Power quality disturbance detection method based on improved adaptive S transformation

Abstract

The invention belongs to the technical field of power quality disturbance detection, and particularly relates to a power quality disturbance detection method based on improved adaptive S transformation, which comprises the following steps of: designing window width control factors of two Gaussian windows, flexibly adjusting time domain window widths and frequency spectrum characteristics of the Gaussian windows, and forming a global adaptive Gaussian window; taking upper and lower limits of time-frequency two-domain resolution as constraint conditions, taking enhanced signal energy concentration ratio as a target function of an optimization scheme, and adopting an interior point method to carry out parameter solving; performing fast Fourier transform on the signal and setting a threshold value, and defining a part higher than a threshold value point as a characteristic frequency point ki; screening calculation at a non-characteristic frequency point, only performing time-frequency transformation at a characteristic frequency point ki, and constructing a fast algorithm for improving adaptive S transformation; and obtaining a final disturbance characteristic information value. According to the method, the global adaptive Gaussian window is constructed as a kernel function for improving adaptive S transformation, the effective window length and the frequency spectrum of the window function can be automatically adjusted along with the change of the detection frequency, window parameters are prevented from being frequently switched for improving the time-frequency resolution, and various power quality disturbances are efficiently and accurately detected.

Description

Technical Field

[0001] The invention belongs to the technical field of power quality disturbance detection, and in particular relates to a power quality disturbance detection method based on improved adaptive S transformation. Background Art

[0002] The penetration rate of renewable energy in the power grid continues to rise, and a large number of power electronic equipment and nonlinear loads are connected to the power system, resulting in the strong randomness, wideband, and multi-coupling characteristics of power quality disturbances in the power system. The resulting power quality problems are complex and diverse, which has a serious impact on industrial production and residents' lives. At the same time, with the improvement of the degree of informatization, various precision electronic equipment have higher and higher requirements for power quality. Rapidly and accurately detecting various power quality disturbances and extracting disturbance data, and performing long-term monitoring or intelligent analysis on them can timely and accurately obtain the characteristics of various disturbance events, providing key information for power quality assessment and governance. However, existing detection methods often cannot meet the requirements of rapid response and high precision at the same time. Therefore, a set of power quality disturbance detection methods that can efficiently detect various disturbances is proposed, which is of great significance to improving the reliability of power supply and the development of smart grid technology.

[0003] At present, the detection methods for power quality disturbances at home and abroad are mainly based on time-frequency analysis methods of signal processing technology, such as short-time Fourier transform, wavelet transform, Hilbert transform, S transform, etc. The short-time Fourier transform window function is fixed and cannot distinguish disturbances of different frequency components. The wavelet transform detects and extracts features of different types of disturbance signals through multi-scale decomposition, but it is easily affected by noise and leads to large detection errors. The Hilbert transform is prone to modal aliasing when extracting characteristic information such as the instantaneous amplitude of the disturbance signal, and has poor adaptability. The S transform can be regarded as a combination of short-time Fourier transform and wavelet transform, which can detect various types of disturbances, but it needs to traverse all frequency points when calculating the Hadamard product of the window function and the signal, which has high computational complexity and large amount of calculation, resulting in poor real-time performance of the algorithm and is difficult to apply in actual engineering. Summary of the invention

[0004] The penetration of renewable energy has caused the power quality disturbances in the power system to have greater diversity and randomness. The existing detection methods cannot fully adapt to the characteristics of the power quality disturbances, resulting in poor detection effect, low real-time performance, and the inability to quickly and accurately reflect the specific characteristic information of the power quality disturbances. The present invention provides a power quality disturbance detection method based on an improved adaptive S transform.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] The power quality disturbance detection method based on improved adaptive S-transformation includes the following steps:

[0007] S1. Design the window width control factors of two Gaussian windows to flexibly adjust the time-domain window width and spectral characteristics of the Gaussian window, and construct a global adaptive Gaussian window ω(τ−t, f) as the kernel function for the improved adaptive S-transform to fully adapt to the transient power quality disturbance characteristics in the power system;

