A method for improving the metering accuracy of electronic gas meters

Abstract

This invention relates to a method for improving the metering accuracy of electronic gas meters. The method is characterized by: determining the sampling time by generating random numbers within a fixed time interval, thus sampling once within each fixed time interval; and using the average value and sample standard deviation of flow data over a past period to determine if there are significant flow fluctuations; calculating the cumulative flow within the sampling interval using a trapezoidal algorithm, comparing the instantaneous flow at the current time with the instantaneous flow at the previous sampling time; and shortening the fixed time interval and increasing the sampling frequency when significant flow fluctuations occur. This invention, by employing random sampling within fixed time intervals, ensures that the sampling points change continuously within each fixed time interval, thus more accurately tracking flow changes. The trapezoidal algorithm used to calculate the cumulative flow is also more effective in reducing the impact of flow fluctuations on metering accuracy.

Description

Technical Field

[0001] This invention relates to a method for improving the metering accuracy of electronic gas meters. Background Technology

[0002] Electronic gas meters are characterized by having no moving parts, small size, easy integration, high sensitivity, and intelligence, and are now widely used in gas metering. As a new type of metering instrument, the electronic gas meter's metering principle is basically based on the instantaneous flow method. The sensor samples within a fixed sampling period to measure the instantaneous flow rate. The cumulative flow rate over each sampling period is then summed to obtain the cumulative flow rate for the entire time period. Existing methods use a fixed sampling frequency and assume that the gas flow is stable within a single sampling period, i.e., the instantaneous flow rate remains constant. The cumulative flow rate Q within a single sampling period is... m It is the instantaneous flow rate q at this moment. m With sampling period t s The product of, i.e., Q m =q m ×t s However, during gas pipeline transportation, airflow fluctuations, especially significant changes in flow rate (such as pulsating flow), lead to substantial errors when using a fixed sampling frequency to directly accumulate flow rate by multiplying instantaneous flow rate by the sampling period. For situations with large flow rate fluctuations, existing technologies use a fixed sampling period, judging flow rate based on changes in instantaneous flow rate at adjacent sampling times, increasing the sampling frequency when flow rate fluctuations are large. However, this method may miss the moments of significant flow rate changes when flow rate undergoes periodic or regular sudden changes, or result in late detection of flow rate changes. Furthermore, when adjacent moments are both high flow rates, incorrect judgments can occur, reducing measurement accuracy. Summary of the Invention

[0003] In view of the problems existing in the prior art, the purpose of this invention is to provide a technical solution for improving the metering accuracy of electronic gas meters.

[0004] The method for improving the metering accuracy of electronic gas meters is characterized by: determining the sampling time by generating random numbers within a fixed time interval, thereby sampling once within each fixed time interval, and judging whether there is a large fluctuation in flow rate by using the average value and sample standard deviation of flow rate data over a past period; the cumulative flow rate within the sampling interval is calculated using a trapezoidal algorithm, by comparing the instantaneous flow rate at the current time with the instantaneous flow rate at the previous sampling time, and the cumulative flow rate within each sampling interval is: Q m It is the cumulative flow rate within a sampling interval, t m It is the m-th sampling time, t m-1It is the m-1 sampling time. It is t m Instantaneous flow rate at the sampling time, It is t m-1 The instantaneous flow rate at the sampling time; when there are large fluctuations in flow rate, the sampling frequency is increased within a shorter fixed time interval.

[0005] The method for improving the metering accuracy of an electronic gas meter is characterized by the following steps: Step (1): Set a fixed time interval and detection cycle, and record instantaneous flow data.

[0006] During normal metering, random numbers are generated at fixed time intervals of T seconds. The first sampling period generates a random number t1 between [0, T]. At the sampling time t1, the instantaneous flow rate is obtained. The second sampling period generates a random number t2 between [T, 2T] to obtain the instantaneous flow rate at sampling time t2. The k-th sampling period generates a random number t in the range [kT-T, kT]. k , to obtain t k Instantaneous flow rate at the sampling time

[0007] The detection period is defined as every h data points, where h is an integer. The first detection period is the [1, h]th instantaneous flow data point, and the second detection period is the [2, h+1]th instantaneous flow data point.

