Flexible DC converter station harmonic evaluation method, device, electronic equipment and medium

Abstract

The present invention discloses a method, device, electronic device and medium for evaluating harmonics in a flexible DC converter station, wherein the method comprises: determining a harmonic source approximate signal according to acquired harmonic data, and performing a local search in the harmonic data based on the harmonic source approximate signal to obtain a local signal; constructing a sparse dictionary according to the local signal, and applying the sparse dictionary to the harmonic data; screening the processed harmonic data according to a preset sparse screening criterion to determine a target local signal in the local signal, and constructing a target global signal according to the target local signal; selecting a target global signal with the strongest non-Gaussianity as a true harmonic source signal, and generating harmonic transmission level evaluation information according to the true harmonic source signal. The embodiment of the present invention can realize the evaluation of the harmonic emission level of a flexible DC converter station in a high-cable scenario, can reduce the influence of the increase in system-side impedance on the evaluation of the harmonic emission level, and can improve the accuracy of the evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of flexible direct current power transmission, and in particular to a method, device, electronic equipment and medium for evaluating harmonics in a flexible direct current converter station. Background Art

[0002] Flexible DC transmission technology is a new generation of high-voltage DC transmission technology based on voltage source converter, self-shutoff device and pulse width modulation technology. It has the characteristics of high efficiency, high stability, high reliability, high flexibility and environmental protection and energy saving. Flexible DC transmission system has been widely used in my country's power system, but harmonic oscillation occurs in the flexible DC transmission system, so it is necessary to evaluate the harmonic emission level of the flexible DC converter station. At present, the existing evaluation of harmonic emission level of harmonic source of flexible DC converter station is mainly based on the two-point assumption that the harmonic impedance of the system side is much smaller than the impedance of the user side and the background harmonic is relatively stable. However, due to the large ground capacitance of the cable, it is easy to cause inductive coupling and increase the harmonic impedance of the system side. Therefore, the impedance of the system side under high cable rate will not meet the assumption that it is much smaller than the impedance of the user side. At present, the traditional evaluation cannot be applied to the harmonic emission level evaluation of the harmonic source of the flexible DC converter station under high cable rate. There is an urgent need for a method that can directly carry out the harmonic emission level evaluation of the flexible DC converter station based on the actual measured harmonic voltage of the common connection point between the flexible DC converter station and the system side and the actual measured harmonic current data of the common connection line. Summary of the invention

[0003] The present invention provides a flexible DC converter station harmonic assessment method, device, electronic device and medium to achieve the assessment of the harmonic emission level of the flexible DC converter station in a high-cable scenario, which can reduce the impact of increased system-side impedance on the harmonic emission level assessment and improve the accuracy of the assessment.

[0004] According to one aspect of the present invention, a method for harmonic assessment of a flexible DC converter station is provided, wherein the method comprises: determining a harmonic source approximate signal based on acquired harmonic data, and performing a local search within the harmonic data based on the harmonic source approximate signal to acquire a local signal; constructing a sparse dictionary based on the local signal, and applying the sparse dictionary to the harmonic data; screening the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal, and constructing a target global signal based on the target local signal; selecting the target global signal with the strongest non-Gaussianity as the true harmonic source signal, and generating harmonic transmission level assessment information based on the true harmonic source signal.

[0005] According to another aspect of the present invention, a harmonic evaluation device for a flexible DC converter station is provided, wherein the device comprises: an approximate local signal module, used to determine a harmonic source approximate signal according to acquired harmonic data, and perform a local search in the harmonic data based on the harmonic source approximate signal to acquire a local signal; a signal sparse processing module, used to construct a sparse dictionary according to the local signal, and apply the sparse dictionary to the harmonic data; a global signal construction module, used to filter the processed harmonic data according to a preset sparse filtering criterion to determine a target local signal in the local signal, and construct a target global signal according to the target local signal; a source signal evaluation module, used to select the target global signal with the strongest non-Gaussianity as the true harmonic source signal, and generate harmonic transmission level evaluation information according to the true harmonic source signal.

[0006] According to another aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the flexible DC converter station harmonic evaluation method described in any embodiment of the present invention.

[0007] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the harmonic evaluation method for a flexible DC converter station described in any embodiment of the present invention when executed.

[0008] The technical solution of the embodiment of the present invention measures harmonic data, determines the harmonic source approximate signal according to the harmonic data, performs local search based on the harmonic source approximate signal to obtain the local signal, generates a sparse dictionary corresponding to the local signal, and applies the sparse dictionary to the harmonic data. The processed harmonic data is screened by a preset sparse criterion to determine the target local signal in each local signal, constructs a target global signal corresponding to each target local signal, selects the target global signal with the strongest non-Gaussianity as the real harmonic source signal, and generates harmonic transmission level evaluation information according to the real harmonic source signal, thereby realizing the evaluation of the harmonic emission level of the flexible DC converter station in a high-cable scenario, reducing the impact of the increase in system-side impedance on the harmonic emission level evaluation, thereby improving the accuracy of the harmonic emission level evaluation, helping to alleviate the harmonic generation of the transmission line, and improving the transmission efficiency of the flexible mainstream converter station.

[0009] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 This is a schematic diagram of harmonic emission level analysis of a flexible DC converter station provided according to an embodiment of the present invention;

[0012] Figure 2 This is a flow chart of a method for evaluating harmonics in a flexible DC converter station provided according to Embodiment 1 of the present invention;

[0013] Figure 3 is a flow chart of another method for evaluating harmonics in a flexible DC converter station provided according to Embodiment 2 of the present invention;

[0014] Figure 4 is a flow chart of another method for evaluating harmonics in a flexible DC converter station provided according to Embodiment 3 of the present invention;

[0015] Figure 5 is a flow chart of another method for evaluating harmonics in a flexible DC converter station provided according to a fourth embodiment of the present invention;

[0016] Figure 6 This is an example diagram of harmonic evaluation of a flexible DC converter station provided according to Embodiment 5 of the present invention;

[0017] Figure 7 2 is a schematic diagram of the structure of a harmonic evaluation device for a flexible DC converter station provided according to a sixth embodiment of the present invention;

