JP2005156211A - Audit method for clock precision among plurality of clocks, and clock audit system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a clock precision audit method and a clock audit system capable of estimating not only time point deviation information in measuring time points between at least two clocks but also the propriety of continuous change in a time deviation between the measuring time points, with a fixed significant level. <P>SOLUTION: This method/system comprises at least two time servers 24, 34. In the method/system, the time deviation between the clocks is found based on time information in the every clock in an optional time, the time deviation obtained by measurement between a pair of internal clocks is classified into periodically measured time deviation groups C, D obtained by executing periodically or in every fixed time interval using a certain time point as a start point, and measured-by-random-extraction time deviation groups E, F obtained by executing in every time determined at random using the certain time point as the start point, and a statistical amount of the periodically measured time deviation groups and a statistical amount of the measured-by-random-extraction time deviation groups obtained by analyzing statistically the periodically measured time deviation groups C, D and the measured-by-random-extraction time deviation groups E, F are found to check the fellow statistical amounts for investigating a significance between the both statistical amounts. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a clock accuracy auditing method and a clock auditing system, and more particularly, from a result of discretely measuring deviations of clock information indicated by at least two clocks in a clock auditing method or calibration system comprising a plurality of clocks. The present invention relates to an audit method and an audit system that can statistically estimate whether or not the deviation amount of a non-measurement section has changed continuously.

  The time distribution and authentication company shall always measure the time deviation between the time of the clock from which it is distributed to the client and the standard time used in that country or region, and publish the information. Can gain the trust of clients.

  For example, in Japan, the agreement obtained by weighted averaging of the time information indicated by the atomic clock of the Communications Research Laboratory (CRL), an independent corporation that establishes Japan Standard Time, and the Standard Time Information of each country by the International Bureau of Weights and Measures (BIPM) When comparing universal time (UTC), an atomic clock mounted on GPS using time information derived from the atomic clock mounted on the satellite of the global positioning system (GPS) as a reference rule A method for measuring the time deviation between clocks, called a common view method, that measures the time difference between each and the NTA clock in each country and uses the result to calculate the time difference between them.

  Furthermore, Patent Document 1 has been proposed as a method for obtaining a time difference with higher accuracy. Patent Document 1 discloses a method of measuring a time deviation between clocks of a ground station, which includes a satellite equipped with an atomic clock and a plurality of ground stations having a communication function, and a clock owned by the ground station. By exchanging signals via satellites with multiple ground stations, the effects of weather and aerodynamic conditions are eliminated as much as possible, so that the time deviation between the clocks of each ground station can be very accurately determined. It becomes possible to ask.

However, even if the invention described in Patent Document 1 is such an excellent method, there was no sudden time difference abnormality in the non-measurement section, or the time deviation obtained in the previous measurement and the current time difference. It is difficult to obtain information such as whether the time deviation obtained by measurement has changed continuously, and it is not possible to estimate the continuity of the time deviation in non-measurement sections of time deviation was there.
JP 2002-82188 A

  Therefore, the technical problem of the present invention is that when time authentication is used as a business, not only the time deviation information at the time of measurement between the standard time of UTC or the country or region and the clock held by the business operator, but also the time deviation between the time points of measurement. It is an object of the present invention to provide a clock accuracy audit method and a clock audit system that can estimate whether or not the frequency of the clock has changed continuously under a certain level of significance.

(1) In order to solve the above technical problem, an inspection method of clock accuracy between a plurality of clocks according to claim 1 of the present invention is composed of at least two clocks, and each clock is set at an arbitrary time. A clock accuracy auditing method having means for obtaining a time deviation between clocks from time information, wherein a time deviation between a pair of clocks obtained by measurement is determined at regular time intervals or at regular time intervals. Time deviation data group measured by random sampling obtained by carrying out every time determined at random starting from the certain point in time And the statistical quantity of the periodically measured time deviation data group obtained by statistical analysis of the time deviation data group measured periodically and the time deviation data group measured by random sampling Random drawing In search of statistics of the measured time difference data group is characterized by performing the test statistic among To investigate the significance of the two statistics.