[0008] S2. Propose a parameter tuning scheme with the enhanced energy concentration ECM as the objective function, use the upper and lower limits of the time-frequency resolution in both domains as constraints, and use the interior point method to solve for the parameters. The results are used as the values of the window width control factors of the global adaptive Gaussian window ω(τ−t, f) in S1;

[0009] S3. Perform a fast Fourier transform on the signal and set a threshold. The part above the threshold point is defined as the characteristic frequency point k i ;

[0010] S4. Screen out the calculations at non-characteristic frequency points and only perform time-frequency transformation at the characteristic frequency point k i to construct a fast algorithm for the improved adaptive S-transform and obtain a time-frequency transformation matrix containing the characteristic information at the characteristic frequency point;

[0011] S5. Extract features based on the time-frequency transformation matrix in S4 to obtain the characteristic information value of the final disturbance.

[0012] The structure of the global adaptive Gaussian window in S1 is as follows:

[0013]

[0014] Among them, the standard deviation σ(f) is a function of the signal frequency f, and the specific structure is σ(f) = 1 / |λ 1 f + λ 2 |. The standard deviation can be flexibly controlled by two window width adjustment factors λ 1 and λ 2 , which is the basis for enabling the window function to adapt to various disturbance signals; the parameter τ is the time shift factor controlling its position in the time domain, t is the time variable in the time-domain signal, and f is the detection signal frequency.

[0015] The definition formula of the improved adaptive S-transform in S1 is:

[0016]

[0017] Among them, x(t) is any time-domain signal.

[0018] The parameter tuning scheme with the enhanced energy concentration ECM as the objective function in S2 is as follows: It includes the following steps:

[0019] S21. After performing the improved adaptive S-transform on the signal, the energy concentration ECM is defined as:

[0020]

[0021] Among them, the normalized IAST matrix is as follows:

[0022]

[0023] The goal of the parameter tuning scheme is to maximize the energy concentration measure (ECM) of the detected signal IAST (λ 1 , λ 2 );

[0024] S22. Using the trade-off between time-frequency resolution as a constraint;

[0025] S23. Obtaining the parameter values that can enhance energy concentration and take into account the trade-off of time-frequency resolution.

[0026] The method of using the trade-off between time-frequency resolution as a constraint in S22 is as follows:

[0027] The overall constraint condition is:

[0028] σ low ≤ σ(f) ≤ σ up

[0029] σ low is the minimum time interval that can be resolved. For a discrete signal sampling sequence, σ low can be replaced by nT s , where T s is the signal sampling period, and n is an integer greater than or equal to 1. Transforming σ low ≤ σ(f) into constraint condition 1:

[0030] nT s *(λ 1 f min + λ 2 ) - 1 ≤ 0

[0031] where f min is the minimum detection frequency, and T s is the signal sampling period;

[0032] σ up determines the upper limit of the frequency resolution. Similarly, transforming σ(f) ≤ σ up into constraint condition 2:

[0033] 1 / (λ 1 f max + λ 2 ) - uT s ≤ 0

[0034] where fmax Determined by the sampling frequency based on the sampling theorem, T s is the sampling period of the signal.

[0035] The method for the parameter value of S23 to enhance energy concentration and balance the time-frequency resolution trade-off is as follows:

[0036] To improve the speed of parameter optimization, it is also necessary to limit the reasonable range of the parameters themselves as the third constraint condition. The overall parameter optimization scheme is as follows:

[0037]

[0038] s.t. nT s *t(λ 1 f min +λ 2 ) ― 1 ≤ 0

[0039] 1 / (λ 1 f max +λ 2 ) ― uT s ≤ 0

[0040] λ 1 ∈(―0.1, ―0.0015)

[0041] λ 2 ∈(15, 35)

[0042] The interior point method is used to solve the parameters of the above tuning scheme to obtain the parameter values that can enhance energy concentration and balance the time-frequency resolution trade-off.

