RU2279115C9 - Method for storing frequency of electric oscillations - Google Patents

Abstract

FIELD: electric engineering, possible use when operating frequency measures.

SUBSTANCE: method includes measuring starting deviation of frequency of frequency measure from standard frequency and at some starting range T₀ beginning speed of systematic shift of frequency of measure relatively to frequency of standard is measured. At the end of time range T₀ frequency of measure is corrected for value of measured starting deviation and further during some time range T additionally periodically frequency of measure is connected in time ranges Δt. Value β of additional correction during time range T is decreased in accordance to predicted decrease of measure frequency shift speed.

EFFECT: possible decrease of systematic frequency shift of frequency measures.

2 dwg

Description

The invention relates to the field of metrology and can be used in the operation of frequency measures (sometimes we will call them simply measures).

There is a method of storing the frequency of electrical oscillations with a frequency measure [1], containing periodically repeated operations of comparing the frequencies of the measure and the standard (ie, measuring the deviation of the frequency of the measure from the frequency of the standard) and correcting the frequency of the measure by the value of the measured deviation.

The deviation of the measure frequency from the frequency of the spatially remote reference is measured by radio, television or satellite signals or by the method of the transported clock, and from the frequency of the reference, located in close proximity to the measure, using frequency or phase comparators and electronically counted frequency meters.

This method of storing the oscillation frequency has a large error, which is determined by the systematic drift of the measure frequency.

на величинуTo reduce this error allows the method [2], containing the operation of periodic (at time intervals T _i ) measuring the frequency deviation of the measure from the standard frequency, correcting the frequency of the measure by the value of the measured deviation and additional periodic correction of the frequency of the measure throughout each time interval T _i , starting with T ₂ at time points

by the amount

(in this case β ₁ = 0);

where Δf _i-1 is the deviation of the measure frequency from the frequency of the standard, measured at the end (or throughout) of the interval T _i-1 .

, при этом прогнозируемое изменение частоты предполагается линейным.The amount of additional periodic correction is determined based on the predicted rate of systematic frequency drift of the measure, and the predicted rate of systematic frequency drift on the time interval T _{i is} assumed to be equal to (or less) the measured speed of systematic frequency drift on the previous time interval T _i-1 , which is equal to

while the predicted change in frequency is assumed to be linear.

The method described in [2] is taken as a prototype.

The disadvantage of the prototype is also a large error in frequency and the speed of the systematic drift of the frequency of the measure, because a linear forecast of the change in the frequency of the measure and, accordingly, an additional periodic correction of the frequency of the measure corresponding to this linear forecast, can reduce the rate of systematic drift of the frequency and the error of the measure in frequency by almost no more than (2 ÷ 3) times. This is because the rate of systematic frequency drift of the frequency measures is in fact not constant, but decreases with time.

The present invention is aimed at reducing the error in frequency and speed of systematic frequency drift of frequency measures.

Storage of the frequency of electrical vibrations is considered on a certain time interval of duration T. If you need to store the frequency on an interval of time greater than T, all operations of the frequency storage method are repeated.

дополнительную периодическую коррекцию частоты меры выполняют на величину β, которую уменьшают со временем на протяжении интервала времени Т в соответствии с прогнозируемым уменьшением скорости систематического дрейфа частоты меры, причем начальную величину дополнительной периодической коррекции устанавливают в соответствии с неравенством 0<|β_нач|<2|ν₀|Δt, а ее с знак - противоположным знаку ν₀. Максимальный положительный эффект достигается при β_нач=-ν₀Δt.The essence of the invention lies in the fact that in a method for storing the frequency of electrical oscillations with a frequency measure, comprising the operations of comparing the frequency of the measure with the frequency of the standard during the initial time interval T ₀ , determining, by the results of these comparisons, the initial deviation of the frequency of the measure from the frequency of the standard Δf ₀ and the initial speed systematic drift of the measure frequency relative to the standard frequency ν ₀ , correction of the measure frequency by the value of the measured initial deviation Δf ₀ at the end of the time interval T ₀ and subsequent on during the time interval T additional periodic correction of the frequency of the measure at time

additional periodic correction of the frequency of the measure is performed by β, which decreases over time over the time interval T in accordance with the predicted decrease in the rate of systematic drift of the frequency of the measure, and the initial value of the additional periodic correction is set in accordance with the inequality 0 <| β _beginning | <2 | ν ₀ | Δt, and its sign is opposite to the sign ν ₀ . The maximum positive effect is achieved when β _beg = -ν ₀ Δt.

The reduction of the error of the measure in frequency in the proposed method is achieved by more accurately predicting the behavior of the frequency of the measure in the time interval between its comparisons with the standard. If the prototype predicts a linear change in frequency, then in this method it is assumed to be non-linear, decreasing with time, which reflects the actual behavior of the frequency of the measure.

The forecast of the behavior of the measure frequency is based on measurements of its deviation from the standard frequency at the initial time interval T ₀ , as well as studies of the behavior of the frequency of measures of this type over long time intervals, much larger than T ₀ , which were carried out earlier. For example, a long (over several years) observation of the behavior of the frequency of quantum hydrogen measures showed that the rate of systematic frequency drift in them decreases with time with a time constant of ~ 1 year. Therefore, if in this time interval the change in the frequency of the measure is not predicted by a linear function, but, for example, by a function of the form e ^{-t / α} , where α = 1 year, and reduce the value of the additional periodic correction depending on time over the time interval T in accordance With this function, it is possible to obtain more accurate compensation of the systematic frequency drift of the measure over time and, accordingly, a lower frequency error and a lower residual rate of the systematic frequency drift of the measure. In this case, the maximum positive effect is achieved provided that β _nach = -ν ₀ Δt. For β _beg <ν ₀ Δt (in absolute value), the magnitude of the positive effect decreases, and for ν ₀ Δt <β _beg <2ν ₀ Δt, the rate of systematic frequency drift of the measure changes sign, but decreases in absolute value.