(2) Further, the watch audit system according to claim 6 of the present invention can communicate directly or via a network by wire or wirelessly, and can send time information unilaterally at an arbitrary time or in response to a request from the other party. Or a clock audit system comprising at least two clocks capable of communicating with each other and capable of measuring a time deviation between its own time information and the other party's time information unilaterally or mutually. Each watch has a means for continuously recording and saving the time deviation from the measured opponent's watch, a means for calculating a statistic of the time deviation data group that has been continuously recorded and saved, and a statistic for the calculation result. At least two statistical recording means to be stored and a means for testing a pair of statistics recorded and stored in the statistical recording means can communicate with the clock itself or the clock. It is characterized by comprising a separate device such as an information device terminals.

  According to the clock accuracy auditing method and the clock auditing system according to the present invention described in the above (1) and (2), a time deviation which is a time lag between clocks between at least two clocks can be set to an arbitrary time. It is possible to estimate whether there is an abnormality in the transition of the time deviation in the non-measurement section in which the time deviation data is not measured by measuring discretely and performing statistical analysis of the measured time deviation data.

  In addition, two data groups, a time deviation data group measured at regular intervals or discretely at regular time intervals, and a time deviation data group measured at random time points are obtained, and the total metric for each system is calculated. And then test the significance of the statistics, test the representative values between the statistics of the two systems, test whether both statistics are significant, and signify significance at a certain level of significance. If there is no sex, it cannot be said that there is a difference in the source of the two systems of data at the significance level, so it can be determined that the time deviation amount is continuous in the section in which the data group is measured. .

  According to the clock accuracy auditing method and the clock auditing system according to the present invention, a time deviation, which is a time lag between clocks between at least two clocks, is discretely measured at an arbitrary time, and measured time deviation data By performing statistical analysis, it is possible to estimate whether there is an abnormality in the transition of the time deviation in the non-measurement section where the time deviation data is not measured, so the time deviation between a pair of clocks is discrete From the measurement data, it is possible to estimate an abnormality in time deviation between the two clocks that occurred in the non-measurement section. Therefore, a business operator who distributes time can obtain the trust of the client by publishing the discrete record of the time deviation from the standard time in the country area and the estimation result of the time deviation of the non-measurement section according to the present invention. it can.

  In the following, when auditing between a plurality of clocks, a pair of target clocks use a clock that becomes a ruler that can measure a time deviation in common, and each pair of watches and a clock that becomes a ruler An example of a common view method for measuring the time deviation of the time and obtaining the time deviation of a pair of watches from the two time deviations will be described.

  FIG. 1 is a block diagram of a clock accuracy audit system for implementing a clock accuracy audit method between a plurality of clocks according to an embodiment of the present invention, wherein 10 is a GPS satellite, 20 is a standard time generator, and 30 is A time information distribution organization 40 is an information processing terminal of a user who receives the distribution.

  The standard time generator 20 includes a GPS receiver 22 that receives radio waves from the GPS satellite 10 via an antenna 21, an atomic clock 23, a time server 24, a transceiver 25, and time deviation recording and statistical analysis. An information processing terminal 26 that performs information communication is included. Similarly, the time information distribution organization 30 also receives a GPS receiver 32 that receives radio waves from the GPS satellite 10 via the antenna 31, an atomic clock 33, a time server 34, a transmitter / receiver 35, and time deviation recording and statistics. It has an information processing terminal 36 that performs analysis and information communication.

  The standard time generating organization 20 and the time information distributing organization 30 communicate with each other via the Internet 61 of the communication network 60 the time indicated by the internal clocks of the time servers 24 and 34 when the GPS time signal is received. The information processing terminal 41 of the user 40 receives time distribution time audit information from the time information distribution organization 30 via the Internet 61.

