Power system rapid frequency detection method based on multi-level set

Abstract

The invention discloses a power system rapid frequency detection method based on a multi-level set, and relates to the technical field of power system frequency detection. According to the method, time domain sampling is selected for signal collection, a signal with a limited frequency spectrum is uniquely represented through sampling values at equal intervals, the aliasing phenomenon in a frequency domain is avoided, and the accuracy and the real-time performance of frequency measurement of the power system are improved; the filter is used for limiting the generation and transmission of harmonic waves, so that the current waveform is smoother, rapid frequency detection can be carried out under the condition of strong distortion of the electrical quantity waveform, and the accuracy is relatively high; when additional interpolation points are added in the Newton interpolation method, the Newton interpolation method can utilize the previous operation result, so that the operand is reduced, the calculation efficiency is improved, and the method has the advantage of high efficiency; when interpolation nodes are increased or decreased, only parts related to newly added points need to be calculated, so that the method is very convenient to use when the number of the nodes is changed.

Description

Technical Field

[0001] The present invention relates to the technical field of power system frequency detection, and particularly to a fast frequency detection method for power systems based on multi-level sets. Background Art

[0002] A power system is an electric energy production and consumption system composed of links such as power generation, transformation, transmission, distribution, and power consumption. Its function is to convert primary energy in nature into electric energy through power generation power devices (mainly including boilers, steam turbines, generators, and auxiliary production systems of power plants, etc.), and then supply the electric energy to each load center through the transmission and transformation systems and the distribution system. Since most power source points and load centers are in different regions and cannot be stored in large quantities, power production must always maintain a balance with consumption. Therefore, the centralized development and decentralized use of electric energy, as well as the continuous supply of electric energy and the random changes of loads, restrict the structure and operation of the power system.

[0003] The main frequency measurement algorithms include: the period method and its improved algorithm methods, the analytical method, the Fourier algorithm, the signal demodulation method, etc. The period method and its improved algorithm methods mainly include the level intersection method, the zero-crossing detection method, the high-order correction function method, etc. Their advantage is better real-time performance, and the disadvantage is relatively low accuracy. The analytical method uses a simple signal model for observation, and there is an approximation process in the derivation of the algorithm, which is only suitable for occasions with low requirements for both accuracy and real-time performance. For the Fourier algorithm, it is difficult to achieve both high measurement accuracy and low computational complexity, and the fence effect and spectral leakage will also reduce the accuracy of frequency measurement calculations. The signal demodulation method requires precise filtering technology to cooperate with it. At the same time, it is necessary to avoid the interference of noise and the interference of impact frequency dynamics. The above measurement methods all have limitations. Therefore, we propose a fast frequency detection method for power systems based on multi-level sets. Summary of the Invention

[0004] The purpose of the present invention is to provide a fast frequency detection method for power systems based on multi-level sets to solve the problems mentioned in the above background art.

[0005] The present invention specifically adopts the following technical solutions to achieve the above purpose:

[0006] A fast frequency detection method for power systems based on multi-level sets, comprising:

[0007] Setting multiple thresholds;

[0008] Collecting voltage signals of the power system;

[0009] Filtering the sampled voltage signals;

[0010] Calculating the frequency values corresponding to each voltage signal respectively for the filtered voltage signals;

[0011] Combine and weight the frequency values to obtain the weighted average frequency;

[0012] Determine whether the weighted average frequency coincides with the threshold. If so, directly output the weighted average frequency. If not, calculate the frequency result using Newton interpolation method.

[0013] Furthermore, the set threshold is based on the historical data of the power system.

[0014] Furthermore, the electrical quantity signals of the power system are collected by time-domain sampling for signal acquisition.

[0015] Furthermore, a low-pass filter is selected for filtering work.

[0016] Furthermore, the expression of the Newton interpolation method is: f(x) = Pn(x) + Rn(x), where Pn(x) is the Newton interpolation polynomial and Rn(x) is the Newton interpolation remainder.

