Frequency band division electric energy quality detection method based on TQWT and interpolation FFT

Abstract

The invention discloses a sub-band power quality detection method based on TQWT and interpolation FFT, and the method comprises the following steps: obtaining a voltage signal from a power grid detection node, and discretizing the voltage signal to obtain a discrete sampling sequence; performing adjustable Q factor wavelet transform TQWT on the discrete sampling signal to obtain a plurality of sub-bands representing different frequency bands; selecting a low-pass sub-band signal containing a fundamental frequency part, performing half-cycle effective value calculation on the low-pass sub-band signal to obtain a fundamental frequency amplitude curve, and obtaining time domain disturbance characteristics; selecting a sub-band signal containing a high-frequency part, recording the starting time and the ending time of transient oscillation, performing fast Fourier transform on a sub-band containing the high-frequency part, and taking the maximum value of the frequency component of the high-frequency part as the oscillation frequency of the transient oscillation; and performing an interpolation FFT algorithm on the sub-band signal containing the intermediate frequency part, and detecting the harmonic and inter-harmonic disturbance of the intermediate frequency band by recording the frequency and amplitude of each harmonic / inter-harmonic component in the frequency spectrum of the sub-band containing the intermediate frequency part.

Description

Technical Field

[0001] The present invention belongs to the field of power quality composite disturbance detection, and particularly considers a frequency-band power quality detection method based on TQWT and interpolation FFT. Background Art

[0002] With the widespread integration of clean energy sources such as photovoltaics and wind power, as well as diverse devices like electric vehicles and home energy storage, power quality disturbances in power systems are becoming increasingly complex and variable. The intermittent, random, and volatile nature of renewable energy generation has led to more significant voltage fluctuations and deviations in power grids. The unordered integration of electric vehicles disrupts the power flow distribution in low-voltage systems, leading to high levels of system voltage uncertainty and even multiple issues such as overvoltage, undervoltage, and voltage sag. Furthermore, the widespread use of devices such as grid-connected converters and power electronic switches has also led to severe harmonic pollution. These disturbances not only affect power supply quality but can also easily cause digital device failures, protective device malfunctions, and damage to sensitive equipment, posing a significant threat to system stability. Therefore, in-depth research on the identification and classification of power quality disturbances is crucial for effectively addressing power quality issues and improving system stability and reliability. This is directly related to the efficient operation and sustainable development of new power systems.

[0003] Joint time-frequency analysis is a classic method for power quality disturbance detection. Researchers at home and abroad have achieved numerous high-quality results in this field, including the Hilbert-Huang transform (HHT), S transform, and wavelet transform. Among these methods, the HHT can decompose the disturbance signal into multiple modal components and analyze them individually, but it is prone to modal aliasing and its screening stop criteria are difficult to determine. The Gaussian window characteristics of the S transform and its large computational complexity make it difficult to directly apply in practice. The wavelet transform is a multi-scale analysis method with variable time-frequency resolution, capable of amplifying signal details at high resolution. However, it also suffers from the difficulty of adjusting the wavelet basis and the number of decomposition levels. Joint time-frequency analysis has inherent drawbacks due to the Heisenberg effect, namely, it cannot simultaneously achieve high time-domain resolution and high frequency-domain resolution. This has led to previous innovations in time-frequency joint detection algorithms being limited to adjusting parameters to allow the algorithm to sacrifice resolution in one dimension as needed.

[0004] As the degree of power electronics in power systems gradually deepens, more and more disturbances no longer appear alone, but frequently occur as composite disturbances with coupling characteristics in the time and frequency domains. In order to accurately detect composite disturbance signals of power quality, a power quality detection algorithm with high accuracy in both the time and frequency domains is needed. To this end, a frequency-band power quality detection method based on TQWT and interpolation FFT is proposed. Summary of the Invention

[0005] To solve the problem that current power quality detection algorithms are difficult to simultaneously have high time-domain and frequency-domain resolutions, the present invention proposes a sub-band power quality detection method based on TQWT and interpolated FFT, which can be applied to power electronic systems with a high proportion of renewable energy for power quality composite disturbance detection. The technical solution of the present invention is as follows:

[0006] A sub-band power quality detection method based on TQWT and interpolated FFT includes the following steps:

[0007] Step 1: Obtain a voltage signal from a power grid detection node and discretize it to obtain a discrete sampling sequence;

[0008] Step 2: Select appropriate parameters to perform an adjustable Q-factor wavelet transform TQWT on the discrete sampling signal to obtain multiple sub-bands representing different frequency bands;

[0009] Step 3: Select a low-pass sub-band signal containing the fundamental frequency part, calculate its half-cycle effective value to obtain a fundamental frequency amplitude curve, and record the maximum value, minimum value, and mutation time point of the fundamental frequency amplitude to obtain time-domain disturbance characteristics;

[0010] Step 4: Select a sub-band signal containing the high-frequency part, record the start time and end time of the transient oscillation, and perform a fast Fourier transform on the sub-band containing the high-frequency part. Use the maximum value of the frequency components of the high-frequency part as the oscillation frequency of the transient oscillation;

[0011] Step 5: Perform an interpolated FFT algorithm on the sub-band signal containing the intermediate frequency part, and detect harmonic and inter-harmonic disturbances in the intermediate frequency band by recording the frequencies and amplitudes of each harmonic / inter-harmonic component in the spectrum of the sub-band containing the intermediate frequency part.

[0012] Further, in step 2, select appropriate TQWT parameters according to the following formula:

[0013]

[0014] In the formula, Q, r, and J are TQWT parameters, and f s is the sampling frequency, f0 is the fundamental frequency. Select appropriate TQWT parameters according to the fundamental frequency of the voltage signal and the sampling frequency of the sampling device in the detection environment.

[0015] Further, let the discrete sampling sequence obtained in step 1 be x s (nT s ), where T s is the sampling interval time of the power grid sampling device, and n is the sampling point number. In step 2, the adjustable Q-factor wavelet transform TQWT is calculated to the Jth layer, and the calculation process includes:

[0016] (1) For the discrete sampling signal xs (nT s ) is subjected to fast Fourier transform to obtain the signal spectrum X(k) as follows:

[0017] where N is the total number of sampling points;

[0018] Design a TQWT filter bank. The expressions of the low-pass filter G0(k) and the high-pass filter G1(k) it contains in the frequency domain are as follows:

[0019]

[0020] In the above formula, the parameters α and β satisfy and

[0021]

[0022] N0 = 2 × round(α J *N / 2), N1 = 2 × round(β × α J-1 ×N / 2)

[0023] round(*) is the floor formula, J is the decomposition level, and N is the length of the signal spectrum X(k).

[0024] (2) In the first-layer decomposition, use the filter bank to filter the signal spectrum X(k) to obtain the low-pass filtering coefficients U0 (1) (k) = X(k)G0(k) and the high-pass filtering coefficients U1 (1) (k) = X(k)G1(k). Perform a low-pass scaling transformation on the low-pass filtering coefficients of the first-layer decomposition, that is, compress the spectrum of U0 (1) (k) from the original bandwidth απ to the low-frequency range (0, απ). The transformation formula is as follows

[0025]

[0026] Perform a high-pass scaling transformation on the high-pass filtering coefficients of the first-layer decomposition, that is, translate the spectrum of U1 (1) (k) from the original transition band ((1-β)π, απ) to the high-frequency range (απ, π). The transformation formula is as follows

[0027]

[0028] Perform an inverse Fourier transform on V0 (1) (k) and V1 (1) (k), and then the time-domain representation v0 (1) (n) of the low-pass subband and the time-domain representation v1 (1)(n); In the second - layer decomposition, the signal to be decomposed is the low - pass sub - band signal V0 obtained from the first - layer decomposition (1) (k); Sequentially perform the second - layer decomposition, third - layer decomposition, and so on until the J - th layer of TQWT iterative decomposition in ascending order

