Search

A new prediction algorithm based on Empirical model decomposition (EMD) and Support vector machine (SVM) is put forward in this paper, and this algorithm solves the problem of the hydrogen atomic clock differences prediction, which is affected by the non-linearity and non-stability. The clock differences were decomposed into Intrinsic mode functions (IMF) and the residual series. The suitable kernel function and parameters were chosen to build the different SVM for predicting each IMF and the residual series. Each prediction result was summed to obtain the clock differences prediction. Results show that the EMD-SVM algorithm is effective compared with the linear regression and single SVM . The relative prediction error is reduced from 0.4327% to 0.2371%, and the dispersion is less than other methods.

In this paper, a new traceable method for measuring RF AM signal is presented. The 50GHz broadband sampling oscilloscope which is the national pulse parameter reference is used to measure the RF AM signal. A waveform measurement system for radio frequency amplitude modulation signal is constructed. After correcting the time-based jitter error of the collected waveform, the phase, amplitude and frequency parameters of radio frequency amplitude modulation signal are accurately restored by the co-directional orthogonal method and the least square method. The experimental results show that the method of using broadband sampling oscilloscope to measure the RF amplitude modulation signal and digitizing demodulation signal has excellent results, which achieves the accurate recovery of waveform parameters, and provides a new idea for traceability measurement of time domain RF modulation signal.

Time-base error is a key factor affecting the accuracy of time-domain measurement of bandwidth sampling oscilloscopes. Therefore, it is necessary to compensate the time-base error of measurement results, and it is also very important to evaluate the uncertainty of time-base error compensation of measurement results. In this paper, the estimation and compensation algorithm of time-base error is analyzed firstly. Then the model of calculating the uncertainty of every point of the measurement waveform which have compensated the time-base error is constructed based on the covariance matrix. Finally, the uncertainty of each point of the high-speed pulse signal after time-base error compensation is calculated, and the maximum voltage value of the uncertainty components of every point of the measurement waveform is about 2. 3mV. The maximum voltage value is calculated as the uncertainty of the rise time of the measured signal, which is about 0. 24ps. The result indicating that the method of evaluation the uncertainty of the measurement waveform by the point-by-point is feasible and reasonable.

The atomic clock prediction is the key step for constructing atomic clock time scales and implementing atomic clocks steering, and its prediction level directly affects the accuracy and stability of atomic clock time scales. Aiming at the non-linear and non-stationary characteristics of the hydrogen atom clock, the least squares support vector machine (LSSVM) hydrogen clock prediction algorithm is proposed, which is verified by the actual hydrogen atomic clock data of the timekeeping laboratory in National Institute of Metrology, China. The results show that compared with the linear prediction algorithm and the support vector machine (SVM) prediction algorithm, the least squares SVM prediction algorithm improves the accuracy of the clock prediction, and its root mean square error is, respectively, reduced by 50% and 29%.

The selection of timescale algorithm will have a great influence on the accuracy and stability of timescale. To improve the accuracy and stability of timescale, this paper studied the stochastic model of the hydrogen clock and Hadamard deviation for researching the Kalman timescale algorithm based on the hydrogen clock ensemble, which is quite different from the traditional Kalman timescale algorithm based on the model of cesium-beam clock ensemble, mixed ensemble and Allen deviation. The Hadamard variance is used to describe the stability of clock frequency in this paper; it will improve the precision of the Hydrogen Clock Ensemble. By using Hadamard variance instead of Allen avariance, it can overcome the influence of frequency drift. In this paper, the clock whose stability is best is chosen as the reference clock so that the accuracy and stability of the hydrogen clock ensemble will be improved. The experiment data of the hydrogen clock ensemble in the timekeeping laboratory of National Institute of Metrology (NIM) in China were processed; the frequency stability can be improved to 10'15. The results verified that the Kalman timescale algorithm based on the hydrogen clock ensemble is effective.

The optical pulse standard method is based on the traditional pulse standard method. This method is that converting the optical pulse into electrical pulse through the detector and then calibrating the broadband sampling oscilloscope. In the paper, according to the optical pulse standard method, a method of evaluating the uncertainty of calibrated bandwidth of oscilloscope is proposed. Firstly, the influencing factors of measurement uncertainty are analyzed and the calculation formula of standard uncertainty is derivated; Then the transfer algorithm of uncertainty is studied; Finally, the composite uncertainty of calibrated bandwidth is evaluated. The results show that the method is scientific and reasonable.