[0017] Furthermore, the expressions of Pn(x) and Rn(x) are shown as follows:

[0018] P n (x) = f[x 0 + f[x 0 , x 1 (x - x 0 ) + f[x 0 , x 1 , x 2 )x - x 0 )(x - x 1 ) + … + f[x 0 , x 1 …x n (x - x 0 )(x - x n-1 )

[0019]

[0020] Furthermore, the calculation formula of the weighted average method is as follows: Y = ∑(Wi * Xi) / ∑Wi,

[0021] where Y represents the result to be calculated;

[0022] Wi represents the weight of each Xi, representing the frequencies in different time periods and different regions in the power grid;

[0023] Xi represents the numerical values of various parameters, representing the frequencies in different time periods and different regions in the power grid;

[0024] The symbol ∑ represents the summation symbol, that is, it sums up all terms.

[0025] A computer-readable storage medium includes a memory, a processor, and a computer program stored on the memory. When the computer program is run by the processor, it executes instructions according to the above-mentioned method.

[0026] A fast frequency detection device for a power system based on multi-level sets. The detection device includes: a signal acquisition module, a signal processing module, and a data storage module. Among them, the signal acquisition module is used to collect voltage signals, the signal processing module is used to preprocess the voltage signals, and the data storage module is used to store the voltage information collected by the signal acquisition module.

[0027] Further, the signal processing module includes a filtering unit and a duplicate removal unit. Among them, the filtering unit is used to filter power systems that do not meet the requirements, and the duplicate removal unit is used to remove duplicate power systems.

[0028] The beneficial effects of the present invention are as follows:

[0029] 1. In the present invention, time-domain sampling is selected for signal acquisition. A spectrally limited signal is uniquely represented by equally spaced sampled values, and aliasing in the frequency domain is avoided, improving the accuracy and real-time performance of power system frequency measurement.

[0030] 2. The present invention uses a filter to limit the generation and transmission of harmonics, making the current waveform smoother. It can perform fast frequency detection under strong distortion of electrical quantity waveforms with high accuracy.

[0031] 3. The Newton interpolation method of the present invention can quickly construct an interpolation polynomial. Its formula structure is simple and easy to calculate, with the advantages of simplicity and directness; when adding additional interpolation points, the Newton interpolation method can utilize the previous operation results, thereby reducing the amount of calculation and increasing the calculation frequency, with the advantage of high efficiency; in theoretical analysis and actual programming calculations, the logic is clear, easy to understand and implement, with the advantage of clear logic; the Newton interpolation method can fit discrete points to obtain relatively accurate function analytical values, with the advantage of wide applicability; when the interpolation nodes are increased or decreased, the Newton interpolation method does not need to recalculate the entire polynomial, but only needs to calculate the part related to the new points, which makes it very convenient to use when the number of nodes changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is the working flow chart of the present invention;

[0033] Figure 2 is the working flow chart of setting the threshold value in the present invention;

[0034] Figure 3 This is the working block diagram of the detection device in the present invention. Detailed implementation mode

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0036] Please refer to Figure 1 - Figure 2 The present invention provides a fast frequency detection method for a power system based on multi-level sets, including:

[0037] Setting multiple thresholds;

[0038] Collecting voltage signals of the power system;

[0039] Filtering the sampled voltage signals;

[0040] For the filtered voltage signals, calculating the frequency values corresponding to each voltage signal respectively;

[0041] Combining and weighting the frequency values to obtain a weighted average frequency;

[0042] Judging whether the weighted average frequency coincides with the threshold. If so, directly output the weighted average frequency. If not, calculate the frequency result using Newton interpolation method.

[0043] In this embodiment, preferably, the thresholds are set based on the historical data of the power system; a calculation model is constructed according to the historical data. The calculation model uses a multi-layer perceptron to build a neural network, where the input layer receives feature data, and the output layer outputs the state quantity information of the electromechanical system. The intermediate hidden layer extracts complex features. After the construction of the calculation model, it is trained. The constructed neural network is trained using the processed historical data of the power system to optimize the network weights and bias parameters, so that the output result is as close as possible to the actual value. The trained model is used in a new data set to observe the coincidence degree between its prediction result and the actual value.