[0029] (3) In the J - th layer, the signal to be decomposed is the low - pass sub - band signal V0 obtained from the (J - 1) - th layer decomposition (J-1) (k), and after decomposition, the low - pass filtering coefficient U0 (J) (k) and high - pass filtering coefficient U1 (J) (K) can be obtained

[0030] U0 (J) (k)=V0 [[ID=1७]] (J-1) (k)G0(k)

[0031] U1 (J) (k)=V0 (J-1) (k)G1(k)

[0032] Perform low - pass scaling transformation and high - pass scaling transformation on both of them respectively

[0033]

[0034] For V0 (J) (k) and V1 (J) (k), perform inverse Fourier transform to obtain the time - domain representation v0 (J) (n) of the low - pass sub - band and the time - domain representation v1 (J) (n) of the high - pass sub - band after the J - th layer decomposition

[0035] (4) After J - layer decomposition, a total of 1 low - pass sub - band and (J - 1) high - pass sub - bands can be obtained.

[0036] Furthermore, the time - domain disturbance characteristics include the disturbance amplitude, start time, and end time of voltage sag, voltage swell, and voltage interruption.

[0037] Furthermore, in step 3, select the last - layer low - pass sub - band after the J - th layer of TQWT decomposition for analysis. Let the last - layer low - pass sub - band be V0 (J) (k), calculate the half - cycle effective value to obtain the fundamental - frequency effective - value change curve. When the fundamental - frequency effective value rises by 10% to 80%, it is detected as a voltage swell; when the fundamental - frequency effective value drops by 10% to 90%, it is defined as a detected voltage sag; when the fundamental - frequency effective value drops by 90% to 100%, it is defined as a detected voltage interruption.

[0038] Furthermore, in step 4, the high - frequency part is 700 hz - 1600 hz. Select the sub - band containing the high - frequency part, denoted as V0 (j)(k), where j is the sub-band number containing high-frequency components. If a transient oscillation occurs, record the start time and end time of the transient oscillation. At the same time, perform a Fourier transform on V0 (j) (k), and record the maximum value of the gradually decaying frequency components that appear within the high-frequency components as the oscillation frequency of the transient oscillation. Among them, the gradually decaying frequency components appear as arch-shaped protrusions in the frequency spectrum.

[0039] Further, the intermediate frequency part is 150 hz to 700 hz. In step 5, select the sub-band containing the intermediate frequency part, perform a fast Fourier transform on the selected sub-band to obtain the frequency spectrum X(k), detect the peak points in X(k) whose amplitudes exceed the preset threshold, and record the frequency point position as k peak , for each detected peak point, extract its main spectral line X0, the left adjacent spectral line X -1 , and the right adjacent spectral line X +1 , calculate the frequency correction amount δ, and the formula is as follows

[0040]

[0041] Calculate the actual frequency according to the frequency correction amount formula:

[0042]

[0043] Calculate the actual amplitude according to the amplitude correction amount formula:

[0044]

[0045] After correction, the frequency and amplitude information of the harmonics and inter-harmonics in this frequency band can be obtained.

[0046] Further, in step 5, within the range of 150 hz to 700 hz, the frequency components that are integer multiples of the fundamental frequency of 50 hz are regarded as harmonic disturbances, and the frequency components that are non-integer multiples of the fundamental frequency of 50 hz are regarded as inter-harmonic disturbances.

[0047] Further, the composite disturbance is the superposition of voltage sag, voltage swell, voltage interruption, harmonics, inter-harmonics and transient oscillation disturbances, and is detected by judging its characteristics according to the judgment methods in steps 3 to 5. The composite disturbance includes: harmonics and sag, harmonics and interruption, harmonics and swell, harmonics and oscillation, inter-harmonics and interruption, inter-harmonics and swell, inter-harmonics and sag, inter-harmonics and oscillation, oscillation and sag, oscillation and interruption, oscillation and swell, harmonics and inter-harmonics, sag and harmonics and oscillation, sag and inter-harmonics and oscillation, sag and harmonics and inter-harmonics. Description of the Drawings

