A small hydropower frequency prediction method based on weighted distribution

Abstract

The application provides a small hydropower frequency prediction method based on weighted distribution; the same prediction interval and standard interval values are set; the frequency of the prediction interval after the current time point is calculated from the reference frequency of the standard interval of the small hydropower operation record; the duration of the small hydropower operation is divided into n standard intervals A; then the weighted value K corresponding to each standard interval A is calculated; and then the average frequency E of the prediction interval is calculated. The value range of the value n is set; the minimum value of the value n is 2, because the values of the standard interval A and the prediction interval C are the same; the prediction frequency of the small hydropower operation in the prediction interval C is calculated from the reference frequency of the small hydropower operation in more than two standard intervals A; if n is not an integer, n is rounded down to an integer, and the complete standard interval is used for calculation; the accuracy is high; the maximum value of the value n is set as 10, the result reliability is improved, and the calculation amount is reduced.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and specifically to a method for predicting the frequency of small hydropower based on weighted allocation. Background Technology

[0002] The problem of grid frequency fluctuations is particularly prominent in small hydropower. Existing frequency regulation methods for small hydropower include one method that adjusts based on frequency changes. However, due to the inertia of the turbine, this method cannot control the frequency within a stable range in a timely manner. Another method is to predict the frequency based on a large amount of data and adjust it in advance. This method usually requires a large amount of memory space, but the control equipment of small hydropower generally does not have sufficient storage space, resulting in a lack of accuracy in this type of regulation method. Summary of the Invention

[0003] This invention provides a method for predicting the frequency of small hydropower stations based on weighted allocation, which accurately predicts the frequency of small hydropower stations based on their current operating time.

[0004] To achieve the above objectives, the technical solution of the present invention is: a method for predicting the frequency of small hydropower based on weighted allocation, comprising the following steps;

[0005] Step (1). Take the last frequency of the last second in each minute of the small hydropower operation as the reference frequency for each minute; record the current quantity of the reference frequency generated by the small hydropower operation; store the value of each reference frequency generated each minute.

[0006] Step (2). Set the values ​​of prediction interval C and standard interval A; prediction interval C is the time period of the frequency to be predicted within X minutes after the current time point; standard interval A is the time period that is close to the time axis of prediction interval C and has a reference frequency; prediction interval C, standard interval A and X have the same value.

[0007] Step (3). Divide the current quantity value of the reference frequency by the standard interval A to obtain the value n; the value n is the quantity of the standard interval A.

[0008] Step (4). Determine if the value n satisfies 10≥n≥2; if yes, proceed to step (4.1); if no, proceed to step (2) to reconfirm the values ​​of C, A and X.

[0009] Step (4.1). Determine if the value n is an integer; if yes, proceed to step (5); if no, round down the value n to the nearest integer and then proceed to step (5).

[0010] Step (5). Using formula k n-1 =10 （1-n） *5 （n-1） And the formula k1+k2+k n-1+k n =1; calculate the weighted value K corresponding to each standard interval A.

[0011] Step (6). Determine k n-1 =k n Is it true? If yes, proceed to step (7); if no, proceed to step (2) to reconfirm the values ​​of C, A, and X; k n It is the weighted value corresponding to the farthest standard interval A on the time axis from the prediction interval C.

[0012] Step (7). Calculate the reference frequency value and f in each standard interval A.

[0013] Step (8). Using the formula F=F=k1*f1+k2*f2+k n-1 *f n-1 + k n *f n ; Calculate the total frequency F of the prediction interval.

[0014] Step (9). Calculate the average frequency E of the prediction interval using the formula E=F / A.