[0008] In step (1) above, the cumulative flow rate within each sampling interval is calculated using a trapezoidal algorithm, i.e., the cumulative flow rate within the first sampling interval is Q1 = q1 × t1.

[0009] Cumulative flow rate within the second sampling interval

[0010] The cumulative flow rate within the kth sampling interval Step (2): Determine if there are large fluctuations in flow rate during the detection period.

[0011] Set the threshold for the average instantaneous flow rate within the detection period. The threshold s for sample standard deviation and the threshold C for coefficient of variation, when the mean of the test data is less than the mean threshold. Use the sample standard deviation to determine if there are large flow fluctuations; when the average value of the test data is greater than the average value threshold. The coefficient of variation, which is the ratio of the sample standard deviation to the mean, is used to determine whether there are large fluctuations in flow.

[0012] Step (3): When the flow rate fluctuates significantly, shorten the fixed time interval and increase the sampling frequency.

[0013] With a fixed time interval of T / b, where b is an integer, a random number is generated within each time interval, and the sampling time is determined by the value of the random number; assuming that m seconds have elapsed when entering step (3), the next sampling time is in the range [m, m+T / b] where a random number t is generated. m1 , sampling t m1 Instantaneous flow rate at any moment If the flow rate continues to fluctuate significantly, the next sampling time will be in the range [m+T / b, m+2T / b] where a random number t will be generated. m2 , sampling t m2 Instantaneous flow rate at any moment Cumulative flow rate within the sampling interval The trapezoidal algorithm is still used, that is After each sampling, the past h detection data are still used as a detection period to obtain the mean and sample standard deviation, and to determine whether the flow rate is still fluctuating significantly.

[0014] The method for improving the metering accuracy of an electronic gas meter is characterized in that, in step (2), the average instantaneous flow rate is measured in the first detection cycle. Sample standard deviation of instantaneous flow rate The following situations exist:

[0015] ① If the average value of instantaneous flow rate detection data If the sample standard deviation s1 of the detection data is less than the sample standard deviation threshold s, it is determined that there is no large fluctuation in the flow rate. Continue to calculate the cumulative flow rate according to the sampling method with a fixed time interval of T seconds in step (1).

[0016] ②If the average value of instantaneous flow rate detection data If the sample standard deviation of the detection data s1≥s, it is determined that there is a large fluctuation in the flow rate, and proceed to step (3);

[0017] ③ If the average value of instantaneous flow rate detection data And the coefficient of variation C1 of the detected data is less than the coefficient of variation threshold C, where At this point, it is determined that there is no significant fluctuation in flow rate, and the cumulative flow rate is calculated using the sampling method with a fixed time interval of T seconds as described in step (1).

[0018] ④ If the average value of instantaneous flow rate detection data Furthermore, the coefficient of variation of the detection data C1≥C, at which point it is determined that there is a large fluctuation in the flow rate, and proceed to step (3).

[0019] The method for improving the metering accuracy of an electronic gas meter is characterized in that, in step (3), t m1 The corresponding detection period for each moment is explained. Within the current detection period, h instantaneous flow data points are... in, It is t m1-h+1 Instantaneous flow rate at the sampling time It is t m1-h+2 Instantaneous flow rate at the sampling time, and the mean of h instantaneous flow rate data. Sample standard deviation The following situations exist:

[0020] 1) If the average value of the instantaneous flow rate detection data Sample standard deviation s of the test data m <a*s, where a<1, indicates that the flow rate is slowly recovering to a stable state. Exit the sampling method in step (3) and calculate the cumulative flow rate according to the sampling method with a fixed time interval of T seconds in step (1).

[0021] 2) If the average value of the instantaneous flow rate detection data Sample standard deviation s of the test data m ≥a*s, at this time the flow rate still fluctuates greatly, continue to calculate the cumulative flow rate according to the sampling method with a fixed time interval of T / b in step (3);

[0022] 3) If the average value of the instantaneous flow rate detection data And the coefficient of variation C of the detection data m <d*C, where d<1, where At this point, it indicates that the flow rate is slowly recovering to a stable state. Exit the sampling method in step (3) and calculate the cumulative flow rate according to the sampling method with a fixed time interval of T seconds in step (1).