[0043] The method for determining the characteristic frequency points of the disturbance characteristic signal in S3 is as follows: It includes the following steps:

[0044] S31. Perform a fast Fourier transform on the signal to obtain the spectrogram of the disturbance signal, preliminarily determine the frequency distribution, and obtain the part containing the main energy in the actual disturbance signal;

[0045] S32. Set a threshold according to the characteristics of the actual disturbance signal, define the frequency components with amplitudes higher than this threshold in the spectrogram as the main frequency points, and automatically filter out the remaining non-main frequency points. The number v of main frequency points determined by the automatic threshold method is much smaller than N, thereby reducing the computational time complexity to O(Nlog 2 N), reducing the amount of computation.

[0046] The calculation method for filtering out non-characteristic frequency points in S4 is as follows: It includes the following steps:

[0047] S41. First, discretize it, and discretize the time variable τ and f in the improved self-adaptive S transform into mT sand k / NT s ;

[0048] S42. During the calculation process of the two-dimensional time-frequency matrix of the disturbance signal, for the rows where the non-dominant frequency points are located, all are calculated according to zeroing, and only the dominant frequency point k i is calculated for the part where it is located. The discrete expression is:

[0049]

[0050] where k i is the dominant frequency point screened out by the signal through the automatic threshold method. The values of m and l are 0, 1, 2,... N - 1, and the value of i is 1, 2... v. X((l + k i ) / NT s ) is the discrete form of X(α + f) containing only the dominant frequency. W(l / NT s , k i / NT s ) is the discrete Fourier transform of the adaptive Gaussian window.

[0051] In the above-mentioned S5, the method for extracting the characteristic information of the final disturbance based on the final time-frequency characteristic map includes the following steps:

[0052] S51. First, extract the time-domain characteristic information of the disturbance signal. In the time-frequency matrices of the two, the time-domain change curve of a certain fixed frequency component in the disturbance signal can be extracted, and the differential curve of this curve is calculated to obtain the starting time of the disturbance of this frequency component. Further, the characteristic information of the magnitude change of the disturbance signal is extracted from the time-domain change curve;

[0053] S52. According to the final two-dimensional time-frequency matrix, extract the frequency-domain characteristic information such as how many frequency components the disturbance signal contains. Then, extract the frequency normalization value of the ordinate in the improved adaptive S-varying two-dimensional time-frequency matrix, and the frequency values of each disturbance signal component can be accurately obtained.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] The present invention constructs a global adaptive Gaussian window as the kernel function of the improved adaptive S transform, which can automatically adjust the effective window length and spectrum of the window function with the change of the detection frequency, avoiding frequent switching of window parameters to improve the time-frequency resolution. And the window parameters are selected with the enhancement of signal energy concentration as the parameter optimization goal to ensure the accurate time-frequency positioning of various disturbances. Finally, a fast algorithm is constructed to perform time-frequency transformation on the disturbance signal, with lower computational complexity, better real-time performance, and stronger time-frequency resolution ability, and can quickly and accurately detect various disturbances. Description of the Drawings

[0056] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings described below are only exemplary. For those of ordinary skill in the art, without creative efforts, other implementation drawings can be obtained by extension based on the provided drawings.

[0057] The structures, proportions, sizes, etc. shown in this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention. Therefore, they do not have technical essence. Any modification of the structure, change of the proportional relationship or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope covered by the technical content disclosed in the present invention.

[0058] Figure 1 It is a flow chart of the improved adaptive S-transform fast algorithm of the present invention;

[0059] Figure 2 It is the IAST time-frequency analysis diagram of the present invention;

[0060] Figure 3 It is the fundamental frequency amplitude curve and the frequency amplitude curve diagram of the present invention. Specific Embodiments

[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. These descriptions are only to further illustrate the features and advantages of the present invention, rather than a limitation on the claims of the present invention; based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope protected by the present application.