The time base drift is a key factor that influences the sampling oscilloscope to measure the high speed pulse signal precisely, so time base drift must be removed as far as possible. Cross-correlation algorithm is an alternative method to remove time base drift effectively. First, this paper does computer simulation, then use the algorithm to do research on the original data that is collected when establishing national pulse parameters primary standard. Both the results of simulation and the research show that the cross-correlation algorithm can remove the time drift effectively.

This paper studied two methods to remove the influence of the microwave reflection in the time domain measurement data of the characteristic impedance of transmission line. Usually, people choose the best measurement region according to the IPC procedure to remove the influence of microwave reflection. Although this method is simple and effective, it inevitably removes plenty of effective measurement data. So it results in the accuracy of the characteristic impedance measurement is not high enough. According to the basic theory of measurement and the feature of measurement waveform, this paper proposed the threshold method of amplitude and mode method of amplitude to remove the influence of the microwave reflection. Analysis and calculation results showed that the two methods could effectively remove the influence of microwave reflection and improve the accuracy of calculation results of the characteristic impedance of transmission line.

The atomic clock device is complicated, and its operation is influenced by a lot of factors, so the date has certain randomness. Predicting atomic clock data by appropriate algorithm is the important step of atomic clock data processing. This article chooses SVM algorithm to predict clock differences, and processed the real atomic clock data, compared relative error with the prediction results and the results of linear regression algorithm, the prediction algorithm based on SVM is effective.

In order to improve the accuracy of measurement the correction of time base error of oscilloscope is essential. In this paper we studied time base error. There are three kinds of time base error including time base drift, time base jitter and time base distortion. Reason of time base error causes and the influence for measurement and calibration, the method for correction or compensation were studied. We also had good results as we process the experiment data which proofed the method in this paper were right.

Atomic clock difference (errors) contains various kinds of noise and it can be characterized to as non-stationary. Hilbert-Huang Transform (HHT), a new adaptive method based on empirical mode decomposition (EMD) and Hilbert spectral analysis, was used to analyze and process atomic clock difference. Taking real data for discussion, the results show that HHT method is practical and versatile to handle noisy data and characterize the clock behavior.

The national standard of transition duration of pulse parameter in China is 7.25 ± 0.40 ps. The standard is not only a pulse signal source standard with the known information about time-domain characters and transition duration, but also a oscilloscope standard with the known impulse response of the oscilloscope. The nominal value of transition duration of Agilent 4016 minus step pulse source is 5ps, faster than the national standard. So Agilent 4016 could not be calibrated by the national standard according to traditional calibration method. The new calibration method has been studied in the paper by using the national standard. Devolution algorithm was also used in measurement data processing. Experiment, data processing and uncertainty estimation were done and calibration of Agilent 4016 by national standard was come true.

GPS common-view is the main method for the time and frequency remote transfer originally, of which the uncertainty can reach several nanoseconds. This paper presents a procedure to simplify the original common-view time transfer system constructed using a powerful dual-frequency multi-channel EURO-160 GPS receiver and a high precision time interval counter SR-620. The relevant data-processing software is designed according to "Technical directives for standardization of GPS time receiver software". A preliminary zero-baseline experiment shows that the new system can reach almost the same accuracy as that ensured by the original system. The structure of the new system is simple, practical and reliable, the simplification techniques for common-view transfer system can provide better services and more choices for the time and frequency remote transfer.

This paper mainly researches atomic time algorithm, and its influence on atomic time scale level. With the improving of atomic clock level internationally, atomic time scale system is continuously promoted. Different atomic time scale system needs different algorithm. An atomic time algorithm should be applied in fit time scale system. ALGOS is adopted by BIPM to calculate TAI. AT1 algorithm and an algorithm based on new TA(NIM) system whose H-maser is steered by Cs fountain clock are mainly introduced in this paper according to the improvement of atomic time scale in NIM. A specific atomic time scale should adopt suitable algorithm, only in this way, we can get a better atomic time scale.

The settings and effects of exponentially filter in the Atomic Time 1(AT1) scale algorithm is researched in this paper. The different effects of time scale algorithm with an exponentially filter or no exponentially filter is discussed. Simulations verified that the exponentially filter makes the result better. Then how to choose the exponential frequency-weighting time constant is discussed. This parameter is determined from the relative levels of white noise and random walk (or flicker) for clock. Different numerical value gives different results. Compared with a series of numerical values, the optimal value of each clock is given in this paper. On the basis of these works, exponentially filter optimized the AT1 time scale algorithm by estimating of the current average frequency of clock. Waveform figures are given and some valuable conclusions are drawn.