[0023] 4) If the average value of the instantaneous flow rate detection data Furthermore, the coefficient of variation of the detection data C1≥d*C, at which point the flow rate still fluctuates significantly. Continue to calculate the cumulative flow rate using the fixed time interval sampling method of T / b in step (3).

[0024] The beneficial effects of this invention are as follows: by employing random sampling within fixed time intervals, the sampling points change continuously within each fixed time interval, enabling more accurate tracking of flow rate changes. A fixed sampling period, with sampling points T, 2T, and 3T each time, may only capture high or low points when the flow rate changes. Secondly, the average value and sample standard deviation of the past h consecutive instantaneous flow rate data are used to determine flow rate fluctuations. When large flow rate fluctuations occur, the flow rate data has a high degree of dispersion, resulting in a large sample standard deviation. When the average value of the instantaneous flow rate data exceeds a set threshold, the instantaneous flow rate of different detection groups may differ significantly. Using the ratio of the sample standard deviation to the average value, i.e., the coefficient of variation, can further eliminate the influence of numerical magnitude. A smaller coefficient of variation indicates a more stable instantaneous flow rate. Furthermore, when flow rate fluctuations are large, the fixed sampling interval is reduced, increasing the sampling frequency. The trapezoidal method used in this invention to calculate cumulative flow rate is more conducive to reducing the impact of flow rate fluctuations on measurement accuracy. When the flow rate returns to normal, the normal sampling mode is resumed, reducing power consumption. Detailed Implementation

[0025] This invention provides a method for improving the metering accuracy of electronic gas meters. The sampling time is determined by generating random numbers within a single time interval at fixed intervals, thereby sampling once within each fixed time interval. The average value and sample standard deviation of instantaneous flow data over a past period are used to determine whether there are large fluctuations in flow. The cumulative flow within the sampling interval is calculated using a trapezoidal algorithm, which compares the instantaneous flow at the current moment with the instantaneous flow at the previous sampling moment.

[0026] This invention discloses a method for improving the metering accuracy of an electronic gas meter, comprising the following steps:

[0027] Step (1): Set a fixed time interval and detection cycle, and record instantaneous flow data.

[0028] During normal metering, random numbers are generated at fixed time intervals of T seconds. The first sampling period generates a random number t1 between [0, T]. At the sampling time t1, the instantaneous flow rate is obtained. The second sampling period generates a random number t2 between [T, 2T] to obtain the instantaneous flow rate at sampling time t2. The k-th sampling period generates a random number t in the range [kT-T, kT]. k , to obtain t k Instantaneous flow rate at the sampling time

[0029] Randomly determining sampling points allows for more accurate tracking of flow rate changes; the detection period is defined as every h data points, where h is an integer. The first detection period is the [1, h]th instantaneous flow rate data point, and the second detection period is the [2, h+1]th instantaneous flow rate data point.

[0030] In step (1) above, if the cumulative flow rate within each sampling interval is calculated using the rectangular method, the fluctuation of airflow during the gas pipeline transportation process will be ignored; therefore, the present invention uses the trapezoidal algorithm to calculate the cumulative flow rate within the sampling interval using the instantaneous flow rate at the current time and the instantaneous flow rate at the previous sampling time.

[0031] That is, the cumulative flow rate within the first sampling interval is Q1 = q1 × t1.

[0032] Cumulative flow rate within the second sampling interval

[0033] The cumulative flow rate within the kth sampling interval Step (2): Determine if there are large fluctuations in flow rate during the detection period.

[0034] Set the threshold for the average instantaneous flow rate within the detection period. The threshold s for sample standard deviation and the threshold C for coefficient of variation, when the mean of the test data is less than the mean threshold. Use the sample standard deviation to determine if there are large flow fluctuations; when the average value of the test data is greater than the average value threshold. The coefficient of variation, which is the ratio of the sample standard deviation to the mean, is used to determine whether there are large fluctuations in flow.