  As a premise, in the common view method, the clock that becomes a ruler is voluntary or according to the request of a person or device / device that measures the time deviation, at a predetermined periodic interval or at regular time intervals. Time information is distributed to each of a pair of watches that are targets for measuring time deviation. This distribution may be a radio signal such as GPS or a standard radio wave, or may be a wired distribution such as a dedicated line or an optical fiber whose total length is known. When accuracy is not so required, distribution via a network such as the Internet may be possible.

  Since GPS is sending the time information of the standard radio wave every minute every second, it is a measurement target at the time when information of a predetermined time is received by a person or device / apparatus trying to measure the time deviation. What is necessary is just to measure the time information of a certain clock.

  The time deviation between a pair of timepieces in the common view method can be calculated at any time later by obtaining the time deviation from the timepiece as a ruler and sending the data to each other or one side. Therefore, this data transmission may be a written fax or mail, or in some cases verbal transmission. Of course, you may utilize the data communication in the information processing terminal via a network. The calculation may be performed manually or automatically by an information processing terminal that is separately prepared based on the obtained time deviation data, and of course, other methods including manual calculation may be used.

  The above-described method for estimating the continuity of time deviation according to the present invention is based on a pair of timepieces for measuring time deviation between timepieces using a common view method, and a regular or regular interval with each timepiece. The time deviation is measured every time.

  The measured time deviation data is continuously recorded and stored as a first series data group, and statistics such as an average value, variance, and standard deviation of the time deviation data are obtained. The data for calculating the statistic may be all data obtained continuously after the measurement is started, or may be a fixed number of time-series data groups measured in the past from the latest data.

  The statistic based on the first data group due to the time deviation is recorded and saved as the regular time deviation statistic at regular or fixed intervals obtained in this way. In addition, using the time point at which collection of data to be calculated as a fixed time deviation is started as a starting point, the time deviation is separately measured to collect the second series of data groups, and the statistics are similarly calculated.

  The timing for collecting the second series of time deviation data is a randomly selected time such as a time derived from a random number generated by a random number generator. It is desirable that the randomly selected time is included at least once in each section with the time when the first series of data groups is measured as a delimiter.

  Further, a statistic based on the data group of the second series of the time deviation measured at the randomly selected time is calculated and recorded and stored as a random time deviation statistic.

  Next, in order to test the fixed time deviation statistic and the random time deviation statistic with the constant significance level, the difference between the average values of both statistics is taken, and the difference between the average values is 0, that is, both average values Are hypothesized that they are equal, test the significance by t-test, or take the ratio of variances, and establish the hypothesis that the variance ratio is 1, that is, the cause system of the variability of both data groups is the same. Test for significant difference by test. Here, the significant difference is tested based on the result of the t-test under a certain significance level.

  If there is no difference in the t-test results under a certain level of significance, it cannot be said that there is a difference in the causal relationship between the values of the two data groups under that level of significance. If there is no significant difference in the result of the F test under the significance level, it cannot be said that there is a difference in the cause system of the variation between the two data groups under the significance level.

  Based on the above results, it is considered that both data groups are controlled by the same causality, and in a section where measurement is made periodically or at intervals of a certain time interval, those with a certain significance level and time average values are continuously It can be judged that it has changed.

  FIG. 2 is a flowchart for determining the continuity of time deviation between a pair of clocks by a common view method using GPS satellites. As shown in FIG. 2, the standard time generating engine 20 performs steps from step S1 to step S1. The operations up to S3 are executed, and the time information distribution organization 30 executes the operations from steps S4 to S6, and both or one of the standard time generation organization 20 and the time information distribution organization 30 executes the operations from steps S7 to S12. To do.