[0015] The above method generates a reference frequency every minute during small hydropower operation. First, the number of reference frequencies is recorded, thus determining the duration of small hydropower operation. Then, the value of prediction interval C is determined, along with its duration. Next, the duration of small hydropower operation is divided into n standard intervals A by subtracting the current value of the reference frequencies from the standard interval A. Then, the weighted value K corresponding to each standard interval A is calculated. Finally, the average frequency E of the prediction interval is calculated. The range of values ​​for n is set; the minimum value of n is set to 2, since the values ​​of standard interval A and prediction interval C are the same. The predicted frequency of small hydropower operation in prediction interval C is calculated using the reference frequencies of small hydropower operation in two or more standard intervals A. If n is not an integer, it is rounded down to the nearest integer, using the complete standard interval for calculation; this method ensures high accuracy. Simultaneously, the maximum value of n is set to 10, improving the reliability of the results while reducing the computational load. Then, the weighted value k corresponding to each standard interval A is calculated. The average frequency E of the prediction interval is calculated using the same weighted value k. n-1 =k n It can guarantee the accuracy of the predicted frequency.

[0016] Furthermore, step (9) is followed by steps (10) to (12).

[0017] Step (10). Determine whether E≤49.5Hz is true; if yes, increase the power of the small hydropower station; if no, proceed to step (11).

[0018] Step (11). Determine whether E≥50.2Hz is true; if yes, reduce the power of the small hydropower station; if no, proceed to step (12).

[0019] Step (12). Do not adjust the power of the small hydropower station.

[0020] Furthermore, the values ​​of C, A, and X are greater than 1 and less than 30.

[0021] Furthermore, step (1) also includes: setting a threshold for the reference frequency; if the current quantity value of the reference frequency is greater than the threshold for the reference frequency; calculating the difference z between the current quantity value of the reference frequency and the threshold for the reference frequency; and clearing z reference frequencies starting from the time point furthest from the prediction interval C.

[0022] The above method clears z reference frequencies; keeps the current value of the reference frequency equal to the threshold of the reference frequency; reduces the storage capacity requirements of small hydropower equipment; avoids too many reference frequency values, which would increase the amount of calculation; and starts from the time point farthest from the prediction interval C, so that the time period for generating the reference frequency is close to the prediction interval C, making the predicted frequency more accurate.

[0023] Further, after completing step (2), proceed to step (3.1); then proceed to steps (4)-(9).

[0024] Step (3.1): Divide the standard interval A by the quantity value of a portion of the reference frequencies in the currently stored reference frequencies to obtain the value n; the value n is the quantity of the standard interval A.

[0025] Furthermore, according to the timeline, the standard interval closest to the prediction interval C is the first standard interval A1; the standard interval farthest from the prediction interval C is the nth standard interval A1. n In step (7), the reference frequency value and f in each standard interval are calculated starting from the first standard interval A1. Attached Figure Description

[0026] Figure 1 The flowchart is for the invention. Detailed Implementation

[0027] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0028] like Figure 1 As shown; a method for predicting the frequency of small hydropower based on weighted allocation, comprising the following steps;

[0029] Step (1). Take the last frequency of the last second in each minute of the small hydropower operation as the reference frequency for each minute; record the current quantity of the reference frequency generated by the small hydropower operation; store the value of each reference frequency generated each minute.

[0030] Step (2). Set the values ​​of prediction interval C and standard interval A; prediction interval C is the time period of the frequency to be predicted within X minutes after the current time point; standard interval A is the time period close to the time axis of prediction interval C and has a reference frequency; prediction interval C, standard interval A and X have the same value. In this implementation, the values ​​of C, A and X are greater than 1 and less than 30.

[0031] Step (3). Divide the current quantity value of the reference frequency by the standard interval A to obtain the value n; the value n is the quantity of the standard interval A.

[0032] Step (4). Determine if the value n satisfies 10≥n≥2; if yes, proceed to step (4.1); if no, proceed to step (2) to reconfirm the values ​​of C, A and X.

[0033] Step (4.1). Determine if the value n is an integer; if yes, proceed to step (5); if no, round down the value n to the nearest integer and then proceed to step (5).

[0034] Step (5). Using formula k n-1 =10 （1-n） *5 （n-1） And the formula k1+k2+k n-1 +k n =1; calculate the weighted value K corresponding to each standard interval A.