Moreover, as in the prototype, ν ₀ can be defined as

where Δf is the change in the frequency of the measure in the time interval T ₀ equal to the difference between the deviations of the frequency of the measure at the end and beginning of the interval T ₀ . However, it is possible to determine ν _{0 by} other methods, for example, by the least squares method, which is more accurate. For this comparison of the frequencies of the measure and the standard on the interval T ₀ spend not only at the beginning and end of this interval, but also at certain time intervals Δt, much smaller than T ₀ . After that, the speed of systematic frequency drift is calculated by the formula [3]

where Δf _k is the deviation of the measure frequency from the standard frequency measured on the k-th day,

n is the number of days during which measurements were taken.

Because the rate of systematic frequency drift of the measure frequency occurs very slowly, then the value of the additional periodic correction β can be changed not at each time interval Δt (usually Δt = 1 day), but at time intervals T _j much longer Δt (for example, T _j is 3 months )

The description of the proposed method is illustrated by figures 1 and 2.

In Fig.1 solid lines depict: 1 - graph of the frequency of the measure from time to time without correction; 2 - the dependence of the value of the additional periodic correction β on time by the proposed method; 3 - one of the implementations of the dependence of β on time by the proposed method; 4 is a graph of the frequency of the measure with correction by the proposed method, starting from time T ₀ . Dashed lines show: 5 - frequency correction value β according to the prototype; and 6 - graph of the frequency change of the measure during frequency storage according to the prototype.

Moreover, curves 4 and 6 do not show frequency changes due to additional periodic corrections that are carried out at time points kΔt (k = 1, 2, ...), due to the smallness of these changes.

Figure 2 shows the structural diagram of a device that implements the proposed method.

In figure 2 is indicated:

7 - a measure of frequency;

8 - standard (frequencies);

9 - frequency comparator;

10 - electronic counting frequency meter.

Figure 1 shows that the error in the frequency of the measure, which is obtained when using the proposed method, Δ ₀ f ₂ , is less than the error in frequency Δ ₀ f ₁ , obtained when using the method adopted for the prototype. The rate of systematic drift of the measure frequency also decreases.

One of the implementations of the proposed method for storing the frequency of electrical vibrations includes the following operations:

- measuring the deviation of the frequency of measure 7 from the frequency of reference 8 every day in the initial time interval T ₀ (10-30 days) using a frequency comparator 9 and an electronically counting frequency meter 10 and determining the rate of systematic drift of the frequency of the measure ν ₀ according to formula (3);

- correction of the measure frequency at time T ₀ by the value of the measured frequency deviation at this moment;

- additional periodic correction of the frequency of measure 7 every day for 3 months by β ₁ = -ν ₀ Δt (Δt = 1 day) (β ₁ in this case is the initial value of the additional periodic correction β _beg );

- correction of the frequency of measure 7 every day by the amount

over the next 3 months;

- correction of the frequency of measure 7 every day by the amount

over the next 3 months;

etc., frequency correction measures 7 every day by the amount

within 3 months following the (n-1) th three-month interval.

The first two operations are carried out at the factory or in the calibration laboratory; further operations are carried out at the consumer automatically without the availability of devices 8-10 using a microprocessor or microcomputer included in the frequency measure, or manually by the operator.

After the intertesting interval, i.e. the time interval over which the measure is verified by the metrological service (usually 1 year, but possibly 2, 3, 5 years), re-determine the magnitude of the systematic drift frequency of the measure frequency and the frequency error by comparing its frequency with the standard frequency and a new cycle of corrections by this method with refined values of ν ₀ and α.

The new value ν ₀ is determined on the basis of comparisons of the measure with the standard on a new initial time interval T ₀ , and the new value α is determined from previous measurements as the time interval over which the rate of systematic drift of the frequency of the measure ν ₀ decreased e times.

If the rate of systematic frequency drift of the measure over a time interval T decreases to an acceptable, sufficiently small value, then further additional periodic frequency correction can not be performed.

Using this method of storing the frequency of electrical oscillations allows to reduce the error in frequency and the speed of systematic drift of the frequency of frequency measures by (2 ÷ 3) times in comparison with known methods.

Literature

1. A.I. Pikhtelev et al. Frequency and time standards based on quantum generators and discriminators. - M .: Owls. Radio 1978, p. 252-262.

2. V.A. Logachev, A.V. Pastukhov. A method of storing the frequency of electrical vibrations. RF patent No. 2178196 with a priority of November 20, 2000. Publ. 01/10/2002, Bull. Number 1.

3. GOST 23512-98. Frequency and time standards. General technical requirements and test methods.

Claims (1)

A method for storing the frequency of electrical oscillations with a frequency measure, comprising the operations of comparing the frequency of the measure with the frequency of the standard during the initial time interval T ₀ , determining from the results of these comparisons the initial deviation of the frequency of the measure from the frequency of the standard Δf ₀ and the initial speed of the systematic drift of the frequency of the measure relative to the frequency of the standard ν ₀ correction of the frequency of the measure by the value of the measured initial deviation Δf ₀ at the end of the time interval T ₀ and the subsequent additional periodic correction of the frequency of the measure at stretching the time interval T at time instants

characterized in that the additional periodic correction of the frequency of the measure is performed by β, which over the time interval T decreases with time in accordance with the predicted decrease in the rate of systematic drift of the frequency of the measure, and the initial value of the additional periodic correction of the frequency β _{nach is} set in accordance with the inequality 0 < | β _beg | <2 | ν ₀ | Δt, and its sign is opposite to the sign ν ₀ .

Images