  That is, the standard time generating engine 20 receives the GPS signal in step S1, determines whether or not it is time information to be measured in step S2, and if it is time information to be measured in step S2, the process proceeds to step S3. Go ahead and record the time indicated by your internal clock. Similarly, the time information distribution organization 30 receives the GPS signal in step S4, determines whether or not it is time information to be measured in step S5, and proceeds to step S6 if it is information to be measured. Record the indicated time.

  Next, both or one of the standard time generation organization 20 and the time information distribution organization 30 calculate and record the time difference indicated by both internal clocks in step S7, and the time of both internal clocks recorded in step S8. The difference is statistically analyzed, and the statistic obtained in step S9 is tested. Next, the process proceeds to step S10, where it is determined whether or not it is significant. If it is not significant, the process proceeds to step S11, where it is determined that the time difference between both clocks is continuous. If there is significance, the process proceeds to step S12, and it is determined that there is no continuity between the two time differences.

  FIG. 3 shows an example of a method for calculating the time deviation of a pair of timepieces in the common view method. In a system for measuring time deviation between a pair of timepieces in the common view method as shown in FIG. Calculate the time deviation of the common view method. In FIG. 3, when GPS signals of 00 and 30 minutes are received every hour, the time indicated by the internal clock of each server is recorded, and the result is sent to each other via the Internet, so that each information processing terminal Is used to calculate the amount of time deviation between servers. Of course, the data may be sent only in one direction, and one information processing terminal may perform the calculation. This is the first data series, that is, the fixed time deviation data groups C and D.

  FIG. 4 is a random time deviation data group that needs to be measured in the present invention, and after a number of seconds randomly determined from 30 minutes, that is, 1800 seconds, with 00 minutes and 30 minutes as an interval for measuring a fixed time deviation. The second series of time deviation data groups E and F obtained by measurement.

Data group C in FIG. 3, D and 4 data groups E, F, respectively, and the GPS time information, the time information T _A the time server 24 in the standard time generating engine 20 is shown, the time information distribution engine 30 Time information T _B indicated by the time server 34 and time deviation information (T _A −T _B ) between the time servers 24 and 34. GPS time information is composed date and hour, minute, second, the time information _{T A} and _{T B} for the time server 24, 34 of the standard time generating engine 20 and the time information distribution engine 30 is shown, date, Consists of hour, minute, second and fractional seconds.

  FIG. 5 shows the statistic analysis result of FIG. 3, and FIG. 6 shows the analysis result of FIG. 4, which is a fixed time deviation data group C and random time deviation from 10:00 to 15:00 on September 10, 2003. The group E and the scheduled time deviation data group D and the random time deviation data group F from 20:30 on September 11, 2003 to 10:00 on September 12, are data measured in the same time zone, If the time deviation is continuously changing, the statistics obtained by statistical calculation should be common. The same can be said for the regular time deviation data group D and the random time deviation data group F.

  It is difficult to find the commonality between the data groups C and E and the data groups D and F only by looking at the results of the statistics. Therefore, the commonality of both statistics is estimated by verifying whether the average values of the data groups to be compared are equal, and whether the variances representing the variation in data are equal. First, as a means for verifying whether or not the average values of the data groups to be compared are statistically equivalent, the difference between the average values is taken and the hypothesis is that the difference is 0, that is, the population averages of both data groups are equal. This hypothesis is tested by t-test.

で表されるｔ_０は自由度ｆ＝ｎ_Ｃ＋ｎ_Ｅ−２のｔ分布となるため、有意水準をＰとして検定すると、
ｔ_０＜ｔ（Ｐ，ｆ）であれば両平均値に有意性があるとはいえないと結論付けられ、ｔ_０＞ｔ（Ｐ，ｆ）であれば両平均値に有意性があると結論付けられる。 Specifically, under the hypothesis that the data group C and the data group E have the same population mean, the data number C of the data group _C , the average value x _C , the deviation sum of squares is S _C , and the data group E N _E , the average value is X _E , and the deviation sum of squares is S _E.