[0035] Step (6). Determine k n-1 =k n Is it true? If yes, proceed to step (7); if no, proceed to step (2) to reconfirm the values ​​of C, A, and X; k n It is the weighted value corresponding to the farthest standard interval A on the time axis from the prediction interval C.

[0036] Step (7). Calculate the reference frequency value and f in each standard interval A.

[0037] Step (8). Using the formula F=k1*f1+k2*f2+k n-1 *f n-1 + k n *f n ; Calculate the total frequency F of the prediction interval.

[0038] Step (9). Calculate the average frequency E of the prediction interval using the formula E=F / A.

[0039] The above method generates a reference frequency every minute during small hydropower operation. First, the number of reference frequencies is recorded, thus determining the duration of small hydropower operation. Then, the value of prediction interval C is determined, along with its duration. Next, the duration of small hydropower operation is divided into n standard intervals A by subtracting the current value of the reference frequencies from the standard interval A. Then, the weighted value K corresponding to each standard interval A is calculated. Finally, the average frequency E of the prediction interval is calculated. The range of values ​​for n is set; the minimum value of n is set to 2, since the values ​​of standard interval A and prediction interval C are the same. The predicted frequency of small hydropower operation in prediction interval C is calculated using the reference frequencies of small hydropower operation in two or more standard intervals A. If n is not an integer, it is rounded down to the nearest integer, using the complete standard interval for calculation; this method ensures high accuracy. Simultaneously, the maximum value of n is set to 10, improving the reliability of the results while reducing the computational load. Then, the weighted value k corresponding to each standard interval A is calculated. The average frequency E of the prediction interval is calculated using the reference frequencies of the same value k. n-1 =k n It can guarantee the accuracy of the predicted frequency.

[0040] In the above methods:

[0041] Step (9) is followed by steps (10) - (12).

[0042] Step (10). Determine whether E≤49.5Hz is true; if yes, increase the power of the small hydropower station; if no, proceed to step (11).

[0043] Step (11). Determine whether E≥50.2Hz is true; if yes, reduce the power of the small hydropower station; if no, proceed to step (12).

[0044] Step (12). Do not adjust the power of the small hydropower station.

[0045] Step (1) further includes: setting a threshold for the reference frequency; if the current quantity value of the reference frequency is greater than the threshold for the reference frequency; calculating the difference z between the current quantity value of the reference frequency and the threshold for the reference frequency; clearing z reference frequencies starting from the time point furthest from the prediction interval C. If the threshold for the reference frequency is 300; the current quantity value of the reference frequency is 330; and the difference between the current quantity value of the reference frequency and the threshold for the reference frequency is calculated to be 30; then clearing the recorded reference frequencies that are 30 minutes furthest from the prediction interval C.

[0046] The above method clears z reference frequencies; keeps the current value of the reference frequency equal to the threshold of the reference frequency; reduces the storage capacity requirements of small hydropower equipment; avoids too many reference frequency values, which would increase the amount of calculation; and starts from the time point farthest from the prediction interval C, so that the time period for generating the reference frequency is close to the prediction interval C, making the predicted frequency more accurate.

[0047] In this embodiment, according to the time axis, the standard interval closest to the prediction interval C is the first standard interval A1; the standard interval furthest from the prediction interval C is the nth standard interval A. n The frequency of prediction interval C is calculated starting from a standard interval close to the prediction interval C, resulting in high accuracy. In step (5), k n k is the weighted value of the standard interval furthest from the prediction interval C; n-1 This is the weighted value of the standard interval preceding the standard interval that is furthest from the prediction interval C. This is determined by judging k. n-1 =k n To determine if the prediction is valid, compare the weighted values ​​of the two standard intervals furthest from the prediction interval C. If they are the same, the accuracy of the prediction frequency can be guaranteed.

[0048] In this implementation, 10 reference frequency values ​​were previously stored; for example, it is necessary to predict the frequency within the next 5 minutes.

[0049] Perform step (2), set the values ​​of prediction interval C and standard interval A to 5; predict the frequency within 5 minutes after the current time point.