[0062] The following will further describe in detail the specific embodiments of the present invention in combination with the drawings and embodiments. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0063] Step S1: Design the window width control factors of two Gaussian windows, flexibly adjust the time-domain window width and spectral characteristics of the Gaussian windows, and construct the global adaptive Gaussian window ω(τ−t, f) as the kernel function of the improved adaptive S-transform to fully adapt to the transient power quality disturbance characteristics in the power system;

[0064] Step S2: Propose a parameter tuning scheme with the enhanced energy concentration ECM as the objective function, use the upper and lower limits of the time-frequency domain resolution as constraints, and adopt the interior point method to solve the parameters. The result is used as the value of the window width control factor of the global adaptive Gaussian window ω(τ―t, f) in Step 1;

[0065] Step S3: Perform a fast Fourier transform on the signal and set a threshold. The part above the threshold point is defined as the characteristic frequency point k i ;

[0066] Step S4: Screen out the calculations at non-characteristic frequency points and only perform time-frequency transformation at the characteristic frequency point k i to construct a fast algorithm for the improved adaptive S transform, and obtain a time-frequency transform matrix containing the characteristic information at the characteristic frequency point;

[0067] Step S5: Extract features based on the time-frequency transform matrix in Step 4 to obtain the characteristic information value of the final disturbance;

[0068] Preferably, the structure of the global adaptive Gaussian window is:

[0069]

[0070] where the standard deviation σ(f) is a function of the signal frequency f, and the specific structure is σ(f) = 1 /

[0071] |λ 1 f + λ 2 |. The standard deviation can be flexibly controlled by two window width adjustment factors λ 1 and λ 2 , which is the basis for enabling the window function to adapt to various disturbance signals. The parameter τ is the time shift factor controlling its position in the time domain, t is the time variable in the time domain signal, and f is the detection signal frequency.

[0072] For any time domain signal x(t), the defined formula for the constructed improved adaptive S transform is:

[0073]

[0074] Step S21: After performing the improved adaptive S transform on the signal, the energy concentration ECM is defined as:

[0075]

[0076] where the normalized IAST matrix is:

[0077]

[0078] The objective of the parameter tuning scheme is to make the energy concentration ECM of the detected signal IAST(λ 1 , λ 2 ) maximum

[0079] Step S22: Taking the trade - off between time - frequency resolution as a constraint, the overall constraint condition is:

[0080] σ low ≤σ(f)≤σ up

[0081] σ low is the minimum time interval that can be resolved. For a discrete - form signal sampling sequence, σ low can be replaced by nT s , where T s is the sampling period of the signal, and n is an integer greater than or equal to 1. Transforming σ low ≤σ(f) into constraint condition one:

[0082] nT s *(λ 1 f min +λ 2 ) ― 1 ≤ 0

[0083] In the formula, f min is the minimum frequency to be detected, and T s is the sampling period of the signal

[0084] σ up determines the upper limit of the frequency resolution. Similarly, transforming σ(f) ≤ σ up into constraint condition two:

[0085] 1 / (λ 1 f max +λ 2 ) ― uT s ≤ 0

[0086] In the formula, f max is determined by the sampling frequency based on the sampling theorem, and T s is the sampling period of the signal

[0087] Step S23: To improve the speed of parameter optimization, it is also necessary to limit the reasonable range of the parameters themselves as constraint condition three. The overall parameter optimization scheme is

[0088]

[0089] s.t. nT s *(λ 1 f min +λ 2 ) ― 1 ≤ 0

[0090] 1 / (λ 1 fmax +λ 2 ) - uT s ≤0

[0091] λ 1 ∈( - 0.1, - 0.0015)

[0092] λ 2 ∈(15, 35)

[0093] The interior point method is used to solve the parameters of the above optimization scheme, and the parameter values that can enhance energy concentration and balance the time - frequency resolution trade - off are obtained.

[0094] Preferably, the specific steps for determining the characteristic frequency points of the perturbation characteristic signal are

[0095] Step S31: Perform a fast Fourier transform on the signal to obtain the spectrogram of the perturbation signal, preliminarily determine the frequency distribution, and obtain the part containing the main energy in the actual perturbation signal.

[0096] Step S32: Set a threshold according to the characteristics of the actual perturbation signal, define the frequency components with amplitudes higher than this threshold in the spectrogram as the main frequency points, and automatically filter out the remaining non - main frequency points. The number v of main frequency points determined by the automatic threshold method is much smaller than N, so the computational time complexity is reduced to O(Nlog 2 N), reducing the amount of computation.