[0018] Figure 8 It is a structural schematic diagram of an electronic device for implementing the flexible DC converter station harmonic evaluation method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] At present, for urban power grids with high cable ratio and flexible DC converter stations, buses with high harmonic voltage content divide the above power grid into system side and user side. Its Norton equivalent model is as follows: Figure 1 As shown, according to the superposition principle Figure 1 The harmonic voltage Δ shown in the figure is related to the equivalent harmonic impedance Z on the system side. S , user-side equivalent harmonic impedance Z C , System side equivalent harmonic source And user-side equivalent harmonic source There is a corresponding relationship between the following formulas:

[0022]

[0023] Based on the above formula (1), the harmonic voltage emission level on the user side and the harmonic voltage emission level on the system side are determined as follows (2):

[0024]

[0025] At present, for the working conditions of traditional nonlinear users, since the harmonic impedances on both sides meet |Z C |>>|Z S |, so we have Z S / Z C ≈0, so the formula (2) can be Resolved to:

[0026]

[0027] That is, we only need to solve the system side harmonic impedance Z S The harmonic emission level can be quantified. However, in a high-cable urban power grid with a flexible DC converter station, as the inductive and capacitive coupling occurs between the cable and the inductive elements, the system-side harmonic impedance Z S At some frequencies, it may increase, which will cause |Z C |>>|Z S | is no longer valid. The harmonic emission level cannot be quantified based on the traditional approximate formula. It is necessary to calculate the system-side harmonic impedance Z at the same time. S And the user side harmonic impedance Z C Only then can the harmonic emission level be quantified.

[0028] Embodiment 1

[0029] Figure 2 This is a flowchart of a method for evaluating harmonics in a flexible DC converter station according to the first embodiment of the present invention. This embodiment is applicable to the evaluation of harmonic emission levels in flexible DC converter stations in high-cable scenarios. The method can be performed by a flexible mainstream device. The flexible DC converter station harmonic evaluation device can be implemented in the form of hardware and / or software. The flexible DC converter station harmonic evaluation device can be configured in a server or terminal device. Figure 2 As shown, the method includes:

[0030] Step 110: determine a harmonic source approximate signal according to the acquired harmonic data, and perform a local search in the harmonic data based on the harmonic source approximate signal to acquire a local signal.

[0031] Among them, the point of common coupling (PCC) can be the connection point of more than one user load in the power system, and the harmonic voltage and harmonic current can be measured at the common connection point. The equivalent harmonic source on the system side and the equivalent harmonic source on the user side can be equivalent harmonic sources determined based on the superposition principle. Harmonics can be waveform distortion caused by the nonlinear rapid time-varying of certain power equipment in the power system, resulting in the non-linear relationship between voltage and current. Harmonics will reduce the efficiency of electric energy generation, transmission and utilization, and harmonic data can be data related to harmonics measured at the common connection point. The harmonic data may include harmonic voltage, harmonic current, equivalent harmonic source on the system side and equivalent harmonic source on the user side, etc. The harmonic source approximate signal can be a signal of an approximate harmonic source determined by analyzing the harmonic data. The harmonic source approximate signal can be obtained by analyzing the main signal data in the harmonic data. The method for obtaining the harmonic source approximate signal may include performing independent component analysis on the harmonic data or generating it through FastICA algorithm processing. The local signal may be a signal located near the harmonic source approximate signal in the harmonic data. The local signal may be searched and determined in the harmonic data with the harmonic source signal as the center.

[0032] In an embodiment of the present invention, the voltage, current and other data of harmonics can be measured at the common connection point of the flexible DC converter station to obtain harmonic data. Since the harmonic data can be data formed by one or more harmonics, the harmonic data can be analyzed to extract the harmonic source approximate signal, and then a local search can be performed based on each harmonic source approximate signal, and the searched signal can be used as a local signal. It can be understood that the local search based on the harmonic source approximate signal can include an exhaustive search within a specified local signal length for each harmonic source approximate signal, or a local signal search according to a specified local signal step length for each harmonic source approximate signal.

[0033] Step 120: construct a sparse dictionary based on the local signal, and apply the sparse dictionary to the harmonic data.

[0034] Among them, the sparse dictionary can be a dimensionality reduction representation of the local signal. The sparse dictionary can be obtained by minimizing the cost using the local signal. The sparse dictionary can use as few atoms as possible to represent the local signal.

[0035] In an embodiment of the present invention, a corresponding sparse dictionary can be created for each searched local signal. Each sparse dictionary can be represented by as few atoms as possible in the corresponding local signal. The sparse dictionary of each local signal can be generated by processing methods such as sparse coding (SC), orthogonal matching pursuit (OMP) or K-SVD algorithm. Each generated sparse dictionary can be used to process the harmonic data separately, so that the processed harmonic data can be subsequently judged using sparse screening criteria, thereby obtaining the local signal of the real harmonic source.

[0036] Step 130 : Screen the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal, and construct a target global signal according to the target local signal.

[0037] Among them, the preset sparse screening criterion can be the basis for judging the harmonic data after sparse dictionary processing, the preset sparse screening criterion can be pre-configured, and the preset sparse screening criterion can at least include an evaluation condition for whether the harmonic data is sparse. The target local signal can be a local signal that meets the harmonic source selection conditions, the local signal can be a local signal that is highly consistent with the real source signal of the harmonic, and the target global signal can be a global signal constructed by the target local signal, and the target global signal can include the overall signal of the local signal.

[0038] In an embodiment of the present invention, a preset sparse screening criterion can be used to judge the harmonic data processed by each sparse dictionary in turn to determine whether the processed harmonic data meets the sparse characteristics. If so, the local signal corresponding to the sparse dictionary is determined to be a target local signal that is highly matched with the harmonic source signal. If not, the preset sparse screening criterion can be used to judge the harmonic data processed by the next sparse dictionary until all the processed harmonic data corresponding to the sparse dictionaries are judged. For all determined target local signals, a target global signal can be generated according to the correspondence between the harmonic impedances in the target local signals.

[0039] Step 140: Select the target global signal with the strongest non-Gaussianity as the real harmonic source signal, and generate harmonic transmission level assessment information according to the real harmonic source signal.

[0040] Among them, non-Gaussianity can be the degree to which the data distribution of the signal deviates from the Gaussian distribution. Non-Gaussianity can be measured by indicators such as kurtosis and negative entropy. Since the stronger the non-Gaussianity, the more independent the signal, the target global signal with the strongest non-Gaussianity can be used as the real harmonic source signal. Harmonic transmission level evaluation information can be information for evaluating the harmonic power emission level, and the harmonic transmission level evaluation information can include impedance on the system side and the user side, etc.