[0048] Figure 1 It is a waveform diagram of 80% voltage sag

[0049] Figure 2 It is the waveform diagram of 30% voltage sag

[0050] Figure 3 It is the waveform diagram of 99% voltage interruption

[0051] Figure 4 It is the waveform diagram of 250hz harmonic disturbance

[0052] Figure 5 It is the waveform diagram of 130hz interharmonic disturbance

[0053] Figure 6 It is the waveform diagram of 1000hz transient oscillation

[0054] Figure 7 It is the waveform diagram of 80% voltage sag + 130hz interharmonic + 1000hz transient oscillation

[0055] Figure 8 It is the structure of the J-layer TQWT filter bank

[0056] Figure 9 It is the detection result of 80% voltage sag from 0.05s to 0.15s

[0057] Figure 10 It is the frequency response of each high-pass filter and the last low-pass filter under 4-layer decomposition

[0058] Figure 11 It is the detection result of transient oscillation at 900hz from 0.28s to 0.32s

[0059] Figure 12 It is the detection result of harmonic and interharmonic disturbances

[0060] Figure 13 It is the detection result of composite disturbances Specific implementation manner

[0061] The following will combine the accompanying drawings and embodiments to elaborate in detail on a method for detecting power quality in sub-bands based on TQWT and interpolated FFT proposed by the present invention.

[0062] (1) Establish 21 different mathematical models of power quality disturbance signals.

[0063] Suppose the signal detected and input by the sampling device in the power system is x(t), then the normal voltage signal model can be established as:

[0064] x(t) = Asin(ω0t)

[0065] Where A is the standard voltage amplitude, t is the time value, the angular frequency ω0 = 2πf0, and the alternating current frequency in China is 50 hz. Therefore, in subsequent modeling, f0 = 50 hz is uniformly adopted. The single disturbances in the power system are divided into voltage sags, voltage swells, voltage interruptions, voltage flickers, harmonics, and transient oscillations. The modeling of the above six single disturbances is as follows:

[0066] Voltage sag:

[0067] x(t) = [1 - A[ε(t - t1) - ε(t - t2)]]sin(ω0t)

[0068] The voltage sag is defined as the rapid drop of the effective value of the supply voltage to 90% - 10% of the rated value. In the formula, ε(t) is the step function, and its defined expression is

[0069]

[0070] t1 is the starting time of the voltage sag, t2 is the ending time of the voltage sag, A is the depth of the voltage sag, and the parameter value range is 0.1 < A < 0.9, T < t2 - t1 < 9T. Here, T is the length of one cycle, which is 0.02 s.

[0071] Voltage swell:

[0072] x(t) = [1 + A[ε(t - t1) - ε(t - t2)]]sin(ω0t)

[0073] The voltage swell is defined as the rapid rise of the effective value of the voltage to 110% - 180% of the rated value. In the formula, t1 is the starting time of the voltage swell, t2 is the ending time of the voltage swell, A is the height of the voltage swell, and the parameter value range is 0.1 < A < 0.8, T < t2 - t1 < 9T. Here, T is the length of one cycle, which is 0.02 s.

[0074] Voltage interruption:

[0075] x(t) = [1 - A[ε(t - t1) - ε(t - t2)]]sin(ω0t)

[0076] The voltage interruption is defined as the rapid drop of the effective value of the voltage to 90% - 100% of the rated value. In the formula, t1 is the starting time of the voltage interruption, t2 is the ending time of the voltage interruption, A is the depth of the voltage interruption, and the parameter value range is 0.9 < A < 1, T < t2 - t1 < 9T. Here, T is the length of one cycle, which is 0.02 s.

[0077] Harmonics:

[0078]

[0079] Harmonics are non-linear components in the power system with frequencies higher than the fundamental frequency. Mathematically, they are represented as the sum of multiple sine waves. In the formula, A i is the amplitude of each harmonic, i is the harmonic order. Harmonics in the power system are usually considered as integer multiples of the fundamental frequency. Therefore, i = 1, 2, 3, …, n.

