阅读说明：本技术 一种使用率较高关口表的评估方法 (Assessment method for gateway table with high utilization rate ) 是由 司浩天 唐亮 冯硕 龚振龙 马一峰 梁晓伟 刘云波 王珍珍 王一博 于 2020-06-28 设计创作，主要内容包括：本发明属于评估方法,具体涉及一种使用率较高关口表的评估方法。它包括：步骤1：采样,在对关口表校准时,对关口表的时钟周期进行连续采样,步骤2：曲线拟合,对步骤1中得到的ΔT<Sub>i</Sub>进行分段,保留最后p个数据,其余数据曲线拟合并进行数据分段,对每段数据分别再次单独进行曲线拟合,步骤3：验证,选取其中误差最小的三个函数保留,步骤4：预测,用步骤3中保留的函数对下个校准时间点做预测,并作出最终判断。本发明的显著效果是：能够评估出未来一个校准周期内,关口表的时钟信号是否会出现超差情况,因而提前避免了因关口表时钟信号误差超差导致的电量示数冻结不同步、算费示数冻结不准确等问题。(The invention belongs to an evaluation method, and particularly relates to an evaluation method of a gateway table with a high utilization rate. It includes: step 1: sampling, namely continuously sampling the clock period of the gateway table when the gateway table is calibrated, and step 2: curve fitting, for the Δ T obtained in step 1 i Segmenting, reserving the last p data, performing curve fitting on the rest data, performing data segmentation, and performing curve fitting on each segment of data independently again, wherein the step 3: and (4) verifying, selecting three functions with the minimum error and reserving, wherein the three functions are as follows: and (4) predicting, namely predicting the next calibration time point by using the function reserved in the step 3, and making a final judgment. The invention has the following remarkable effects: whether the clock signal of the gateway meter can be out of tolerance in a calibration period in the future can be evaluated, so that the problems of asynchronous freezing of electric quantity readings, inaccurate freezing of calculation expense readings and the like caused by the out of tolerance of the clock signal of the gateway meter are solved in advance.)

1. A method for evaluating a high-usage correlation table, comprising the steps of:

step 1: sampling

When the gateway table is calibrated, the clock period of the gateway table is continuously sampled,

step 2: fitting of curves

For the delta T obtained in step 1_iSegmenting, reserving the last p data, using the rest data for subsequent calculation,

for delta T capable of participating in calculation_iPerforming curve fitting, the function after fitting is f₀(t) represents a number of times,

for function f₀(t) deriving to obtain a first derivative function

At the point t obtained as described above₁、t₂、…、t_mFor the demarcation point, the Δ T which can participate in the calculation_iSegmenting, and independently fitting a curve in each segment to respectively obtain f₁(t)、f₂(t)、…、f_m(t) wherein f₁(t) corresponds to 0-t₁Fitted curve of time period, f₂(t) corresponds to t₁-t₂Fitted curve for time period, …, f_m(t) corresponds to t_m-1-t_mA fitted curve of the time period is obtained,

and step 3: authentication

Using the curve f obtained in step 2₀(t)、f₁(t)、f₂(t)、…、f_m(t) predicting the time period of the data retained in the step 2, comparing the time period with the retained data, selecting three functions with minimum error for retention,

and 4, step 4: prediction

Predicting the next calibration time point by using the functions reserved in the step 3, and if the prediction results of the three functions are errors meeting the requirement, judging that the problem of clock out-of-tolerance cannot be generated in the next calibration period of the gateway table; if the predicted results of the three functions are error out-of-tolerance, judging that the gateway table can generate a clock out-of-tolerance problem in a calibration period in the future, and proposing a suggestion of replacing the gateway table; if the prediction results of the three functions are inconsistent, the accurate result cannot be judged, and a suggestion of manually checking the gateway table is provided.

2. A method of evaluating a higher usage correlation table as defined in claim 1, wherein: the step 1 comprises the steps of,

the number of samples is more than 1000, and the sampling result is expressed by the following formula

T_i＝T₀+ΔT_i

Wherein i is the number of sampling times; t is_iRepresents the result of the ith sampling; t is₀The clock period is rated for the gateway table, and the data can be looked up from the use instruction of the gateway table; delta T_iIn order to measure the error between the ith sampling result and the rated period, the sampling interval t is recorded, the maximum value of the sampling times is n, namely the value range of i is a positive integer from 1 to n,

the sampling interval can be selected at will within the range given by the application, and the sampling interval is required to be kept unchanged for the same sampling behavior of the same gateway table; the sampling interval may be different for different correlation tables or different sub-samples of the same correlation table.