T ₀ represented by is a t distribution with f = n _C + n _E −2 degrees of freedom.
If t ₀ <t (P, f), it can be concluded that both average values are not significant. If t ₀ > t (P, f), both average values are significant. It can be concluded.

  Next, as a means of verifying whether or not the variance of each data group to be compared is statistically equivalent, the ratio of variance is taken, and the ratio is 1, that is, the population variance of both data groups is made equal, This hypothesis is tested by F test.

で表されるＦ_０は自由度（ｆ_Ｅ＝ｎ_Ｅ−１，ｆ_Ｃ＝ｎ_Ｃ−１）のＦ分布となるため、仮説のσ_Ｃ＝σ_ＥよりＦ_０＝Ｓ_Ｅ ^２／Ｓ_Ｃ ^２を、有意水準をＰとして検定すると、
Ｆ_０＜Ｆ（Ｐ，ｆ_Ｅ，ｆ_Ｃ）であれば両標本分散の比に有意性があるとはいえないと結論付けられ、
Ｆ_０＞Ｆ（Ｐ，ｆ_Ｅ，ｆ_Ｃ）であれば両標本分散の比に有意性があると結論付けられる。 Specifically, under the hypothesis that the population variance of the data group C and the data group E is equal, the number of data in the data group C is n _C , the sample variance is S _C ² , the population variance is σ _C ² , Assuming that the number of data of data E is n _E , the sample variance is S _E ² , and the population variance is σ _E ² , when S _E > S _C ,

F ₀ represented by the following _equation is an F distribution with degrees of freedom (f _E = n _E −1, f _C = n _C −1), and therefore F ₀ = S _E ² / S _C from the hypothesis σ _C = σ _E. ² with a significance level of P,
If F ₀ <F (P, f _E , f _C ), it is concluded that the ratio of both sample variances is not significant,
If F ₀ > F (P, f _E , f _C ), it can be concluded that the ratio of both sample variances is significant.

  FIG. 7 shows the results of the t test for the difference between the average values of the data groups C and E in FIG. 5 and the data group D and F in FIG. 6 and the F test for the ratio of variance. In the data group of C and E, the significance level of 5% is not seen in both the t-test for the difference between the mean values and the F-test for the ratio of variance, so between 10:00 on September 10, 2003 and 14:30 on the same day Then, it is determined that the time deviation between the pair of timepieces to be processed has continuity.

  Data groups D and F do not show a significant difference of 5% in the t-test of the difference between the mean values, but the F-test of the variance ratio is significant at the significance level of 5%. In the period between 20:30 on September 11, 2003 and 01:00 on the next 12th, it is determined that there is no continuity in the time deviation between the pair of clocks.

  The reason why it is determined that there is no continuity this time is that it is caused by an abnormality in the time indicated by the time server 34 side of the time information distribution organization 30 when the 23: 1158 signal of the data group F is received. However, it is not impossible to determine which part of the data of which clock is abnormal if the calculation is performed while shifting the statistics of the same number of data by one data.

  FIG. 8 shows a clock audit system according to another embodiment of the present invention, in which one clock of a pair of clocks is directly connected to the information processing terminal, and the other clock is connected to the information processing terminal via the network. FIG. 5 is a block diagram when a time deviation is measured by comparing time information of both clocks, and statistical processing and testing are performed. In FIG. 8, reference numeral 27 denotes one clock part, and 37 denotes the other clock part. The information processing terminal includes an arithmetic device 43 and a storage device 47. The computing device 43 includes a time deviation computing unit 44, a statistical computing unit 45, and a test computing unit 46. The storage device 47 includes a scheduled time deviation recording unit 48, a random time deviation recording unit 49, a scheduled time deviation statistic recording unit 50, a random time deviation statistic recording unit 51, and a test result recording unit 52.