[0080] Interharmonics:

[0081]

[0082] Interharmonics refer to harmonics that are not integer multiples of the fundamental frequency. In the formula, B i is the amplitude of each interharmonic, f i is the frequency of each interharmonic, i = 1, 2, …, n, which are non-integer multiples of 50 Hz.

[0083] Transient oscillation:

[0084]

[0085] Among them, A is the oscillation amplitude, taking values from 0.1 to 0.8; t1 and t2 are the start and end times of the oscillation, 0.5T < t2 - t1 < 3T; τ is the oscillation decay factor, taking values from 25 to 330, and f is the oscillation frequency, generally taking values from 700 to 1600 Hz.

[0086] Six single perturbation waveforms established using Matlab according to the above mathematical model are as Figures 1 to 6 shown. From this, composite perturbations can be constructed. Generally speaking, voltage sags, voltage swells, and voltage interruptions can be defined as time-domain perturbations, while harmonics, interharmonics, and transient oscillations are defined as frequency-domain perturbations. The composite perturbation of the power system is composed of the combination of time-domain perturbations and frequency-domain perturbations. Twenty-one different power quality perturbation signals are shown in Table 1. Now, a typical example of the composite perturbation is given, and others are not elaborated. The example is voltage sag + interharmonic + transient oscillation perturbation, and its waveform diagram is Figure 7 .

[0087] Table 1

[0088] C1 Sag C8 Harmonics + Interruption C15 Oscillation + Sag C2 Interruption C9 Harmonics + Swell C16 Oscillation + Interruption C3 Swell C10 Harmonics + Oscillation C17 Oscillation + Swell C4 Harmonics C11 Interharmonics + Interruption C18 Harmonics + Interharmonics C5 Interharmonics C12 Interharmonics + Swell C19 Sag + Harmonics + Oscillation C6 Transient Oscillation C13 Interharmonics + Sag C20 Sag + Interharmonics + Oscillation C7 Harmonics + Sag C14 Interharmonics + Oscillation C21 Sag + Harmonics + Interharmonics

[0089] (2) In the power system, time-domain perturbations (sags, swells, interruptions) are manifested as the replication change of the sine signal with a fundamental frequency of 50 Hz in a certain time period, while frequency-domain perturbations (harmonics, interharmonics, transient oscillations) are manifested as the appearance of other components in the spectrum except for the 50 Hz component. If the 50 Hz frequency component can be accurately extracted, the time-domain perturbations can be detected more accurately. TQWT uses a two-channel filter bank to achieve signal decomposition and reconstruction. Its Q factor is easy to adjust, which can provide flexibility for filter design under specific spectrum decomposition conditions. Figure 8The diagram of the J-layer TQWT filter bank is shown in Figure 1. After the disturbance signal is input, the first decomposition will be divided into a low-pass subband and a high-pass subband. The next decomposition will repeat the above operation on the low-pass subband, thereby decomposing the signal layer by layer according to the frequency band to obtain more detailed information. The filter design depends entirely on the low-pass scale parameter α and the high-pass scale parameter β. These two parameters will be calculated by the user-set oscillation factor Q, redundancy r and decomposition layer number J. The calculation formula is as follows:

[0090]

[0091] The quality factor Q, defined as the ratio of the center frequency to the bandwidth, mainly affects the oscillation properties of the wavelet and should be selected according to the oscillation characteristics of the signal. The redundancy factor r is defined as the ratio of the sum of the wavelet coefficients to the signal length, which characterizes the local refinement ability of the reconstruction. The number of decomposition layers J is the number of cascaded filter banks and determines the number of decomposition subbands. It should be noted that there is a maximum value for the number of decomposition layers J, and the user should not exceed the maximum number of decomposition layers J when setting the parameters. max ,With the increase of the number of decomposition layers, the ,number of sampling points in each sub-band layer will decrease. ,The appropriate number of decomposition layers should be selected to ensure ,sufficient sampling points to prevent sub-band signal distortion.