[0097] Preferably, for the calculation of filtering out non - characteristic frequency points, only the improved time - frequency transformation is performed at the characteristic frequency point k i The specific steps for constructing the improved adaptive S - transform fast algorithm are

[0098] Step S41: To facilitate the implementation of the IAFST algorithm in a digital signal processor, first discretize it. The time variables τ and f in the improved adaptive S - transform are discretized into mT s and k / NT s .

[0099] Step S42: During the two - dimensional time - frequency matrix calculation of the perturbation signal, for the rows where the non - main frequency points are located, all are calculated as zero, and only the part where the main frequency point k i is located is calculated. The discretized expression is:

[0100]

[0101] In the formula, k i is the main frequency point screened out by the signal through the automatic threshold method. The values of m and l are 0, 1, 2,... N - 1, the value of i is 1, 2... v, X((l + k i ) / NT s) is the discrete form where X(α + f) in Equation (9) only contains the main frequency, and W(l / NT s ,k i / NT s ) is the discrete Fourier transform of the adaptive Gaussian window. The specific implementation process of the IAFST algorithm is as Figure 1 shown.

[0102] Preferably, based on the final time-frequency feature map, the characteristic information of the final disturbance is extracted. The specific steps are

[0103] Step S51: First, extract the time-domain characteristic information of the disturbance signal. In the time-frequency matrices of the two, the time-domain change curve of a certain fixed frequency component in the disturbance signal can be extracted, and the differential curve of this curve is calculated to obtain the starting time of the disturbance of this frequency component. Further characteristic information such as the magnitude change of the disturbance signal is extracted from the time-domain change curve.

[0104] Step S52: Based on the final two-dimensional time-frequency matrix, extract the frequency-domain characteristic information such as how many frequency components the disturbance signal contains. Then, extract the frequency normalization value of the ordinate in the improved adaptive S-transform two-dimensional time-frequency matrix, and the frequency values of each disturbance signal component can be accurately obtained.

[0105] The following further illustrates the technical solution claimed in the present invention through a specific embodiment. This embodiment adopts the above-mentioned real-time detection method for power quality disturbances based on the improved adaptive S-transform.

[0106] In this embodiment, the power quality disturbance signal to be detected is a composite disturbance signal containing multiple frequency components. The specific mathematical model is as shown in the following formula:

[0107]

[0108] Among them, f0 is the power frequency of 50 Hz, ε(t) is the step function, and this signal contains a voltage sag with an amplitude of 0.6 pu within 0.08 - 0.16 s, accompanied by 5th and 14th time-varying harmonics between 0.08 s - 0.18 s and 0.02 s - 0.12 s, and a transient oscillation with a frequency of 29 times the power frequency within 0.11 s - 0.14 s.

[0109] To verify the effectiveness of this method, first, perform an improved adaptive S-transform analysis on this disturbance signal to obtain a two-dimensional time-frequency matrix, as Figure 2 shown.

[0110] As Figure 2 can be seen, the IAST with the enhanced energy concentration optimization scheme as the parameter selection basis has the optimal energy concentration performance, maintains good time-frequency resolution on the entire time-frequency plane, and thus is conducive to improving the accuracy of disturbance feature extraction. The extracted fundamental frequency amplitude curve and frequency amplitude curve are asFigure 3 as shown

[0111] from Figure 3 It can be seen that the improved adaptive S-transform can accurately extract the amplitude transformation trend of the disturbance at the fundamental frequency and extract the accurate values of each frequency component.

[0112] To further verify the detection effect of this method in practical applications, a power quality disturbance detection platform was built. Among them, the disturbance signal source uses the semi-physical simulation platform dSPACE based on Matlab / simulink. The power quality disturbance signal is collected in real time by the power quality signal acquisition chip ADE9000. The DSP uses the TMS320F28377D with a dual-core architecture. The DSP and ADE9000 realize real-time data exchange through SPI communication, obtain the power signal of the acquisition chip and are responsible for the execution of the detection algorithm.