[0041] In an embodiment of the present invention, the non-Gaussianity of each target global signal is determined, and the non-Gaussianity can be determined by at least one of the indicators such as the kurtosis and the negative entropy of the target global signal. The non-Gaussianity of each target global signal can be compared, and the target global signal with the highest non-Gaussianity can be used as a real harmonic source signal. According to the voltage, current and other information of the real harmonic source signal, the system side harmonic impedance indicating the harmonic transmission level and the user side harmonic impedance and other information can be determined.

[0042] In an embodiment of the present invention, harmonic data at a common connection is measured, and a harmonic source approximate signal is determined according to the harmonic data. A local search is performed based on the harmonic source approximate signal to obtain a local signal, a sparse dictionary corresponding to the local signal is generated, and the sparse dictionary is applied to the harmonic data. The processed harmonic data is screened by a preset sparse criterion to determine a target local signal in each local signal, a target global signal corresponding to each target local signal is constructed, and a target global signal with the strongest non-Gaussian property is selected as a true harmonic source signal. Harmonic transmission level evaluation information is generated according to the true harmonic source signal, thereby realizing the evaluation of the harmonic emission level of a flexible DC converter station in a high-cable scenario. The evaluation process does not need to consider the assumption that the system side impedance is much smaller than the user side impedance, and can reduce the influence of the increase in the system side impedance on the harmonic emission level evaluation, thereby improving the accuracy of the harmonic emission level evaluation, helping to alleviate the harmonic generation of the transmission line, and improving the transmission efficiency of the flexible mainstream converter station.

[0043] Embodiment 2

[0044] Figure 3 is a flowchart of another flexible DC converter station harmonic evaluation method provided according to the second embodiment of the present invention. The embodiment of the present invention is a specific embodiment of the above invention, and the processing method of harmonic data and the search process of local signals are described in detail. Figure 3 The method provided in the embodiment of the present invention specifically includes the following steps:

[0045] Step 210: decouple the real and imaginary parts of the harmonic voltage, harmonic current, system-side equivalent harmonic source, and user-side equivalent harmonic source in the harmonic data to construct a real and imaginary part decoupling matrix.

[0046] Among them, the harmonic voltage can be the voltage data of the harmonic measured at the common connection point at different times, the harmonic current can be the current data of the harmonic measured at the common connection point at different times, and the equivalent harmonic source on the system side and the equivalent harmonic source on the user side can be the equivalent harmonic source determined by the load equivalent on the system side or the user side respectively, the real and imaginary part decoupling can be the process of separating the real and imaginary parts of the harmonic data, the real and imaginary part decoupling matrix can be the data matrix generated by the harmonic data after the real and imaginary part separation processing, the real and imaginary part decoupling matrix can include the real and imaginary information of the harmonic voltage, the harmonic current, the equivalent harmonic source on the system side and the equivalent harmonic source on the user side, and further, the real and imaginary part decoupling matrix can also include mixed sparseness, and the value of the mixing coefficient can be determined by the equivalent harmonic impedance on the user side and the equivalent harmonic impedance on the system side.

[0047] In an embodiment of the present invention, in order to determine the approximate signal of the harmonic source, it is necessary to convert the harmonic data into independent data. Therefore, this can be achieved by extracting the fast-changing components of the harmonic data, and generating the fluctuation amounts of different signals based on the fast-changing components, so that the equivalent harmonic source on the system side and the equivalent harmonic source on the user side are approximately independent of each other. The real and imaginary parts of the harmonic voltage, harmonic current, equivalent harmonic source on the system side, and equivalent harmonic source on the user side in the harmonic data can be decoupled, so that the real and imaginary decoupling matrix is ​​composed of the real and imaginary values ​​of the harmonic voltage, harmonic current, equivalent harmonic source on the system side, and equivalent harmonic source on the user side.

[0048] Step 220, construct a first matrix based on the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the harmonic imaginary part determined by the real and imaginary part decoupling matrix, and construct a second matrix based on the real part of the system side equivalent harmonic source, the imaginary part of the system side equivalent harmonic source, the real part of the user side equivalent harmonic source and the imaginary part of the user side equivalent harmonic source determined by the real and imaginary part decoupling matrix.

[0049] Specifically, the first matrix can be a matrix composed of the real part of harmonic voltage, the imaginary part of harmonic voltage, the real part of harmonic current and the imaginary part of harmonic, and the elements in the first matrix can be the real part of harmonic voltage, the imaginary part of harmonic voltage, the real part of harmonic current and the imaginary part of harmonic data, respectively, and the second matrix is ​​constructed by the real part of the equivalent harmonic source on the system side, the imaginary part of the equivalent harmonic source on the system side, the real part of the equivalent harmonic source on the user side and the imaginary part of the equivalent harmonic source on the user side.

[0050] Step 230: Based on the blind source separation model matrix determined by the first matrix and the second matrix, determine the harmonic source approximation signal according to a preset FastICA algorithm.

[0051] The blind source separation model matrix may be a matrix for classifying harmonic source approximation signals from harmonic data, and the blind source separation model matrix may be determined by the correspondence between the first matrix and the second matrix, and the correspondence may include X=AS, where X is the first matrix, S is the second matrix, and A is the mixing matrix. The FastICA algorithm may be an algorithm for amplifying the non-Gaussianity of the signal by recursion, thereby separating the signal, and the FastICA algorithm may be composed of processes such as data preprocessing, centering operation, data whitening, ICA iteration, and blind source separation.

[0052] In an embodiment of the present invention, the first matrix and the second matrix can be used to construct a blind source separation matrix, which can meet the basic requirements of data preprocessing and centralization operations of the preset FastICA algorithm. The elements in the blind source separation matrix can be subjected to data whitening and ICA iteration and other processes to determine a separation matrix W for separating the harmonic source approximation signal. The separation matrix W can be called to process the blind source separation matrix to generate a harmonic source approximation signal with non-Gaussianity. It can be understood that the separation matrix W determined by the preset FastICA algorithm can multiply the blind source separation matrix by the transpose of the separation matrix W to restore the harmonic source approximation signal as an independent signal.

[0053] Step 240: Search and obtain a local signal within a preset local signal length range of the harmonic source approximation signal of the harmonic data.