[0035] Taking the first detection cycle as an example, the average instantaneous flow rate Sample standard deviation of instantaneous flow rate The following situations exist:

[0036] ① If the average value of instantaneous flow rate detection data This indicates that the instantaneous flow rate values ​​at each sampling point are not large, and there may be cases where no flow passes through. If the coefficient of variation is used for calculation, the denominator value is very small or close to zero, resulting in a large deviation in the calculation result. At this time, the sample standard deviation of the detection data is used as the judgment condition. If the sample standard deviation s1 of the detection data is less than the sample standard deviation threshold s, it indicates that the dispersion of the h instantaneous flow rate data is not high, and the difference in instantaneous flow rate at each sampling point is not large. At this time, it is determined that there is no large fluctuation in flow rate, and the cumulative flow rate calculation is continued according to the sampling method with a fixed time interval of T seconds in step (1).

[0037] ②If the average value of instantaneous flow rate detection data The standard deviation of the sample data s1≥s indicates that the dispersion of the h instantaneous flow data is high and the difference in instantaneous flow at each sampling point is large. At this time, it is determined that there is a large fluctuation in flow and proceed to step (3).

[0038] ③ If the average value of instantaneous flow rate detection data When the instantaneous flow rate is large, the coefficient of variation is used for determination, which can further eliminate the influence of the magnitude of values ​​from different detection groups; if the coefficient of variation C1 of the detection data is less than the coefficient of variation threshold C, then... This indicates that the h instantaneous flow data have good stability. At this point, it is determined that there is no large fluctuation in flow. Continue to calculate the cumulative flow according to the sampling method with a fixed time interval of T seconds in step (1).

[0039] ④ If the average value of instantaneous flow rate detection data Furthermore, the coefficient of variation C1 ≥ C of the detected data indicates poor stability of the h instantaneous flow data. At this point, it is determined that there is a large fluctuation in flow, and the process proceeds to step (3). Step (3): When the flow fluctuation is large, shorten the fixed time interval and increase the sampling frequency.

[0040] With a fixed time interval of T / b, where b is an integer, a random number is generated within each time interval, and the sampling time is determined by the value of the random number.

[0041] Assuming m seconds have elapsed before step (3) is reached, the next sampling time will be in the range [m, m+T / b] where a random number t is generated. m1 , sampling t m1 Instantaneous flow rate at any moment If the flow rate continues to fluctuate significantly, the next sampling time will be in the range [m+T / b, m+2T / b] where a random number t will be generated. m2 , sampling t m2 Instantaneous flow rate at any moment Cumulative flow rate within the sampling interval The trapezoidal algorithm is still used, that is After each sampling, using the past h detection data points as a detection period, the mean and sample standard deviation are obtained to determine whether the flow rate is still fluctuating significantly. (Using t...) m1 The detection period corresponding to each moment is explained, and the h instantaneous flow rates within the current detection period are... It is t m1-h+1 Instantaneous flow rate at the sampling time It is t m1-h+2 Instantaneous flow rate at the sampling time, and the mean of h instantaneous flow rate data. Sample standard deviation The following situations exist:

[0042] 1) If the average value of the instantaneous flow rate detection data Sample standard deviation s of the test data m <a*s, where a<1, considering that the instantaneous flow rate will not remain in a state of large fluctuations indefinitely and will gradually return to a stable state, the sample standard deviation threshold a*s (a<1) for exiting step (3) is set to be smaller than that for entering. When the sample standard deviation s of the detection data m When <a*s, it indicates that the instantaneous flow rate has gradually stabilized, and the sampling method in step (3) is exited. The cumulative flow rate is calculated according to the sampling method with a fixed time interval of T seconds in step (1).

[0043] 2) If the average value of the instantaneous flow rate detection data Sample standard deviation s of the test data m ≥a*s, at this time the flow rate still fluctuates greatly, continue to calculate the cumulative flow rate according to the sampling method with a fixed time interval of T / b in step (3);

[0044] 3) If the average value of the instantaneous flow rate detection data And the coefficient of variation C of the detection data m <d*C, where d<1, where Considering that instantaneous flow rate will not remain in a state of large fluctuations indefinitely and will gradually return to a stable state, the differences between instantaneous flow rate data within the detection period become smaller. Therefore, the coefficient of variation threshold d*C (d<1) for exiting step (3) is set to be smaller than that for entering. When the coefficient of variation C of the detection data... m When <d*C, it indicates that the instantaneous flow rate has gradually stabilized. Exit the sampling method in step (3) and calculate the cumulative flow rate according to the sampling method with a fixed time interval of T seconds in step (1).