[0113] Five types of composite disturbance signals including time-frequency two domains were set, and the accuracy and rapidity of the algorithm detection were tested respectively. Among them, the accuracy is measured by the relative error between the detected value and the true value of various disturbance characteristic information. The main characteristic information of various disturbances includes disturbance amplitude, duration, and frequency components. The relative error value during the composite disturbance test is the average relative error of its components. The detected accuracy results are shown in Table 1.

[0114] Table 1 Accuracy of IAFST detection for different types of disturbances

[0115]

[0116] As can be seen from Table 1, whether it is a single disturbance or a composite disturbance including both time domain and frequency domain, IAFST can adapt to the characteristics of the disturbance type, so as to more accurately extract the characteristic information such as disturbance duration and amplitude.

[0117] To verify the real-time performance of various algorithms in practical applications, the execution time of this method for different types of disturbance detections was tested in the CCS environment of the DSP development platform. The test results are shown in Table 2.

[0118] Table 2 Execution time of IAFST detection for different types of disturbances

[0119]

[0120] As can be seen from Table 2, on the basis of ensuring the detection accuracy, IAFST has good real-time performance, which benefits from the universality of the global adaptive Gaussian window and the fast calculation scheme based on the automatic threshold method, further verifying the effectiveness of the proposed method.

[0121] The above only elaborates in detail on the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those of ordinary skill in the art, various changes can be made without departing from the gist of the present invention, and all such changes should be included within the protection scope of the present invention.

Claims

1. A power quality disturbance detection method based on improved adaptive S transform, characterized in that: The following steps are involved: S1. Design the window width control factors of two Gaussian windows, flexibly adjust the time domain window width and spectrum characteristics of the Gaussian window, and form a global adaptive Gaussian window ω(τ―t, f) as the kernel function of the improved adaptive S transform to fully adapt to the transient power quality disturbance characteristics in the power system; S2. A parameter tuning scheme with enhanced energy concentration ECM as the objective function is proposed. The upper and lower limits of the time-frequency domain resolution are used as constraints, and the interior point method is used to solve the parameters. The result is used as the value of the window width control factor of the global adaptive Gaussian window ω(τ―t, f) in S1. S3, perform fast Fourier transform on the signal and set the threshold. The part above the threshold is defined as the characteristic frequency point k i ; S4, filter out the calculations at non-characteristic frequency points, and only calculate the characteristic frequency point k i Perform time-frequency transformation at the characteristic frequency points, construct a fast algorithm for improving adaptive S transformation, and obtain a time-frequency transformation matrix containing characteristic information at characteristic frequency points; S5. Perform feature extraction based on the time-frequency transformation matrix in S4 to obtain the feature information value of the final disturbance.

2. The power quality disturbance detection method based on improved adaptive S transform according to claim 1 is characterized in that: The structure of the global adaptive Gaussian window in S1 is: Among them, the standard deviation σ(f) is a function of the signal frequency f, and the specific structure is σ(f)=1 / |λ1f+λ2|. The standard deviation can be flexibly controlled by two window width adjustment factors λ1 and λ2, which is the basis for the window function to adapt to various types of disturbance signals; the parameter τ is the time shift factor that controls its position in the time domain, t is the time variable in the time domain signal, and f is the detection signal frequency.

3. The power quality disturbance detection method based on improved adaptive S transform according to claim 1 is characterized in that: The improved adaptive S transform definition in S1 is: Where x(t) is any time domain signal.

4. The power quality disturbance detection method based on improved adaptive S transform according to claim 1, characterized in that: The parameter tuning scheme in S2 with enhancing energy concentration ECM as the objective function includes the following steps: S21. After the signal is subjected to improved adaptive S transform, the energy concentration ECM is defined as: Among them, the normalized IAST matrix is: The goal of the parameter tuning scheme is to make the energy concentration of the detected signal ECM IAST (λ1,λ2) is the largest; S22, using the trade-off between time and frequency resolution as a constraint; S23. Obtain parameter values ​​that can enhance energy concentration while taking into account the trade-off between time-frequency resolution.