3. A method of evaluating a higher usage correlation table as defined in claim 2, wherein: the sampling interval is greater than or equal to 100 gate table clock cycles.

4. A method of evaluating a higher usage correlation table as defined in claim 3, wherein: the step 2 comprises

The value range of p is 100-10% n, namely the minimum value of p is 100, and the maximum value of p is 10% n.

5. A method of evaluating a higher usage correlation table as defined in claim 4, wherein: the step 3 comprises the steps of,

the minimum error is determined according to the following method,

first using f₀(t)、f₁(t)、f₂(t)、…、f_m(t) predicting the time period of data retention in the step 2, wherein the value of n is determined and the value range of p is also determined as the sampling number is determinedThe value of p is determined by external input, and the sampling interval is determined and is constant in the same sampling, so that the time coordinate of each sampling point can be obtained and the time coordinate is substituted into f₀(t)、f₁(t)、f₂(t)、…、f_m(t), the predicted values at different time coordinates can be obtained,

the values of the same function at different time coordinates are then accumulated, since the fit is to Δ T in step 1_iSo that the prediction result is also Δ T at different time coordinates_iTherefore, the corresponding delta T can be obtained by substituting different time coordinates into the formula_iP Δ T_iAdding up to obtain total delta T of the function_i，

Thirdly, comparing the accumulated values corresponding to different functions, taking the function corresponding to the three numbers with the minimum accumulated value as the output of the step, and preferentially selecting f if the minimum accumulated value corresponds to a plurality of functions₀(t) then with f_iThe value of i in (t) is larger than the preference function.

Technical Field

The invention belongs to an evaluation method, and particularly relates to an evaluation method of a gateway table with a high utilization rate.

Background

The gateway meter is an electric energy meter installed and operated in gateway electric energy metering devices such as power generation enterprises for surfing the internet, cross-regional connecting lines, provincial connecting lines, intra-provincial networks and the like, is used for trade settlement and check of internal economic indexes, is mostly an imported 0.25-grade multifunctional electric energy meter, and plays an important role in electric energy metering of the whole power grid.

The clock information of the table of gates is one of the key parameters. The inaccurate clock information can cause the problems of asynchronous freezing of the electric quantity readings, inaccurate freezing of the calculation fee readings and the like. Clock out-of-tolerance defects are formed when the clock deviation exceeds a standard range allowed by regulations, and the clock out-of-tolerance directly causes deviation of electricity statistics of the meter.

In the traditional power grid, the parameters of the gateway meter are mainly collected manually and regularly, but the distribution of the gateway meter is very wide, the field inspection periods of different transformers have differences, and operators on duty can not easily find the faults every day during routing inspection and meter reading, so that the faults can seriously affect the accuracy of the electric energy metering of the gateway for a long time, and the faults can be found only after larger electric energy metering errors are caused or the regular field inspection of the gateway meter is carried out.

According to the requirements of the electric energy metering device on-site inspection operation instruction book, the time difference between the internal clock of the electric energy meter running on site and Beijing is not more than 5 minutes every year in principle; the calibration period must not be less than 1 time per year or shortened as appropriate. Before field calibration, whether a calendar clock in a gateway table is correct or not is firstly checked, the time difference with Beijing is within 5 minutes, and the time is adjusted on the spot; the time error with Beijing is more than 5 minutes, the reason is analyzed, and the gateway table is replaced when necessary.

An excessively long calibration period may lead to poor calibration effect, and therefore, a method capable of determining whether the clock information of the gate table can satisfy the requirement for a long time during calibration is needed.

Disclosure of Invention

Aiming at the defects of the prior art, the application provides an evaluation method of a gateway table with a high utilization rate.