[0050] In step (3), the value of the current reference frequency 10 is divided by the value of the standard interval A to obtain the value of n 2; the number of standard intervals A is 2.

[0051] Proceed to step (4), where the value of n satisfies 10 ≥ n ≥ 2.

[0052] Proceed to step (4.1) to determine if n is an integer.

[0053] Proceed to step (5) to calculate that the values ​​of k1 and k2 are both 0.5.

[0054] Proceed to step (6), K n-1 =K n Established.

[0055] Perform step (7) to calculate the standard interval A1, which is the sum of the reference frequencies from the 5th to the 10th minute and f1; calculate the standard interval A2, which is the sum of the reference frequencies from the 0th to the 5th minute and f2. The weight corresponding to f1 is k1, and the weight corresponding to f2 is k2.

[0056] Proceed to steps (8)-(12).

[0057] Previously, 10 reference frequency values ​​were stored; for example, we need to predict the frequency within the next 4 minutes.

[0058] Perform step (2), set the values ​​of prediction interval C and standard interval A to 4; predict the frequency within 4 minutes after the current time point.

[0059] In step (3), the value of the current reference frequency 10 is divided by the value of the standard interval A to obtain the value of n, which is 2.5; the number of standard intervals A is 2.5.

[0060] Proceed to step (4), where the value of n satisfies 10 ≥ n ≥ 2.

[0061] Proceed to step (4.1), determine if n is not an integer, and round n down to the nearest integer; that is, take n as 2; thus, the number of standard intervals A is 2. Proceed to step (5), calculate that the values ​​of k1 and k2 are both 0.5.

[0062] Proceed to step (6), K n-1 =K n Established.

[0063] Perform step (7) to calculate the standard interval A1, which is the value of the reference frequency from the 7th to the 10th minute and f1; calculate the standard interval A2, which is the value of the reference frequency from the 3rd to the 6th minute and f2. The weight corresponding to f1 is k1, and the weight corresponding to f2 is k2.

[0064] Proceed to steps (8)-(12).

[0065] In the above method, step (2) further includes: after completing step (2), proceed to step (3.1); then proceed to steps (4)-(9).

[0066] Step (3.1): Divide the standard interval A by the quantity value of a portion of the reference frequencies in the currently stored reference frequencies to obtain the value n; the value n is the quantity of the standard interval A.

[0067] Take, for example, a database containing 300 reference frequencies; and the task of predicting the frequency over the next 30 minutes.

[0068] Perform step (2), set the values ​​of prediction interval C and standard interval A to 30; predict the frequency within 30 minutes after the current time point.

[0069] Perform step (3.1), using a portion of the currently stored reference frequency 300; these can be 60, 90, 120, 150, 210, etc.; if the value of the portion of the reference frequency is 90; remove the value of the standard interval A to obtain the value of n, which is 3; the number of standard intervals A is 3.

[0070] Proceed to step (4), where the value of n satisfies 10 ≥ n ≥ 2.

[0071] Proceed to step (4.1) to determine if n is an integer.

[0072] Perform step (5) to calculate the value of k1 as 0.5; the value of k2 as 0.25; and the value of k3 as 0.25.

[0073] Proceed to step (6), K n-1 =K n Established.

[0074] Perform step (7) to calculate the standard interval A1, which is the reference frequency value and f1 within the 270th to 300th minute; calculate the standard interval A2, which is the reference frequency value and f2 within the 240th to 270th minute; calculate the standard interval A3, which is the reference frequency value and f3 within the 210th to 240th minute. The weight corresponding to f1 is k1, the weight corresponding to f2 is k2, and the weight corresponding to f3 is k3.

[0075] Proceed to steps (8)-(12).

[0076] If in step (3.1), the value of some reference frequencies is 150; after removing the value of standard interval A, the value of n is 5; the number of standard intervals A is 5.

[0077] Proceed to step (4), where the value of n satisfies 10 ≥ n ≥ 2.

[0078] Proceed to step (4.1) to determine if n is an integer.

[0079] Perform step (5) to calculate the value of k1 as 0.5; the value of k2 as 0.25; the value of k3 as 0.125; the value of k4 as 0.0625; and the value of k5 as 0.0625.