  The time deviation calculation unit 44 of the calculation device 43 receives the clock information from one clock unit 27 via the interface 42a, and receives the time information from the other clock unit 37 described above via the Internet 61 and the interface 42b. Then, time deviation data is calculated. The scheduled time deviation recording unit 48 and the random time deviation recording unit 49 of the storage device 47 record the time deviation information obtained by the time deviation calculating unit 44. The statistical calculation unit 45 of the calculation device 43 reads the fixed time deviation data group from the fixed time deviation recording unit 48 and also reads the random time deviation data group from the random time deviation recording unit 49, and based on these data groups. Calculate the statistic.

  The scheduled time deviation statistic recording unit 50 reads the scheduled time deviation statistic from the statistical calculation unit 45. The random time deviation statistic recording unit 51 reads the random time deviation statistic from the statistical calculation unit 45. The test calculation unit 46 reads the fixed time deviation statistic from the fixed time deviation statistic recording unit 50 and also reads the random time deviation statistic from the random time deviation statistic recording unit 51, and tests based on these statistics. Perform the operation. The test result recording unit 52 records the test result and causes the display device 53 to display the test result.

  The scheduled time deviation statistic recording unit 50 and the scheduled time deviation recording unit 48 continuously record and store the measured time deviation data as a data group of the first series, and the statistical calculation unit 45 includes an average value of the time deviation data, Find statistics such as variance and standard deviation. The data for calculating the statistic may be all data obtained continuously after the measurement is started, or may be a fixed number of time-series data groups measured in the past from the latest data.

  The scheduled time deviation statistic recording unit 50 records and saves the statistic based on the first data group due to the time deviation as the scheduled time deviation statistic obtained in this way at regular or regular intervals. In addition, the statistical calculation unit 45 separately collects the second series of data groups by measuring the time deviation starting from the time when the collection of data to be calculated as the scheduled time deviation is started, and similarly calculates the statistics. calculate.

  Further, the statistical calculation unit 45 calculates a statistic based on the second series data group of the time deviation measured at the randomly selected time as described above, and the random time deviation statistic recording unit 51 as a random time deviation statistic. Record and save.

  The test calculation unit 46 tests the fixed time deviation statistic and the random time deviation statistic with the constant significance level, takes the difference between the average values of the two statistics, and the difference between the average values is 0, that is, both Establishing the hypothesis that the mean values are equal and testing for a significant difference by t-test, or taking the ratio of variances, and establishing the hypothesis that the variance ratio is 1, that is, the cause system of the variance of both data groups is common The significant difference is tested by F test.

  If there is no significant difference between the two statistical instruments as a result of t-test and F-test, it is considered that both data groups are governed by the same causality, and the interval is divided periodically or at regular time intervals. Therefore, it can be determined that the average value and the time average value are continuously changing.

  Since this embodiment is not a method for obtaining time deviation between a plurality of clocks but a method for estimating the continuity of the obtained time deviation, the time at which the time information can be transmitted and received between the networks is obtained. Of course, the present invention can be applied even to a time deviation obtained by a method of directly measuring a time deviation such as two computer internal clocks performed between a server and a PC.

1 is a block diagram of a clock audit system for implementing a clock accuracy audit method between a plurality of clocks according to an embodiment of the present invention. The operation | movement flowchart of the timepiece auditing system for enforcing the timepiece auditing method between several timepieces by embodiment of this invention. A group of data showing an example of a method for calculating a fixed time deviation in a pair of watches in the common view method. A data group showing an example of a method for calculating a random time deviation in a pair of watches in the common view method. The data table of the statistic analysis result of FIG. The data table of the statistic analysis result of FIG. Data table of statistical test results. The block diagram in the case of measuring a time deviation, performing a statistical process, and a test in the timepiece audit system according to another embodiment of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... GPS satellite 20 ... Standard time generation organization 22 ... GPS receiver 23 ... Atomic clock 24 ... Time server 26 ... Information processing terminal 27 ... Clock part 30 ... Time information distribution organization 32 ... GPS receiver 33 ... Atomic clock 34 ... Time Server 36 ... Information processing terminal 37 ... Clock unit 40 ... User 41 ... Information processing terminal 43 ... Computing device 44 ... Time deviation computing unit 45 ... Statistical computing unit 46 ... Test computing unit 47 ... Storage device 48 ... Scheduled time deviation recording unit 49 ... Random time deviation recording part 50 ... Regular time deviation statistical quantity recording part 51 ... Random time deviation statistical quantity recording part 52 ... Test result recording part

