,其核心环节可以概括为通过N-1组观测量(钟差)计算得到N台钟的权重和预测值。传统算法主要关注如何调整权重来提高时间尺度的稳定度,本文算法通过调整预测值来提高时间尺度的稳定度。本文算法使用Kalman滤波器对观测钟差进行状态估计,在Kalman滤波器每一次递推的过程中,调整一次预测值,通过每次实时调整预测值来建立时间尺度。理论推导和仿真实验都表明,本文建立的时间尺度滤除了频率白噪声,主要只含有频率随机游走噪声,所以具有很高的中短期稳定度。该时间尺度是一个连续、实时、可预测的时间尺度。
[Abstract]:The time scale algorithm is used to synthesize all the atomic clocks in the clock group and establish a time scale with higher frequency stability. The core link of the algorithm can be summarized as the calculation of the weight and prediction value of the N clock by the measurement (clock error) of the Nm 1 set of observations. The traditional algorithm mainly focuses on how to adjust the weight to improve the stability of the time scale. In this paper, we adjust the predictive value to improve the stability of the time scale. In this paper, the Kalman filter is used to estimate the state of the observed clock error. In each recursive process of the Kalman filter, the predicted value is adjusted once, and the time scale is established by adjusting the predicted value each time in real time. Theoretical derivation and simulation experiments show that the time scale filter proposed in this paper contains only frequency random walk noise, so it has high short-term and medium-term stability. The time scale is a continuous, real-time, predictable time scale.
【作者单位】： 国防科学技术大学电子科学与工程学院;
【分类号】：P228.4
，
本文编号：2457101
[Abstract]:The time scale algorithm is used to synthesize all the atomic clocks in the clock group and establish a time scale with higher frequency stability. The core link of the algorithm can be summarized as the calculation of the weight and prediction value of the N clock by the measurement (clock error) of the Nm 1 set of observations. The traditional algorithm mainly focuses on how to adjust the weight to improve the stability of the time scale. In this paper, we adjust the predictive value to improve the stability of the time scale. In this paper, the Kalman filter is used to estimate the state of the observed clock error. In each recursive process of the Kalman filter, the predicted value is adjusted once, and the time scale is established by adjusting the predicted value each time in real time. Theoretical derivation and simulation experiments show that the time scale filter proposed in this paper contains only frequency random walk noise, so it has high short-term and medium-term stability. The time scale is a continuous, real-time, predictable time scale.
【作者单位】： 国防科学技术大学电子科学与工程学院;
【分类号】：P228.4
，
本文编号：2457101
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2457101.html