[0054] Among them, the preset local signal length range can be the signal length for searching the local signal. The preset local signal length range requirement can be set based on experience. In some embodiments of the invention, the preset local signal length range can be set to 5.

[0055] In the embodiment of the present invention, for each determined harmonic source approximation signal, a search signal may be traversed within the range of a preset local signal length of the harmonic source approximation signal, and the searched signal may be used as a local signal.

[0056] Step 250: construct a sparse dictionary based on the local signal, and apply the sparse dictionary to the harmonic data.

[0057] Step 260 : Screen the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal, and construct a target global signal according to the target local signal.

[0058] Step 270: Select the target global signal with the strongest non-Gaussianity as the real harmonic source signal, and generate harmonic transmission level assessment information according to the real harmonic source signal.

[0059] In an embodiment of the present invention, the real and imaginary parts of the harmonic voltage, harmonic current, system-side equivalent harmonic source and user-side equivalent harmonic source in the harmonic data are decoupled to generate a real and imaginary part decoupling matrix, the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the imaginary part of the harmonic current in the real and imaginary part matrix are extracted as a first matrix, and the real part of the system-side equivalent harmonic source, the imaginary part of the system-side equivalent harmonic source, the real part of the user-side equivalent harmonic source and the imaginary part of the user-side equivalent harmonic source are generated into a second matrix, the blind source separation model matrix corresponding to the first matrix and the second matrix is ​​determined, and the blind source separation model matrix is ​​separated into harmonic source approximate signals according to the Fastit algorithm, and local signals are searched and obtained within a preset local signal length range of the harmonic source approximate signal of the harmonic data, and the processed harmonic data are screened by a preset sparse criterion so as to obtain local signals in each local area. The target local signal is determined in the signal, and the target global signal corresponding to each target local signal is constructed. The target global signal with the strongest non-Gaussianity is selected as the real harmonic source signal, and the harmonic transmission level evaluation information is generated according to the real harmonic source signal, so as to realize the evaluation of the harmonic emission level of the flexible DC converter station in the high cable scenario. The harmonic data is separated into independent harmonic source approximation signals through the FasitICA algorithm to estimate the equivalent harmonic source impedance on the system side and the equivalent harmonic source impedance on the user side, so that the evaluation process does not need to consider the assumption that the system side impedance is much smaller than the user side impedance, which can reduce the impact of the increase in system side impedance on the harmonic emission level evaluation, thereby improving the accuracy of the harmonic emission level evaluation, helping to alleviate the harmonic generation of the transmission line and improve the transmission efficiency of the flexible mainstream converter station.

[0060] Embodiment 3

[0061] Figure 4 is a flowchart of another flexible DC converter station harmonic evaluation method provided according to the third embodiment of the present invention. The embodiment of the present invention is a specific embodiment of the above invention embodiment, and the sparse processing process of harmonic data is described in detail. Figure 4 The method provided in the embodiment of the present invention specifically includes the following steps:

[0062] Step 310: determine a harmonic source approximate signal according to the acquired harmonic data, and perform a local search in the harmonic data based on the harmonic source approximate signal to acquire a local signal.

[0063] Step 320: Generate a sparse dictionary of the local signal according to a preset dictionary learning algorithm and the sparsity of the local signal.

[0064] The preset dictionary learning (KSVD) algorithm may be a dictionary learning method in sparse representation, wherein the preset dictionary learning algorithm requires K iterations, and each iteration uses SVD decomposition. The sparsity may be the number of non-quantized elements of the signal after the local signal is sparsely represented.

[0065] In the embodiment of the present invention, each local signal may be processed according to its corresponding sparsity by using a preset dictionary learning algorithm, thereby generating a sparse dictionary corresponding to the local signal.

[0066] Step 330: transform the first matrix corresponding to the harmonic voltage and the harmonic current in the harmonic data according to the sparse dictionary.

[0067] In an embodiment of the present invention, the first matrix may be composed of real data and imaginary data of harmonic voltage and harmonic current, and the sparse dictionary may be used to transform the first matrix, which may include multiplying the first matrix by the inverse transpose of the dictionary matrix corresponding to the sparse dictionary.

[0068] Step 340: Determine a third matrix in the processed harmonic data, wherein the third matrix is ​​composed of the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current, and the imaginary part of the harmonic current processed by the sparse dictionary.

[0069] In an embodiment of the present invention, the third matrix can be determined by the harmonic data that has been sparsely processed, and the data corresponding to the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current, and the imaginary part of the harmonic current can be extracted from the sparsely processed harmonic data as the third matrix. It can be understood that when there are multiple local signals, the third matrix generated by the sparse dictionary processing corresponding to the local signals can also be multiple.

[0070] Step 350: When the third matrix is ​​a sparse matrix, it is determined that the harmonic data after the sparse dictionary meets the preset sparse screening criterion, and the local signal corresponding to the current sparse dictionary is used as the target local signal.

[0071] In an embodiment of the present invention, the third matrix can be judged to determine whether the generated third matrix is ​​a sparse matrix, that is, whether the elements in the third matrix have sparse characteristics. If so, it is determined that the harmonic data after the sparse dictionary conversion meets the preset sparse filtering conditions. Otherwise, it is determined that the harmonic data after the sparse dictionary conversion does not meet the preset sparse filtering conditions. For the sparse dictionary that meets the preset sparse filtering conditions, the local signal corresponding to the sparse dictionary can be used as the target local signal.

[0072] Step 360: Obtain a third matrix corresponding to the target local signal, and determine the clustering line slope of the third matrix.

[0073] Among them, the clustering line slope can be used to distinguish the slopes of the clustering lines of different target local signals, the clustering execution line can be determined by the linear relationship satisfied respectively between each target local signal, and the clustering line slope can include the proportional relationship between the elements of each target local signal.

[0074] In an embodiment of the present invention, the corresponding third matrix can be extracted for each target local signal, that is, the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the imaginary part of the harmonic current are extracted in each target local signal to form a third matrix, and then the linear relationship between the elements of each third matrix can be determined as the clustering straight line slope of the third matrix. For example, the proportional relationship between the elements in the same position on different rows in the third matrix can be used as the clustering straight line slope of the third matrix.

[0075] Step 370: Generate a target global signal according to the clustering line slope and the target local signal.