[0092]

[0093] Experimental verification shows that harmonics and interharmonics within 0-150 Hz have little effect on time domain disturbance detection. It is more in line with the actual situation to exchange complete decomposition of harmonics for better time domain disturbance detection accuracy. This paper proposes a parameter selection standard that will optimize the parameters with the goal of completely extracting the fundamental frequency component. Assume that the last layer of low-pass filter in TQWT is Its center frequency It can be obtained by the following formula

[0094]

[0095] where f s is the sampling frequency, then Setting the fundamental frequency to 50 Hz allows the fundamental frequency component to be decomposed into the final low-pass subband signal. The constraint formulas for the parameters (r, Q, J) are as follows. Users can select appropriate parameters based on the sampling frequency of the sampling device.

[0096]

[0097] After the decomposition is completed, extract the sub-band signal of the last layer and perform a half-cycle effective value detection on it to obtain the fundamental frequency effective value curve. According to the curve characteristic values, it is possible to judge what kind of time-domain disturbance has occurred. Take an example for verification. At a sampling frequency of 12,800 hz, select the parameters r = 1.4, Q = 1.46, J = 4 that meet the requirements. Under this condition, detect the 80% voltage sag signal occurring from 0.05 s to 0.15 s, as Figure 9 shown. The detection result is that an 80.2% sag disturbance occurs from 0.051 s to 0.152 s, and the detection result is good.

[0098] (3) Transient oscillation is a frequency component whose oscillation frequency band is 700 hz to 1600 hz but the oscillation gradually decays. In this method, the sub-band that can contain the frequency band of 700 hz to 1600 hz should be extracted from the sub-band signals decomposed by TQWT. Still taking the TQWT parameters r = 1.4, Q = 1.46, J = 4 as an example, the frequency responses of the high-pass filters of each layer and the last low-pass filter under this parameter are as Figure 10 shown. It can be seen that this frequency band should be decomposed into the second-layer high-pass sub-band. Take the transient oscillation with a frequency of 900 hz and a duration from 0.28 s to 0.32 s as an example to show the results, as Figure 11 shown. The detection result of the duration is from 0.278 s to 0.322 s; the detection result of the oscillation frequency is 899.297 hz, and the detection result is good.

[0099] (4) 150 hz to 700 hz is a frequency band where harmonics and inter-harmonics occur frequently. Since the harmonic frequency is an integer multiple of the fundamental frequency, the general Fourier algorithm can achieve a frequency resolution of 5 hz and can detect harmonics well. However, the inter-harmonic frequency is a non-integer multiple of the fundamental frequency. This special frequency makes it difficult to achieve synchronous sampling of inter-harmonics during sampling. Therefore, the true frequency points of inter-harmonics often do not fall on the discrete frequency points of the spectrum, and it is necessary to correct and calculate the frequencies and amplitudes of the discrete frequency points on both sides of the true frequency. Now, take an example to generate a 250 hz harmonic disturbance with a normalized amplitude of 0.5 and a 342 hz inter-harmonic disturbance with a normalized amplitude of 0.3. The detection results are as Figure 12 shown. It can be seen that the detected disturbances are a harmonic with a frequency of 250.00 hz and an amplitude of 0.499; an inter-harmonic with a frequency of 342.11 hz and an amplitude of 0.291, and the detection error is small.

[0100] (5) To show the detection results of the composite disturbance of this method, this paper shows sag + inter-harmonic + oscillation. The set values of each disturbance are: duration from 0.04 s to 0.05 s, 80% voltage sag; duration from 0.38 s to 0.42 s, 1000 hz transient oscillation; 156 hz, inter-harmonic with a normalized amplitude of 0.3; 248 hz, inter-harmonic with a normalized amplitude of 0.2.

[0101] The detection results are as follows Figure 13 , the method proposed in this paper detects an 80.4% voltage sag at 0.042 s to 0.050 s; a transient oscillation of 999.22 hz at 0.378 s to 0.421 s; interharmonics with a frequency of 155.90 hz and an amplitude of 0.261; and interharmonics with a frequency of 247.89 hz and an amplitude of 0.194. The method in this paper has a small error and can better detect complex disturbances.