[0045] 4) If the average value of the instantaneous flow rate detection data Furthermore, the coefficient of variation of the detection data C1≥d*C, at which point the flow rate still fluctuates significantly. Continue to calculate the cumulative flow rate using the fixed time interval sampling method of T / b in step (3).

[0046] Example

[0047] In step (1), the fixed sampling interval T can be set to 1 second. In normal metering mode, sampling is performed once in a random manner every 1 second. The detection period h of instantaneous flow fluctuation is 30. The mean and standard deviation of every 30 instantaneous flow data are used to determine whether there is a large flow fluctuation.

[0048] In step (2), taking a G2.5 grade household electronic gas meter as an example, the average threshold value is... 0.014m can be taken. 3 / s, the sample standard deviation threshold s is set to 0.052m 3 / s, the coefficient of variation threshold C is set to 0.65, when the average value of the detected data is less than 0.014m 3 At a rate of / s, is the sample standard deviation of the detection data greater than 0.052m? 3 / s determination; when the average value of the detection data is greater than 0.014m 3 When the value is / s, the determination is made by whether the coefficient of variation of the detected data is greater than 0.65;

[0049] In step (3), parameter a can be 0.6, parameter b can be 4, and parameter d can be 0.8, which is equivalent to changing the sampling interval to 250ms. When the flow rate fluctuates significantly, a random sample is taken every 250ms. When the average value of the 30 detection data is less than 0.014m... 3 When the speed is 1 / s, the sample standard deviation of the detection data is less than 0.0312m. 3 When / s, exit the sampling mode of step (3) and sample once randomly every 1 second; when the average value of 30 detection data is greater than 0.014m 3 When the coefficient of variation of the detected data is less than 0.52 at / s, exit the sampling mode of step (3) and sample once in a random manner every 1 second.

[0050] This embodiment employs random sampling, ensuring that the sampling points change continuously within each fixed time interval, thus more accurately tracking flow rate variations. The trapezoidal method is used to calculate cumulative flow, further reducing the impact of flow fluctuations on measurement accuracy when flow rates change slowly. To minimize the impact of large flow rate variations on measurement accuracy, this embodiment uses the average value and sample standard deviation of the past h consecutive instantaneous flow rate data to determine flow rate fluctuations. When large flow rate fluctuations occur, the fixed sampling interval is shortened, increasing the sampling frequency. When the flow rate returns to normal, it reverts to normal sampling mode, reducing power consumption.

Claims

1. A method for improving the metering accuracy of an electronic gas meter, characterized in that: The sampling time is determined by generating random numbers within a single fixed time interval, thus sampling once within each fixed time interval. The average value and standard deviation of the instantaneous flow rate data within the detection period are used to determine whether there are significant fluctuations in flow rate. The step of setting the detection period includes: [The text abruptly ends here, so the translation stops as well.] Each set of test data is used as a test cycle. It is an integer, and the first detection period is the [number]. The first instantaneous flow data, the second detection cycle is the first... Instantaneous flow data; The cumulative flow within the sampling interval is calculated using a trapezoidal algorithm, which compares the instantaneous flow at the current time with the instantaneous flow at the previous sampling time. The cumulative flow within each sampling interval is: ,in It is the cumulative flow rate within a sampling interval. It is the first Next sampling time yes Next sampling time yes Instantaneous flow rate at the sampling time, yes Instantaneous flow rate at the sampling time; When significant fluctuations in traffic occur, shorten the fixed time interval and increase the sampling frequency; the steps of shortening the fixed time interval and increasing the sampling frequency include: a fixed time interval, It is an integer; a random number is generated within each time interval, and the sampling time is determined by the value of the random number; it is assumed that this step has already been performed. If the time is seconds, then the next sampling time is... Generate random numbers ,sampling Instantaneous flow rate at any moment If the flow rate continues to fluctuate significantly, the next sampling time will be... Generate random numbers ,sampling Instantaneous flow rate at any moment Cumulative flow rate within the sampling interval The trapezoidal algorithm is still used, that is After each sampling, the data is still based on the past. Each test data point constitutes one testing period. The mean and sample standard deviation are obtained to determine whether the flow rate is still fluctuating significantly.