[0080] Proceed to step (6), K n-1 =K n Established.

[0081] Perform step (7) to calculate the standard interval A1, i.e., the reference frequency values ​​and f1 within the 270th to 300th minute; calculate the standard interval A2, i.e., the reference frequency values ​​and f2 within the 240th to 270th minute; calculate the standard interval A3, i.e., the reference frequency values ​​and f3 within the 210th to 240th minute; calculate the standard interval A4, i.e., the reference frequency values ​​and f4 within the 180th to 210th minute; calculate the standard interval A5, i.e., the reference frequency values ​​and f5 within the 150th to 180th minute. The weight corresponding to f1 is k1, the weight corresponding to f2 is k2, the weight corresponding to f3 is k3, the weight corresponding to f4 is k4, and the weight corresponding to f5 is k5.

[0082] In the above method, if the currently stored reference frequency value is 300, two, three, four, five, six, seven, eight, nine, or ten standard intervals can be used to predict the frequency within 30 minutes after the current time point; the more standard intervals used, the higher the accuracy of the predicted frequency.

Claims

1. A method for frequency prediction of small hydropower based on weighted allocation, characterized in that: It comprises the following steps: Step (1). Take the last frequency of the last second in each minute as the reference frequency of each minute; Record the current value of the reference frequency generated by the small hydropower operation to obtain the duration of the small hydropower operation; Respectively store the value of each reference frequency generated in each minute; Step (2). Set the values of the prediction interval C and the standard interval A; the prediction interval C is the time period of the frequency to be predicted within X minutes after the current time point; the standard interval A is the time period close to the time axis of the prediction interval C and has the reference frequency; the prediction interval C, the standard interval A and the value of X are the same; Step (3). Divide the duration of the small hydropower operation into n standard intervals A using the value of part of the reference frequency currently stored in the reference frequency and the standard interval A, to obtain the value n; the value n is the number of standard intervals A, which divides the duration of the small hydropower operation into n standard intervals A; after step (9), steps (10)-(12) are included; Step (4). Determine whether the value n satisfies 10≥n≥2; if yes, proceed to step (4.1); if no, proceed to step (2) to reconfirm the values of C, A and X; Step (4.1). Determine whether the value n is an integer; if yes, proceed to step (5); if no, the value n is rounded down to an integer, and then proceed to step (5); Step (5). Calculate the weighted value K corresponding to each standard interval A by the formula k n-1 =10 （1-n） *5 （n-1） and the formula k1+k2+k n-1 +k n =1; Step (6). Determine whether k n-1 = k n is true; if yes, proceed to step (7); if no, proceed to step (2) to reconfirm the values of C, A and X; k n is the farthest standard interval A corresponding weighted value on the time axis of the distance prediction interval C. Step (7). Calculate the value and f of the reference frequency in each standard interval A, respectively; Step (8). Calculate the total frequency F of the prediction interval by the formula F = k1*f1 + k2*f2 + k n-1 *f n-1 + k n *f n ; Step (9). Calculate the average frequency E of the prediction interval by the formula E=F / A; After completing step (2), proceed to step (3); then proceed to steps (4)-(9); Step (10). Determine whether E≤49.5Hz is true; if yes, increase the power of the small hydropower; if no, proceed to step (11); Step (11). Determine whether E≥50.2Hz is true; if yes, decrease the power of the small hydropower; if no, proceed to step (12); Step (12). Do not adjust the power of the small hydropower.

2. A frequency prediction method for small hydropower plants based on weighted allocation according to claim 1, characterized in that: The values of C, A and X are greater than 1 and less than 30.

3. A frequency prediction method for small hydropower plants based on weighted allocation according to claim 1, characterized in that: According to the time axis, the standard interval closest to the prediction interval C is the first standard interval A1, and the standard interval farthest from the prediction interval C is the nth standard interval A n ; in step (7), the sum f of the reference frequencies in each standard interval is calculated starting from the first standard interval A1.

Images