[0102] In summary, the present invention proposes a method for detecting power quality in sub - frequency bands based on TQWT and interpolated FFT, which can be applied to modern high - proportion renewable - energy power - electronic systems for detecting complex power - quality disturbances.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific implementation manners of the present invention can still be modified or equivalently replaced, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A method for detecting power quality in sub - bands based on TQWT and interpolated FFT, comprising the following steps: Step 1: Obtain a voltage signal from a power grid detection node and discretize it to obtain a discrete sampling sequence; Step 2: Select appropriate parameters to perform an adjustable Q - factor wavelet transform TQWT on the discrete sampling signal to obtain multiple sub - bands representing different frequency bands; Step 3: Select the low - pass sub - band signal containing the fundamental - frequency part, calculate the half - cycle effective value of it to obtain the fundamental - frequency amplitude curve, and record the maximum value, minimum value, and mutation time point of the fundamental - frequency amplitude to obtain time - domain disturbance characteristics; Step 4: Select the sub - band signal containing the high - frequency part, record the start time and end time of the transient oscillation, and perform a fast Fourier transform on the sub - band containing the high - frequency part. Use the maximum value of the frequency components in the high - frequency part as the oscillation frequency of the transient oscillation; Step 5: Perform an interpolated FFT algorithm on the sub - band signal containing the intermediate - frequency part, and detect harmonic and inter - harmonic disturbances in the intermediate - frequency band by recording the frequencies and amplitudes of each harmonic / inter - harmonic component in the spectrum of the sub - band containing the intermediate - frequency part.

2. According to the power quality fast disturbance detection method described in claim 1, in step 2, select appropriate TQWT parameters according to the following formula: where Q, r, J are TQWT parameters, and f s is the sampling frequency, f0 is the fundamental frequency. Appropriate TQWT parameter selection is carried out according to the fundamental frequency of the voltage signal and the sampling frequency of the sampling device in the detection environment.

3. The power quality rapid disturbance detection method according to claim 1, characterized in that Let the discrete sampling sequence obtained in step 1 be x s (nT s ), where T s is the sampling interval time of the power grid sampling device, n is the sampling point number, and the adjustable Q-factor wavelet transform TQWT in step 2 is calculated to the J-th layer. The calculation process includes: (1) Perform a fast Fourier transform on the discrete sampled signal x s (nT s ) to obtain the signal spectrum X(k) as follows: Where N is the total number of sampling points; Design a TQWT filter bank. The expressions of its low - pass filter G0(k) and high - pass filter G1(k) in the frequency domain are as follows: In the above formula, the parameters α and β satisfy and N0 = 2 × round(α J *N / 2), N1 = 2 × round(β × α J-1 ×N / 2) round(*) is the floor function formula, J is the decomposition level, and N is the length of the signal spectrum X(k). (2) In the first - layer decomposition, the signal spectrum X(k) is filtered using a filter bank to obtain the low - pass filtering coefficients U0 (1) (k) = X(k)G0(k) and the high - pass filtering coefficients U1 (1) (k) = X(k)G1(k). A low - pass scaling transformation is performed on the low - pass filtering coefficients of the first - layer decomposition, that is, the spectrum of U0 (1) (k) is compressed from the original bandwidth απ to the low - frequency range (0, απ), and the transformation formula is as follows Perform a high-pass scaling transformation on the high-pass filter coefficients of the first-layer decomposition, that is, shift the spectrum of U1 (1) (k) from the original transition band ((1-β)π, απ) to the high-frequency range (απ, π). The transformation formula is as follows For C0 (1) (k) and V1 (1) (k) performs an inverse Fourier transform to obtain the time-domain representation v0 of the low-pass subband after the first-layer decomposition (1) (n) and the time-domain representation v1 of the high-pass subband (1) (n); In the second-layer decomposition, the signal to be decomposed is the low-pass subband signal V0 obtained from the first-layer decomposition (1) (k); In order from low to high, perform the second-layer decomposition, third-layer decomposition, and so on until the J-layer TQWT iterative decomposition (3) In the J-th layer, the decomposed signal is the low-pass sub-band signal V0 (J-1) (k) obtained by decomposing the (J-1)-th layer. After decomposition, the low-pass filtering coefficients U0 (J) (k) and the high-pass filtering coefficients U1 (J) (k) of the J-th layer decomposition can be obtained U0 (J) (k) = V0 (J-1) (k)G0(k) U1 (J) (k) = V0 (J-1) (k)G1(k) Perform low - pass scaling transformation and high - pass scaling transformation on the two respectively For V0 (J) (k) and V1 (J) Performing the inverse Fourier transform on (k) and V1(k) can obtain the time-domain representation v0(n) of the low-pass sub-band after the Jth layer of decomposition (J) and the time-domain representation v1(n) of the high-pass sub-band (J) (n); (4) After J - layer decomposition, a total of 1 low - pass sub - band and J - 1 high - pass sub - bands can be obtained.