[0076] In an embodiment of the present invention, the slope of the clustering line may be a linear relationship that the target local signal conforms to, and the linear relationship may be satisfied by the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current, and the imaginary part of the harmonic current in the target local signal. On this basis, based on the correspondence between the harmonic voltage and the harmonic current and the harmonic impedance in the target local signal, the user-side harmonic impedance and the system-side harmonic impedance of the target local signal may be determined by the above-mentioned clustering line, and the user-side harmonic impedance and the system-side harmonic impedance determined by the target local signal may be used to reconstruct the target global signal. Since the target global signal may conform to the above-mentioned determined user-side harmonic impedance and system-side harmonic impedance, on the basis of determining the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current, and the imaginary part of the harmonic current, the user-side equivalent harmonic source and the system-side equivalent harmonic source may be determined as the target global signal in combination with the user-side harmonic impedance and the system-side harmonic impedance.

[0077] Step 380: Select the target global signal with the strongest non-Gaussianity as the real harmonic source signal, and generate harmonic transmission level assessment information according to the real harmonic source signal.

[0078] In an embodiment of the present invention, a harmonic source approximate signal is determined by independently analyzing the harmonic data, and a local signal is searched in the harmonic data according to the harmonic source approximate signal, a sparse dictionary corresponding to the local dictionary is generated according to the sparsity of the local signal and a preset dictionary learning algorithm, and a first matrix composed of harmonic voltage and harmonic current in the harmonic data is sparsely transformed by the sparse dictionary, and the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current, and the imaginary part of the harmonic current generated by each sparse dictionary after processing are extracted to construct a third matrix. When it is determined that the third matrix is ​​a sparse matrix, the local signal corresponding to the sparse dictionary that generates the third matrix is ​​used as the target local signal, and the third matrix is ​​generated according to the third matrix. The matrix generates the clustering line slope of the target local signal, generates the target global signal according to the clustering line slope and the target local signal, selects the target global signal with the strongest non-Gaussianity as the real harmonic source signal, and generates the harmonic transmission level assessment information according to the real harmonic source signal, thereby realizing the estimation of the equivalent harmonic source impedance on the system side and the equivalent harmonic source impedance on the user side. The assessment process does not need to consider the assumption that the system side impedance is much smaller than the user side impedance, which can reduce the impact of the increase in system side impedance on the harmonic emission level assessment, thereby improving the accuracy of the harmonic emission level assessment, helping to alleviate the harmonic generation of the transmission line and improve the transmission efficiency of the flexible mainstream converter station.

[0079] Embodiment 4

[0080] Figure 5 is a flowchart of another flexible DC converter station harmonic evaluation method provided according to the fourth embodiment of the present invention. The embodiment of the present invention is a specific embodiment of the above invention embodiment, and the screening process of the total negative entropy is described. Figure 5 The method provided in the embodiment of the present invention specifically includes the following steps:

[0081] Step 410: determine a harmonic source approximate signal according to the acquired harmonic data, and perform a local search in the harmonic data based on the harmonic source approximate signal to acquire a local signal.

[0082] Step 420: construct a sparse dictionary based on the local signal, and apply the sparse dictionary to the harmonic data.

[0083] Step 430 : Screen the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal, and construct a target global signal according to the target local signal.

[0084] Step 440: preprocess the target global signal so that the target global signal has zero mean and unit variance.

[0085] In an embodiment of the present invention, the target global signal may be preprocessed, and the preprocessing process may include transforming the display form or value representation mode of the target global signal, so that the target global signal can meet the requirements for subsequent determination of negative entropy. The preprocessed target global signal needs to have zero mean and unit variance. In some embodiments, the preprocessing process of the target global signal may include: Where S represents the target global signal, represents the mean of the target global signal, and std(S) represents the standard deviation of the target global signal.

[0086] Step 450, determine the negative entropy of the target global signal according to the preset negative entropy approximate calculation formula, and extract the maximum negative entropy of the real part, the minimum negative entropy of the real part, the maximum negative entropy of the imaginary part, and the minimum negative entropy of the imaginary part from the negative entropy.

[0087] Among them, the preset negative entropy approximate calculation formula can be a formula for performing negative entropy estimation calculation on the target global signal, and the real maximum negative entropy, real minimum negative entropy, imaginary maximum negative entropy and imaginary minimum negative entropy of the target global signal can be the maximum negative entropy value of the real part, the minimum negative entropy value of the real part, the maximum negative entropy value of the imaginary part and the minimum negative entropy value of the imaginary part in a target global signal, respectively.

[0088] In an embodiment of the present invention, the target global signal can be input into a preset negative entropy approximation calculation formula to determine the negative entropy of the target global signal, and the negative entropy of the target global signal can be compared to select the maximum real part negative entropy, the minimum real part negative entropy, the maximum imaginary part negative entropy and the minimum imaginary part negative entropy of the target global signal with the largest or smallest value.

[0089] Based on the above-mentioned embodiment of the invention, a negative entropy approximate calculation formula is preset, including:

[0090] J(s)≈[E{G(s)}-E{G(s Gauss )}] 2

[0091] Where s represents the target global signal, G(s Gauss ) represents a Gaussian random signal, E{} represents the expectation, G(s) = -exp(-s 2 / 2), and exp represents a natural exponential function.

[0092] Step 460: For each target global signal, the negentropy is normalized according to its respective maximum real part negentropy, minimum real part negentropy, maximum imaginary part negentropy and minimum imaginary part negentropy.

[0093] Among them, the normalization processing can be a processing of normalizing the values ​​of the negative entropy of different target global signals. The normalization processing can unify the comparison standard of the negative entropy of different target global signals, thereby realizing the comparison of the negative entropy of different target global signals and determining the target global signal with the strongest non-Gaussianity.

[0094] In an embodiment of the present invention, the negative entropy can be normalized according to its maximum negative entropy, minimum negative entropy of the real part, maximum negative entropy of the imaginary part and minimum negative entropy of the imaginary part, respectively. The proportion of the real part of the negative entropy in the difference between the maximum negative entropy of the real part and the minimum negative entropy of the real part can be used as the normalized real negative entropy. Similarly, the proportion of the imaginary part of the negative entropy in the difference between the maximum negative entropy of the imaginary part and the minimum negative entropy of the imaginary part can be used as the normalized imaginary negative entropy.