5. The power quality disturbance detection method based on improved adaptive S transform according to claim 4 is characterized in that: The method of using the trade-off between time and frequency resolution as a constraint in S22 is: The overall constraints are: s low ≤σ(f)≤σ up σ low For a discrete signal sampling sequence, σ can be low Replace with nT s , T s is the sampling period of the signal, n is an integer greater than or equal to 1, and σ low ≤σ(f) is transformed into constraint one: nT s *(λ1f min +λ2)―1≤0 Among them, f min is the minimum frequency for detection, T s is the sampling period of the signal; σ up Determines the upper limit of frequency resolution. Similarly, σ(f)≤σ up Transformed into constraint 2: 1 / |λ1f max +λ2|―uT s ≤0 Among them, f max Based on the sampling theorem, T is determined by the sampling frequency. s is the sampling period of the signal.

6. The power quality disturbance detection method based on improved adaptive S transform according to claim 5 is characterized in that: The method for selecting the parameter values ​​of S23 that can enhance energy concentration and take into account the trade-off between time-frequency resolution is: In order to improve the speed of parameter optimization, it is also necessary to limit the parameters themselves to a reasonable range as constraint condition three. The overall parameter optimization solution is: s.t.nT s *(λ1f min +λ2)―1≤0 1 / (λ1f max +λ2)―uT s ≤0 λ1∈(―0.1,―0.0015) λ2∈(15,35) The interior point method is used to solve the parameters of the above tuning scheme to obtain parameter values ​​that can enhance energy concentration and take into account the trade-off between time and frequency resolution.

7. The power quality disturbance detection method based on improved adaptive S transform according to claim 1, characterized in that: The method for determining the characteristic frequency point of the disturbance characteristic signal in S3 comprises the following steps: S31, performing fast Fourier transform on the signal to obtain a spectrum of the disturbance signal, preliminarily determining the frequency distribution, and obtaining the portion of the actual disturbance signal that contains the main energy; S32. Set a threshold according to the characteristics of the actual disturbance signal, define the frequency components in the spectrum graph with amplitudes higher than this threshold as the main frequency points, and automatically screen out the remaining non-main frequency points. The number of main frequency points v determined by the automatic threshold method is much smaller than N, thereby reducing the time complexity of the calculation to O(Nlog2N) and reducing the amount of calculation.

8. The power quality disturbance detection method based on improved adaptive S transform according to claim 1, characterized in that: The calculation method for filtering out non-characteristic frequency points in S4 includes the following steps: S41, first discretize it, and discretize the time variables τ and f in the improved adaptive S transform into mT s and k / NT s ; S42, in the process of calculating the two-dimensional time-frequency matrix of the disturbance signal, all rows where non-main frequency points are located are calculated as zero, and only the main frequency point k is calculated. i The discretized expression of the part is: Among them, k i is the main frequency point of the signal selected by the automatic threshold method, the values ​​of m, l are 0, 1, 2, ... N-1, the values ​​of i are 1, 2... v, X((l+k i ) / NT s ) is the discrete form of X(α+f) containing only the main frequency, W(l / NT s ,k i / NT s ) is the discrete Fourier transform of the adaptive Gaussian window.

9. The power quality disturbance detection method based on improved adaptive S transform according to claim 1 is characterized in that: The method for extracting the characteristic information of the final disturbance according to the final time-frequency characteristic graph in S5 comprises the following steps: S51, first extracting the time domain characteristic information of the disturbance signal, extracting the time domain variation curve of a fixed frequency component in the disturbance signal from the time-frequency matrix, calculating the difference curve of the curve to obtain the starting time of the frequency component disturbance, and further extracting the amplitude variation characteristic information of the disturbance signal from the time domain variation curve; S52. According to the final two-dimensional time-frequency matrix, the frequency domain characteristic information such as several frequency components contained in the disturbance signal is extracted, and then, the frequency normalized value of the vertical coordinate in the improved adaptive S-variable two-dimensional time-frequency matrix is ​​extracted to accurately obtain the frequency value of each disturbance signal component.

Images