The invention is realized by the following steps: a method for evaluating a high-usage gate list comprises the following steps:

step 1: sampling

When the gateway table is calibrated, the clock period of the gateway table is continuously sampled,

step 2: fitting of curves

For the delta T obtained in step 1_iSegmenting, reserving the last p data, using the rest data for subsequent calculation,

for delta T capable of participating in calculation_iPerforming curve fitting, the function after fitting is f₀(t) represents a number of times,

for function f₀(t) deriving to obtain a first derivative function f₀(t) then find f₀Points where (t) is 0, and these points are denoted as t₁、t₂、…、t_m，

At the point t obtained as described above₁、t₂、…、t_mFor the demarcation point, the Δ T which can participate in the calculation_iSegmenting, and independently fitting a curve in each segment to respectively obtain f₁(t)、f₂(t)、…、f_m(t) wherein f₁(t) corresponds to 0-t₁Fitted curve of time period, f₂(t) corresponds to t₁-t₂Fitted curve for time period, …, f_m(t) corresponds to t_m-1-t_mA fitted curve of the time period is obtained,

and step 3: authentication

Using the curve f obtained in step 2₀(t)、f₁(t)、f₂(t)、…、f_m(t) predicting the time period of the data retained in the step 2, comparing the time period with the retained data, selecting three functions with minimum error for retention,

and 4, step 4: prediction

Predicting the next calibration time point by using the functions reserved in the step 3, and if the prediction results of the three functions are errors meeting the requirement, judging that the problem of clock out-of-tolerance cannot be generated in the next calibration period of the gateway table; if the predicted results of the three functions are error out-of-tolerance, judging that the gateway table can generate a clock out-of-tolerance problem in a calibration period in the future, and proposing a suggestion of replacing the gateway table; if the prediction results of the three functions are inconsistent, the accurate result cannot be judged, and a suggestion of manually checking the gateway table is provided.

The method for evaluating a high usage rate correlation table as described above, wherein the step 1 comprises,

the number of samples is more than 1000, and the sampling result is expressed by the following formula

T_i＝T₀+ΔT_i

Wherein i is the number of sampling times; t is_iRepresents the result of the ith sampling; t is₀The clock period is rated for the gateway table, and the data can be looked up from the use instruction of the gateway table; delta T_iIn order to measure the error between the ith sampling result and the rated period, the sampling interval t is recorded, the maximum value of the sampling times is n, namely the value range of i is a positive integer from 1 to n,

the sampling interval can be selected at will within the range given by the application, and the sampling interval is required to be kept unchanged for the same sampling behavior of the same gateway table; the sampling interval may be different for different correlation tables or different sub-samples of the same correlation table.

The method for evaluating the high-usage gate table as described above, wherein the sampling interval is greater than or equal to 100 gate table clock cycles.

The method for evaluating a high-usage relationship table as described above, wherein the step 2 includes

The value range of p is 100-10% n, namely the minimum value of p is 100, and the maximum value of p is 10% n.

The method for evaluating a high usage rate correlation table as described above, wherein the step 3 includes,

the minimum error is determined according to the following method,

first using f₀(t)、f₁(t)、f₂(t)、…、f_m(t) predicting the time period of data retention in the step 2, wherein the value of n is determined because the sampling number is determined, the value range of p is also determined, the value of p is also determined through external input, and simultaneously, because the sampling interval is determined and the sampling interval is not changed in the same sampling, the time coordinate of each sampling point can be obtained, and the time coordinate is substituted into the f₀(t)、f₁(t)、f₂(t)、…、f_m(t), the predicted values at different time coordinates can be obtained,

the values of the same function at different time coordinates are then accumulated, since the fit is to Δ T in step 1_iSo that the prediction result is also Δ T at different time coordinates_iTherefore, the corresponding delta T can be obtained by substituting different time coordinates into the formula_iP Δ T_iAdding up to obtain total delta T of the function_i，

Thirdly, comparing the accumulated values corresponding to different functions, taking the function corresponding to the three numbers with the minimum accumulated value as the output of the step, and preferentially selecting f if the minimum accumulated value corresponds to a plurality of functions₀(t) then with f_iThe value of i in (t) is larger than the preference function.

The invention has the following remarkable effects: the evaluation method can evaluate whether the clock signal of the gateway meter can be out of tolerance in a future calibration period in one-time clock calibration of the gateway meter, and if the situation of predicting the clock out of tolerance occurs, the gateway meter can be replaced in time, so that the problems of unsynchronized freezing of the electric quantity readings, inaccurate freezing of the calculation expense readings and the like caused by the error out of tolerance of the clock signal of the gateway meter are avoided in advance.

Detailed Description

A method for evaluating a high-usage gate list comprises the following steps:

step 1: sampling

When the gateway table is calibrated, the clock period of the gateway table is continuously sampled, the sampling interval is more than or equal to 100 clock periods of the gateway table, the sampling number is more than 1000, and the sampling result is expressed by the following formula

T_i＝T₀+ΔT_i

Wherein i is the number of sampling times; t is_iRepresents the result of the ith sampling; t is₀The clock period is rated for the gateway table, and the data can be looked up from the use instruction of the gateway table; delta T_iIs the error between the measured ith sampling result and the rated period. The sampling interval t is additionally recorded. The maximum value of the sampling times is n, namely the value range of i is a positive integer from 1 to n.