[0044] In this embodiment, preferably, when collecting the electrical quantity signals of the power system, time-domain sampling is used for signal acquisition; a spectrally limited signal is uniquely represented by equally spaced sampled values, and aliasing phenomena in the frequency domain are avoided, improving the accuracy and real-time performance of power system frequency measurement. Select the sampling frequency fs and the sampling length N, and sample the signal at equal intervals. The sampling frequency fs should be not less than 2 times the highest harmonic component frequency contained in the signal.

[0045] In this embodiment, preferably, a low-pass filter is used for filtering work; a filter is used to limit the generation and transmission of harmonics, making the current waveform smoother.

[0046] In this embodiment, preferably, the expression of Newton interpolation method is: f(x) = Pn(x) + Rn(x), where Pn(x) is the Newton interpolation polynomial and Rn(x) is the Newton interpolation remainder.

[0047] In this embodiment, preferably, the expressions of Pn(x) and Rn(x) are as shown in the following formula:

[0048] P n (x) = f[x 0 + f[x 0 , x 1 (x - x 0 ) + f[x 0 , x 1 , x 2 )x - x 0 )(x - x 1 ) + … + f[x 0 , x 1 …x n (x - x 0 )(x - x n-1 )

[0049]

[0050] In this embodiment, preferably, the calculation formula of the weighted average method is as follows: Y = ∑(Wi * Xi) / ∑Wi,

[0051] where Y represents the result to be calculated;

[0052] Wi represents the weight of each Xi, which represents the frequencies in different time periods and different regions in the power grid;

[0053] Xi represents the numerical values of various parameters, which represents the frequencies in different time periods and different regions in the power grid;

[0054] ∑ represents the summation symbol, that is, summing up all terms.

[0055] A computer-readable storage medium includes a memory, a processor, and a computer program stored on the memory. When the computer program is run by the processor, it executes the instructions according to the above method.

[0056] Please refer to Figure 3 , a fast frequency detection device for a power system based on multi-level sets. The detection device includes: a signal acquisition module, a signal processing module, and a data storage module. Among them, the signal acquisition module is used to collect voltage signals, the signal processing module is used to preprocess the voltage signals, and the data storage module is used to store the voltage information collected by the signal acquisition module.

[0057] The signal processing module includes a filtering unit and a duplicate removal unit. Among them, the filtering unit is used to filter power systems that do not meet the requirements, can quickly determine the power systems to be inspected, and avoid errors; the duplicate removal unit is used to remove duplicate power systems; it can increase the working frequency and reduce the error rate.

[0058] The working principle and usage process of the present invention are as follows:

[0059] Set multiple thresholds based on the historical data of the power system; it includes the following steps:

[0060] Construct a calculation model. Build a calculation model according to the historical data. The calculation model uses a multi-layer perceptron to build a neural network. Among them, the input layer receives feature data, and the output layer outputs the state quantity information of the electromechanical system. The intermediate hidden layer extracts complex features;

[0061] Model training. After completing the construction of the calculation model, train it. Use the processed historical data of the power system to train the built neural network to optimize the network weights and bias parameters so that the output result is as close as possible to the actual value;

[0062] Model verification. Use the trained model in a new data set and observe the degree of agreement between its prediction result and the actual value.

[0063] Collect the voltage signals of the power system. Time-domain sampling is selected for signal acquisition; a spectrally limited signal is uniquely represented by equally spaced sampled values, and aliasing in the frequency domain is avoided. Select the sampling frequency fs and the sampling length N, and perform equally spaced sampling on the signal. The sampling frequency fs should be no less than 2 times the frequency of the highest harmonic component contained in the signal.

[0064] Filter the sampled voltage signals to eliminate the influence of high-order harmonics and noise on the signals; select a low-pass filter for filtering work, and use the filter to limit the generation and transmission of harmonics to make the current waveform smoother.

[0065] For the voltage signals after low-pass filtering, calculate the frequency values corresponding to each voltage signal respectively.

[0066] Perform combined weighting on the frequency values. There are n different frequencies f1, f2,..., fn in the power grid, and their corresponding values (such as power or energy) are P1, P2,..., Pn respectively. The calculation formula for the weighted average frequency is: weighted average frequency = (f1*P1 + f2*P2 +... + fn*Pn) / (P1 + P2 +... + Pn), and the weighted average frequency is obtained.