[0095] Step 470: determine the comprehensive negative entropy value corresponding to each normalized negative entropy, and find the maximum comprehensive negative entropy value among the comprehensive negative entropy values.

[0096] Among them, the comprehensive negative entropy value can be determined by the normalized negative entropy of the target global signal, and the comprehensive negative entropy value can be determined by the real part of the negative entropy and the imaginary part of the negative entropy. For example, the comprehensive negative entropy value can include the average value of the real part of the negative entropy and the imaginary part of the negative entropy or the weighted sum of the real part of the negative entropy and the imaginary part of the negative entropy, etc.

[0097] In an embodiment of the present invention, for each target global signal, its normalized negative entropy can be extracted, and a comprehensive negative entropy value can be generated according to the real part and imaginary part of its negative entropy. In combination with the specific application scenario, the real part and imaginary part of the negative entropy may have different or the same influence on the comprehensive negative entropy value. For example, the comprehensive negative entropy value may be determined by the weighted sum of the real part of the negative entropy and the imaginary part of the negative entropy, and the weight of the real part of the negative entropy may be different from the weight of the imaginary part of the negative entropy. The comprehensive negative entropy values ​​of each target global cycle can be compared, and the comprehensive negative entropy value with the largest value is obtained as the maximum comprehensive negative entropy value.

[0098] Step 480: Use the target global signal with the maximum comprehensive negative entropy value as the real harmonic source signal, and determine the harmonic impedance of the target global signal as harmonic transmission level assessment information.

[0099] In an embodiment of the present invention, the target global signal corresponding to the maximum comprehensive negative entropy value can be extracted, and the target global signal can be used as a real harmonic signal. The corresponding user-side harmonic impedance and system-side harmonic impedance can be determined for the real harmonic signal as harmonic transmission level assessment information.

[0100] In an embodiment of the present invention, harmonic data at a common connection is measured, and a harmonic source approximate signal is determined according to the harmonic data. A local search is performed based on the harmonic source approximate signal to obtain a local signal, a sparse dictionary corresponding to the local signal is generated, and the sparse dictionary is applied to the harmonic data. The processed harmonic data is screened by a preset sparse criterion to determine a target local signal in each local signal, a target global signal corresponding to each target local signal is constructed, and the target global signal is preprocessed so that the real part and the imaginary part of the target global signal have a zero mean and a unit equation respectively, and a negative entropy is called. The approximate calculation formula determines the negative entropy of the target global signal, and determines the maximum real negative entropy, the minimum real negative entropy, the maximum imaginary negative entropy and the minimum imaginary negative entropy in the negative entropy. According to the maximum real negative entropy, the minimum real negative entropy, the maximum imaginary negative entropy and the minimum imaginary negative entropy, the negative entropy of the corresponding target global signal is normalized to generate a comprehensive negative entropy value of the normalized negative entropy. The comprehensive negative entropy values ​​of the target global signal are compared to determine the target global signal with the maximum comprehensive negative entropy value as the real harmonic source signal, and determine the harmonic impedance of the target global signal as the harmonic transmission level evaluation information. The estimation of the equivalent harmonic source impedance on the system side and the equivalent harmonic source impedance on the user side is realized. The evaluation process does not need to consider the assumption that the impedance on the system side is much smaller than the impedance on the user side, which can reduce the impact of the increase in the impedance on the system side on the harmonic emission level evaluation, thereby improving the accuracy of the harmonic emission level evaluation, helping to alleviate the harmonic generation of the transmission line and improve the transmission efficiency of the flexible mainstream converter station.

[0101] Embodiment 5

[0102] Figure 6 is an example diagram of harmonic evaluation of a flexible mainstream converter station provided according to the fifth embodiment of the present invention, see Figure 6 In an embodiment of the present invention, the collected harmonic data can be separated by the FastICA algorithm. However, due to the large signal separation error of the FastICA algorithm, the real source signal S is not completely consistent with the source signal Y obtained by separation, but is located near the source signal Y determined by separation, but there is a certain difference with Y. The embodiment of the present invention combines SCA with FastICA and adopts a global-local-global approach. First, the FastICA algorithm is used to separate the global approximate signal of the source signal, and then a sparse screening criterion is constructed by SCA. A local search is performed near the separated source signal to find a local signal that may be consistent with the real source signal; finally, the global signal corresponding to the local signal is reconstructed, and the harmonic impedance is obtained by comprehensive negative entropy screening.

[0103] First, the harmonic data is analyzed by the FastICA algorithm. According to the central limit theorem, the Gaussianity of the linear combination of independent random signals will be enhanced. Based on this, the FastICA algorithm considers the independence between the source signals and applies a linear change W to the observed signal X (that is, the collected harmonic data). T X, find the value that makes W T X has the maximum non-Gaussianity W, then W T X is the separated source signal. Among them, non-Gaussianity can be measured by indicators such as kurtosis and negative entropy. Among them, since the robustness of negative entropy has a wider application value, in an embodiment of the present invention, the observation signal X can be processed based on the FastICA algorithm of maximum negative entropy, wherein the negative entropy approximate calculation formula may include:

[0104] J(s)≈[E{G(s)}-E{G(s Gauss )}] 2 (4)

[0106] Where s represents the observed signal, G(s Gauss ) represents a Gaussian random signal, E{} represents the expectation, G(s) represents a non-square nonlinear function, exp represents a natural exponential function, s and s Gauss They all have zero mean and unit equation. The larger the value of J(s), the stronger the non-Gaussianity of the observed signal s.

[0107] The FastICA algorithm requires that each source signal is independent of each other. This can be achieved by extracting the fast-changing components of each harmonic source signal. The fast-changing components can be extracted by calculating the fluctuation of each signal, and then the equivalent harmonic source on the system side is obtained. And user-side equivalent harmonic source are approximately independent, and the fluctuations of the real and imaginary parts of the source signal are also approximately independent. and The real and imaginary parts of are decoupled as follows:

[0108]

[0109] Among them, ΔU PCC-x , ΔU PCC-y , ΔI PCC-x and ΔI PCC-y They are and The real and imaginary parts, ΔI c-x , ΔI c-y , ΔI s-x and ΔI s-y They are and The real and imaginary parts of ijis the mixing coefficient, i and j = 1, 2, 3 or 4 respectively. The value of the mixing coefficient is only related to the harmonic impedance on both sides. The subscript x can represent the real part, and the subscript y can represent the imaginary part.