The sampling interval may be chosen arbitrarily within the ranges given in this application. For the same sampling behavior of the same gate table, the sampling interval needs to be kept unchanged; the sampling interval may be different for different correlation tables or different sub-samples of the same correlation table. For example, the sampling interval of a certain gateway table is set to 150 clock cycles of the gateway table, and the total number of samples is 3000, then 150 clock cycles of the gateway table must be kept for all 3000 samples. After the sampling is finished, the clock period of the same gateway table needs to be sampled again, so that the sampling interval can be set to 200 clock periods of the gateway table, the total sampling amount is 2000, and 200 clock periods of the gateway table need to be kept unchanged for the 2000 samples.

Step 2: fitting of curves

For the delta T obtained in step 1_iAnd (3) segmenting, reserving the last p data, using the rest data for subsequent calculation, wherein the value range of p is 100-10% n, namely the minimum value of p is 100, the maximum value of p is 10% n, for example, when n is 1000, the value of p is 100, and when n is 2000, the value range of p is 100-200. The p value is directly input from the outside.

For delta T capable of participating in calculation_iPerforming curve fitting, the function after fitting is f₀(t) represents.

The curve fitting may be any existing fitting method, and for example, interpolation, least squares, and the like are optional methods.

For function f₀(t) deriving to obtain a first derivative function f₀(t) then find f₀Points where (t) is 0, and these points are denoted as t₁、t₂、…、t_m。

At the point t obtained as described above₁、t₂、…、t_mFor the demarcation point, the Δ T which can participate in the calculation_iSegmenting, and performing curve fitting in each segment by the same fitting method to respectively obtain f₁(t)、f₂(t)、…、f_m(t)。

This step is to fit the global data and the individual data separately, since Δ T_iIs randomly present, the overall trend and the trend for small ranges of data are not necessarily the same. The specific data of which part can be closer to the final result cannot be judged in the step, so that each data segment is fitted.

And step 3: authentication

Using the curve f obtained in step 2₀(t)、f₁(t)、f₂(t)、…、f_m(t) predicting the time period for retaining the data in the step 2, comparing the time period with the retained data, and selecting three functions with the minimum error for retention.

The minimum error is determined by the following method.

First using f₀(t)、f₁(t)、f₂(t)、…、f_m(t) predicting the time period for which data is retained in step 2. The sampling number is determined, so that the value of n is determined, the value range of p is also determined, the value of p input from outside is also determined, meanwhile, the sampling interval is determined, and the sampling interval is unchanged in the same sampling, so that the time coordinate of each sampling point can be obtained, and the time coordinate is substituted into the f₀(t)、f₁(t)、f₂(t)、…、f_mIn (t), the predicted values at different time coordinates can be obtained.

The values of the same function at different time coordinates are then accumulated. Since the fitting is to Δ T in step 1_iSo that the prediction results are also in different time coordinatesΔ T of_iTherefore, the corresponding delta T can be obtained by substituting different time coordinates into the formula_iP Δ T_iAdding up to obtain total delta T of the function_i。

And thirdly, comparing the accumulated values corresponding to different functions, and taking the function corresponding to the three numbers with the minimum accumulated value as the output of the step. If there are multiple functions corresponding to the minimum accumulated value, then f is preferably selected₀(t) then with f_iThe value of i in (t) is larger than the preference function. For example, have f₀(t)、f₂(t)、f₅(t)、f₆(t)、f₉(t) the corresponding accumulated values all belong to the range of the three lowest values, then f is selected first₀(t) then selecting f_iI in (t) is large, i.e. f₆(t)、f₉(t), so the final output of this step is f₀(t)、f₆(t)、f₉(t)。

And 4, step 4: prediction

Predicting the next calibration time point by using the functions reserved in the step 3, and if the prediction results of the three functions are errors meeting the requirement, judging that the problem of clock out-of-tolerance cannot be generated in the next calibration period of the gateway table; if the predicted results of the three functions are error out-of-tolerance, judging that the gateway table can generate a clock out-of-tolerance problem in a calibration period in the future, and proposing a suggestion of replacing the gateway table; if the prediction results of the three functions are inconsistent, the accurate result cannot be judged, and a suggestion of manually checking the gateway table is provided.