[0067] Determine whether the weighted average frequency coincides with the threshold. If so, directly output the weighted average frequency. If not, calculate the frequency result using Newton interpolation method. The expression of Newton interpolation method is: f(x) = Pn(x) + Rn(x), where Pn(x) is the Newton interpolation polynomial and Rn(x) is the Newton interpolation remainder.

[0068] The expressions of Pn(x) and Rn(x) are shown as follows:

[0069] P n (x) = f[x 0 + f[x 0 , x 1 (x - x 0 ) + f[x 0 , x 1 , x 2 )x - x 0 (x - x 1 ) + … + f[x 0 , x 1 …x n (x - x 0 )(x - x n-1 )

[0070]

[0071] The frequency result is directly calculated to obtain the final fundamental frequency estimation value based on the weighted average method. The calculation formula of the weighted average method is as follows: Y = ∑(Wi * Xi) / ∑Wi,

[0072] where Y represents the result to be calculated;

[0073] Wi represents the weight of each Xi, which represents the frequencies in different time periods and different regions in the power grid;

[0074] Xi represents the values of various parameters, which represents the frequencies in different time periods and different regions in the power grid;

[0075] ∑ represents the summation symbol, that is, summing up all terms.

[0076] Store the data after the detection is completed and generate a detection report.

[0077] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A fast frequency detection method for power system based on multi-level sets, characterized in that: include: Set multiple thresholds; Collect voltage signals of power systems; Filtering the sampled voltage signal; For the filtered voltage signals, the frequency values ​​corresponding to the voltage signals are calculated respectively; Combine and weight the frequency values ​​to obtain a weighted average frequency; Determine whether the weighted average frequency coincides with the threshold. If so, directly output the weighted average frequency. If not, use Newton interpolation to calculate the frequency result.

2. The method for rapid frequency detection of a power system based on multiple level sets according to claim 1, characterized in that: The set threshold is based on historical data of the power system.

3. The method for rapid frequency detection of a power system based on multiple level sets according to claim 1, characterized in that: The electrical quantity signals of the power system are collected by using time domain sampling.

4. The method for rapid frequency detection of a power system based on multiple level sets according to claim 1, characterized in that: The filtering uses a low-pass filter to perform filtering.

5. The method for rapid frequency detection of a power system based on multiple level sets according to claim 1, characterized in that: The expression of the Newton interpolation method is: f(x)=Pn(x)+Rn(x), wherein Pn(x) is the Newton interpolation polynomial and Rn(x) is the Newton interpolation remainder.

6. The method for rapid frequency detection of a power system based on multiple level sets according to claim 5, characterized in that: The expressions of Pn(x) and Rn(x) are as follows: P n (x)=f[x0]+f[x0, x1](x-x0)+f[x0, x1, x2](x-x0)(x-x1)+…+f[x0, x1…x n ](x-x0)…(xx n-1 ) 7. The method for rapid frequency detection of a power system based on multiple level sets according to claim 1, characterized in that: The weighted average calculation formula is as follows: Y = ∑ (Wi*Xi) / ∑Wi, Among them, Y represents the result to be calculated; Wi represents the weight of each Xi; Xi represents the values ​​of various parameters, and represents the frequencies of different time periods and regions in the power grid; ∑ represents the summation symbol, which means adding up all items.

8. A computer-readable storage medium, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: When the computer program is executed by the processor, the computer program executes the instructions of the method according to any one of claims 1 to 7.

9. A fast frequency detection device for power system based on multi-level sets, characterized in that: The detection device comprises: a signal acquisition module and a signal processing module, wherein: The signal acquisition module is used to collect voltage signals of the power system; The signal processing module is used to set multiple thresholds; filter the sampled voltage signal; calculate the frequency value corresponding to each voltage signal for the filtered voltage signal; combine and weight the frequency values ​​to obtain the weighted average frequency; determine whether the weighted average frequency coincides with the threshold, if so, directly output the weighted average frequency, if not, use Newton interpolation method to calculate the frequency result.

10. The power system fast frequency detection device based on multi-level sets according to claim 9, characterized in that: The signal processing module includes a filtering unit and a deduplication unit, wherein the filtering unit is used to filter the power systems that do not meet the requirements, and the deduplication unit is used to remove duplicate power systems.

Images