Claims (11)

A clock accuracy auditing method comprising at least two clocks and having means for obtaining a time deviation between clocks from time information for each clock at an arbitrary time,
A time deviation data group obtained by performing a time deviation between a pair of clocks obtained by measurement at regular intervals or every predetermined time interval starting from a certain time point, and the certain time point Time deviation data group measured by random sampling obtained by carrying out at each randomly determined time as a starting point, time deviation data group measured periodically and time measured by random sampling Obtain statistical values of the time deviation data group measured periodically and statistical data of the time deviation data group measured by random sampling, obtained by statistical analysis of the deviation data group. A method for auditing clock accuracy among a plurality of clocks, characterized by performing a test between statistics in order to check the sex.   2. The clock accuracy between a plurality of timepieces according to claim 1, wherein the randomly determined time is included at least once in a section in which measurement at regular or regular time intervals is divided. Audit method.   The test for examining the significance is a t-test on the difference between the average value of the time deviation data group measured periodically and the average value of the time deviation data group measured by random sampling. The method of auditing clock accuracy among a plurality of clocks according to claim 1.   The test for examining the significance is an F test relating to a ratio of a variance of time deviation data groups measured periodically and a variance of time deviation data groups measured by random sampling. Item 2. A method for auditing clock accuracy among a plurality of clocks according to Item 1.   When at least one clock is restarted after changing the time information, the time deviation data and statistics collected up to the point of time when the time information was changed and restarted are deleted, and the data of the data is restarted with the restarted time as the origin. 2. The clock accuracy auditing method for a plurality of times according to claim 1, wherein collection recording is started. It consists of at least two watches that can communicate directly or via a network by wire or wireless, and can communicate time information either unilaterally or at the request of the other party at any time. A watch audit system capable of measuring a time deviation between the time information of the other party and the time information of the other party unilaterally or mutually,
At least one timepiece means for continuously recording and storing the time deviation from the measured opponent's timepiece;
Means for calculating statistics of time deviation data groups continuously recorded and stored;
At least two statistical recording means for recording and storing statistics of calculation results;
A timepiece audit system comprising means for examining a pair of statistics recorded and stored in the statistic recording means in a timepiece itself or in another device such as an information processing terminal capable of communicating with the timepiece.   A time deviation data group obtained by carrying out time deviations obtained by measurement by communication of time information periodically or at regular time intervals based on a certain time point, and starting from the certain time point Divided into time deviation data groups measured by random sampling obtained by carrying out at each time determined as random, and time deviation data groups measured periodically and times measured by random sampling 7. The clock audit system according to claim 6, wherein a statistic of the deviation data group is obtained, and the statistic is tested to check the significance of both statistics.   7. The clock audit system according to claim 6, wherein the randomly determined time is included at least once in a section that is divided periodically or at intervals of a certain time interval.   The test for examining the significance is a t-test on the difference between the average value of the time deviation data group measured periodically and the average value of the time deviation data group measured by random sampling. The watch inspection system according to claim 6.   The test for examining the significance is an F test relating to a ratio of a variance of time deviation data groups measured periodically and a variance of time deviation data groups measured by random sampling. Item 7. The watch audit system according to item 6.   When at least one clock is restarted after changing the time information, the time deviation data and statistics collected up to the point of time when the time information was changed and restarted are deleted, and the data of the data is restarted with the restarted time as the origin. The timepiece audit system according to claim 6, wherein the collection record is started.

Images