[0110] Let X = [ΔU PCC-x ,ΔU PCC-y ,ΔI PCC-x ,ΔI PCC-y ] T , S=[ΔI c-x ,ΔI c-y ,ΔI s-x ,ΔI s-y ] T , the corresponding blind source separation model matrix can be obtained by formula (5), which is as follows:

[0111] X=AS(6)

[0112] The FastICA algorithm aims to maximize the negative entropy. It first performs centralization and whitening preprocessing on X, and then optimizes the unmixing matrix W row by row to separate the source signal from the observed signal one by one, and obtains the harmonic source approximate signal Y = [Y c-x ,Y c-y ,Y s-x ,Y s-y ] T , corresponding to [ΔI c-x ,ΔI c-y ,ΔI s-x ,ΔI s-y ] T , but the corresponding order is unknown, so there is Through matrix transformation:

[0113]

[0114] Through the matrix The harmonic impedance of the system side and the user side can be calculated. However, the obtained harmonic impedance is different from the harmonic impedance Z of the system side. S And the user side harmonic impedance Z C The corresponding relationship is uncertain, but since the real part of the impedance is always positive, The elements of any column j of are:

[0115]

[0116] Among them, Z S-X and Z C-X Z S and Z C The real part of .

[0117] Then the matrix The same column of can be used to obtain the imaginary part of the corresponding harmonic impedance as shown in formula (9). So far, the FastICA algorithm can be used to obtain Z S and Z C .

[0118]

[0119] When Z S and Z C The amplitude satisfies |Z C |>>|Z S |, the FastICA algorithm can accurately calculate the harmonic impedance Z on the smaller side S ; But when Z S and Z C When the amplitudes are close, such as at certain frequencies in urban power grids with high cable rates, the errors in the harmonic impedances on both sides are large. Since the FastICA algorithm takes the maximum negative entropy of each separated signal as the optimization goal, the sources of its calculation errors can be discussed from the following two points:

[0120] (1) Under-optimization: The non-Gaussianity of the separated source signal Y is smaller than that of the true source signal S, that is, the optimization is not thorough enough, resulting in large calculation errors.

[0121] (2) Over-optimization: In theory, the central limit theorem is strictly valid only when the number of random signals tends to infinity, but in practice the number of source signals is usually small, taking the number of source signals of the present invention as 4 as an example. After a small number of source signals are linearly combined, the Gaussianity will usually still increase, but the possibility of decrease is not ruled out. FastICA establishes the objective function with the maximum negative entropy, and sometimes over-optimization occurs, that is, the negative entropy value of the separated source signal is larger than that of the real source signal.

[0122] In the embodiment of the present invention, the dimension of the source signal is relatively large, and the central limit theorem holds true in most cases. Therefore, most of the FastICA errors come from under-optimization, and a few come from over-optimization.

[0123] In order to accurately calculate the harmonic impedance on both sides of the flexible DC converter station with a high cabling rate, the present invention aims at the above two points. The embodiment of the present invention combines SCA with FastICA, adopts the idea of ​​"global-local-global", first separates the global approximation of the source signal by FastICA; then constructs a sparse screening criterion by SCA, searches locally near the separated source signal, and finds the local signal that may be consistent with the real source signal; finally, reconstructs the global signal corresponding to the local signal, and uses comprehensive negative entropy screening to obtain the harmonic impedance. The principle of the SCA algorithm is as follows.

[0124] SCA uses the sparsity of the signal to extract the source signal. When the source signal S is sparse, its value is zero most of the time and non-zero a few times. It can be considered that the mixed signal X is generated by at most one of the source signals at each moment. At this time, X is linearly clustered, and the mixing matrix can be calculated from the slope of the clustering line.

[0125] In practical applications, the source signal S is usually not a sparse signal and needs to be converted into a sparse domain through a sparse dictionary D to obtain the sparse source signal S (D -1 ) T For equation (6), the sparse dictionary D of the source signal S is used to transform S and the mixed signal X at the same time, and we have

[0126] X(D -1 ) T =AS(D -1 ) T (10)

[0127] It can be seen that this transformation does not change the mixing matrix A.

[0128] A local search is performed near the source signal Y separated by FastICA. The local signal y found corresponds to the three rows of the real source signal S. y corresponds to [ΔI c-x ,ΔI c-y ,ΔI s-x ] T For the i-th group of local signals y i , the KSVD algorithm can be used to obtain the sparse dictionary D according to the required sparsity (the number of non-zero elements of the signal after sparseness), and D is applied to the mixed signal X at the corresponding time. From formula (10), it can be seen that D is equivalent to acting on the real source signal S. If y i If it matches the corresponding moment S, then S can also be sparse, corresponding to X(D -1 ) T If only one source signal is generated, the local sampling point is called a single source point; on the contrary, if y i If it does not match S, then the corresponding X(D -1 ) T Generated by each source signal, this local sampling point is called a multi-source point

[0129] Single source point:

[0130] At this time i It matches S of the corresponding period, sets the sparsity to 1 and the non-zero values ​​of the signal are at the same time, and uses y i The sparse dictionary D can be [ΔI c-x ,ΔI c-y ,ΔI s-x ] T Sparse. iThe sampling point with value zero, X(D -1 ) T Only the sparse transformed ΔI s-y Produced, so for a single source point:

[0131]

[0132] The subscript “sparse” indicates that the signal is obtained after sparse transformation by D.

[0133] Then we can get:

[0134]

[0135] It can be seen that after sparse transformation, X(D -1 ) T Linear clustering: The proportional relationship between the elements in a column in A can be obtained from the slope of the clustering line.

[0136] Multiple sources:

[0137] At this time i It does not coincide with the corresponding time S, through y i The sparse dictionary D cannot make S sparse, X(D -1 ) T It is still generated by multiple source signals. Only when equation (13) holds true, X(D -1 ) T In order to achieve linear clustering, most local signals do not meet this condition.

[0138]

[0139] Therefore, X(D -1 ) T Linear clustering is used as the searched i-th group of local signals y i The necessary condition for the real source signal S to be consistent with the corresponding time. i Sparse screening criterion for consistency with S:

[0140]

[0141] Where: Re(.) and Im(.) represent the real and imaginary parts respectively; ε represents the threshold set for sparse screening, which can be set to 0.05 according to experience; j, k = 1, 2, 3, 4, and j ≠ k; m represents the local signal y i The locations of sample points whose values ​​are zero after thinning.