4. According to the power quality fast disturbance detection method described in claim 1, in step 3, the time - domain disturbance characteristics include the disturbance amplitude, start time, and end time of voltage sag, voltage swell, and voltage interruption.

5. The power quality rapid disturbance detection method according to claim 1, characterized in that In step 3, the last low-pass sub-band after the J-layer TQWT decomposition is selected for analysis. Let the last low-pass sub-band be V0 (J) (k). The root-mean-square (RMS) value of half a cycle is calculated for it to obtain the curve of the fundamental frequency RMS value change. When the fundamental frequency RMS value rises by 10% to 80%, a voltage sag is detected; when the fundamental frequency RMS value drops by 10% to 90%, it is defined as detecting a voltage dip; when the fundamental frequency RMS value drops by 90% to 100%, it is defined as detecting a voltage interruption.

6. The power quality rapid disturbance detection method according to claim 1, characterized in that In step 4, the high-frequency part is 700 hz to 1600 hz, and the sub-band containing the high-frequency part is selected and denoted as V0 (j) (k), where j is the sub-band number containing the high-frequency component. If a transient oscillation occurs, record the start time and end time of the transient oscillation. At the same time, perform a Fourier transform on V0 (j) (k), and record the maximum value of the gradually decaying frequency component that appears within the high-frequency component as the oscillation frequency of the transient oscillation. Among them, the gradually decaying frequency component appears as an arched bulge in the frequency spectrum.

7. The method for quickly detecting power quality disturbances according to claim 1, characterized in that The intermediate frequency part is 150 hz to 700 hz. In step 5, a sub-band containing the intermediate frequency part is selected, and the selected sub-band is subjected to a fast Fourier transform to obtain a spectrum X(k). Peak points with amplitudes exceeding a preset threshold are detected in X(k), and the frequency point position is recorded as k peak , for each detected peak point, its main spectral line X0, the left adjacent spectral line X -1 , and the right adjacent spectral line X +1 are extracted, and the frequency correction amount δ is calculated. The formula is as follows Calculate the actual frequency according to the frequency correction formula: Calculate the actual amplitude according to the amplitude correction formula: After correction, the frequency and amplitude information of harmonics and inter - harmonics in this frequency band can be obtained.

8. The power quality rapid disturbance detection method according to claim 7, wherein In step 5, in the range of 150 hz to 700 hz, regard the frequency components that are integer multiples of the fundamental frequency of 50 hz as harmonic disturbances, and regard the frequency components that are non - integer multiples of the fundamental frequency of 50 hz as inter - harmonic disturbances.

9. The power quality detection method according to claim 1, wherein The composite disturbance is the superposition of voltage sag, voltage swell, voltage interruption, harmonics, inter - harmonics, and transient oscillation disturbances. Detect it by judging its characteristics according to the judgment methods in steps 3 - 5. The composite disturbances include: harmonics and sag, harmonics and interruption, harmonics and swell, harmonics and oscillation, inter - harmonics and interruption, inter - harmonics and swell, inter - harmonics and sag, inter - harmonics and oscillation, oscillation and sag, oscillation and interruption, oscillation and swell, harmonics and inter - harmonics, sag and harmonics and oscillation, sag and inter - harmonics and oscillation, sag and harmonics and inter - harmonics.

Images