2. The method for improving the metering accuracy of an electronic gas meter according to claim 1, characterized in that... Includes the following steps: Step (1): Set a fixed time interval and detection cycle, and record instantaneous flow data. During normal measurement processes, A fixed time interval of seconds is used to generate random numbers within each interval. The first sampling period is... Generate random numbers between ,exist At the sampling time, the instantaneous flow rate is obtained. The second sampling period is in Generate random numbers between ,get Instantaneous flow rate at the sampling time ;No. Each sampling period is in Generate random numbers ,get Instantaneous flow rate at the sampling time ; In step (1) above, the cumulative flow rate within each sampling interval is calculated using a trapezoidal algorithm, i.e., the cumulative flow rate within the first sampling interval is calculated as follows: , Cumulative flow rate within the second sampling interval , ..., No. Cumulative flow rate within each sampling interval ; Step (2): Determine if there are large fluctuations in flow rate during the detection period. Set the threshold for the average instantaneous flow rate within the detection period. Threshold for sample standard deviation Threshold for coefficient of variation When the average value of the detected data is less than the average value threshold The standard deviation of the sample is used to determine whether there are large flow fluctuations; when the average value of the test data is greater than the average value threshold... The coefficient of variation, which is the ratio of the sample standard deviation to the mean, is used to determine whether there are large fluctuations in flow. Step (3): When the flow rate fluctuates greatly, shorten the fixed time interval and increase the sampling frequency.

3. The method for improving the metering accuracy of an electronic gas meter according to claim 2, characterized in that... In step (2), the average instantaneous flow rate is measured in the first detection cycle. Sample standard deviation of instantaneous flow rate The following situations exist: ① If the average value of instantaneous flow rate detection data Average threshold Sample standard deviation of the test data Sample standard deviation threshold At this point, it is determined that there is no significant fluctuation in flow rate, and the process continues according to step (1). Accumulated flow is calculated using a sampling method with fixed second intervals. ②If the average value of instantaneous flow rate detection data Sample standard deviation of the test data At this point, it is determined that there is a large fluctuation in traffic, and the process proceeds to step (3). ③ If the average value of instantaneous flow rate detection data And the coefficient of variation of the detection data Coefficient of variation threshold ,in At this point, it is determined that there is no significant fluctuation in flow rate, and the process continues according to step (1). Accumulated flow is calculated using a sampling method with fixed second intervals. ④ If the average value of instantaneous flow rate detection data And the coefficient of variation of the detection data At this point, it is determined that there is a large fluctuation in traffic, and the process proceeds to step (3).

4. The method for improving the metering accuracy of an electronic gas meter according to claim 2, characterized in that... In step (3), with The corresponding detection period for each moment is explained, and within the current detection period... The instantaneous flow data is , … ,in, yes Instantaneous flow rate at the sampling time yes Instantaneous flow rate at the sampling time The mean of the instantaneous flow data Sample standard deviation The following situations exist: 1) If the average value of instantaneous flow rate detection data Sample standard deviation threshold Sample standard deviation of the test data ,in This indicates that the flow rate is slowly recovering to a stable state. Exit the sampling method in step (3) and proceed according to the steps in step (1). Accumulated flow is calculated using a sampling method with fixed second intervals. 2) If the average value of the instantaneous flow rate detection data Sample standard deviation of the test data At this point, the flow rate still fluctuates significantly, so continue following step (3)... Accumulated flow is calculated using a fixed time interval sampling method; 3) If the average value of the instantaneous flow rate detection data And the coefficient of variation of the detection data ,in ,in This indicates that the flow rate is slowly recovering to a stable state. Exit the sampling method in step (3) and proceed according to the steps in step (1). Accumulated flow is calculated using a sampling method with fixed second intervals. 4) If the average value of the instantaneous flow rate detection data And the coefficient of variation of the detection data At this point, the flow rate still fluctuates significantly, so continue following step (3)... The cumulative flow is calculated using a fixed time interval sampling method.

Images