[0142] The slope of the clustered straight line can be used to find the proportional relationship between the elements in a column of the mixing matrix A. Then, the corresponding side harmonic impedance can be calculated by combining equations (8), (9) and (12).

[0143] When performing a local search near the source signal Y separated by FastICA, if the local signal length Llocal<5, then formula (14) is always valid, and the sparse screening criterion fails, so Llocal≥5 is required. However, a large Llocal will increase the burden of local search. In the embodiment of the present invention, Llocal=5 can be taken. At this time, the search space dimension is 15, and the search space dimension can be reduced to 6 by further performing a dimensionality reduction transformation on the signal. Taking formula (14) as the objective function, the above search process can be completed by performing an exhaustive search near Y.

[0144] In the embodiment of the present invention, the sparse screening criterion is the local signal y i The necessary condition for the local signal y to be consistent with the real source signal S at the corresponding moment. i In the above equation, there are signals that are highly consistent with S, and there are also signals that make X(D -1 ) T Therefore, it is necessary to filter the local signals y obtained after sparse screening. i Conduct secondary screening.

[0145] The screening method is studied from the perspective of the non-Gaussianity of the signal. The error of FastICA comes from under-optimization or over-optimization of the non-Gaussianity of the signal. Since the central limit theorem is approximately valid, under-optimization is the main problem in practice, while over-optimization is rare. Even if over-optimization occurs, it often only occurs on a certain source signal, while the other source signals are still in a sub-optimized state. Therefore, when the error of the separated signal is large, there are generally two phenomena: 1) The non-Gaussianity of each separated signal is weak; 2) The non-Gaussianity of one of the separated signals is strong, while the Gaussianity of the other signals is weak. For the real source signal, each signal has a strong non-Gaussianity. For a local signal y i , the corresponding side harmonic impedance can be calculated by equation (8), equation (9) and equation (12), and then the corresponding global signal can be reconstructed by equation (15).

[0146]

[0147] Where: and is the impedance on both sides obtained from the local signal.

[0148] by Take this as an example to illustrate, if is highly consistent with the corresponding real source signal, then and All have strong non-Gaussian properties; otherwise and The non-Gaussianity of all signals is weak, or only one of the signals is strongly non-Gaussian, while the other signal is weakly non-Gaussian. Non-Gaussianity is measured by the negative entropy shown in formula (4). In formula (4), the maximum value of J(s) is generally obtained at the maximum value of E{G(s)}, so E{G(s)} can be used to approximate the negative entropy between the signals. The larger the value, the greater the negative entropy and the stronger the non-Gaussianity. Formula (16) is selected as the non-square nonlinear function G{.} in formula (4).

[0149] G(s)=-exp(-s 2 / 2) (16)

[0151] In calculation and Before the negative entropy is obtained, it is necessary to pre-process it. and The signal s is preprocessed as follows:

[0152]

[0153] Where: It means to find the mean of signal s. std(.) means to find the standard deviation of signal s.

[0154] Let the negative entropy of the real and imaginary parts of the global signal corresponding to the i-th group of local signals after sparse screening be J x,i With J y,i , J can be expressed as x,i Normalized to J x,i .J y,i The normalization method is similar.

[0155] J x,i =(J x,i -J x,min ) / (J x,max -J x,min ) (18)

[0157] Where: J x,min With J x,max They are respectively the minimum and maximum values ​​of the real part negative entropy of the global signal corresponding to each group of local signals after sparse screening.

[0158] by and The comprehensive negative entropy maximization is used as the screening target. The comprehensive negative entropy σ of the i-th group of local signals corresponding to the global signal gauss,i Measured by formula (19), the larger the value, the stronger the comprehensive non-Gaussianity of the real and imaginary parts.

[0159] σ gauss,i =(J x,i+J y,i ) / 2 (19)

[0161] After sparse screening, the local signal that is highly consistent with the real source signal can be found through comprehensive negative entropy screening, and under-optimization and over-optimization can be avoided. The harmonic impedance of the user side and the system side of the global signal can then be obtained, and the harmonic emission level can be evaluated through the harmonic impedance determined above.

[0162] Embodiment 6

[0163] Figure 7 1 is a schematic diagram of the structure of a harmonic evaluation device for a flexible DC converter station according to Embodiment 6 of the present invention. Figure 7 As shown, the device comprises:

[0164] The approximate local signal module 510 is used to determine a harmonic source approximate signal according to the acquired harmonic data, and perform a local search in the harmonic data based on the harmonic source approximate signal to acquire a local signal.

[0165] The signal sparse processing module 520 is used to construct a sparse dictionary according to the local signal, and apply the sparse dictionary to the harmonic data.

[0166] The global signal construction module 530 is used to filter the processed harmonic data according to a preset sparse screening criterion to determine a target local signal in the local signal, and construct a target global signal according to the target local signal.

[0167] The source signal evaluation module 540 is used to select the target global signal with the strongest non-Gaussianity as the real harmonic source signal, and generate harmonic transmission level evaluation information according to the real harmonic source signal.

[0168] In the embodiment of the present invention, harmonic data is measured by an approximate local signal module, and a harmonic source approximate signal is determined according to the harmonic data. A local search is performed based on the harmonic source approximate signal to obtain a local signal. A signal sparse processing module generates a sparse dictionary corresponding to the local signal, and the sparse dictionary is applied to the harmonic data. The processed harmonic data is screened by a preset sparse criterion to determine a target local signal in each local signal. A global signal construction module constructs a target global signal corresponding to each target local signal. A source signal evaluation module selects a target global signal with the strongest non-Gaussianity as a true harmonic source signal, and generates harmonic transmission level evaluation information according to the true harmonic source signal, thereby realizing the evaluation of the harmonic emission level of a flexible DC converter station in a high-cable scenario, reducing the impact of an increase in system-side impedance on the harmonic emission level evaluation, thereby improving the accuracy of the harmonic emission level evaluation, helping to alleviate the harmonic generation of transmission lines, and improving the transmission efficiency of flexible mainstream converter stations.

[0169] Based on the above embodiments of the invention, the approximate local signal module 510 includes:

[0170] The decoupling processing unit is used to decouple the real and imaginary parts of the harmonic voltage, harmonic current, system-side equivalent harmonic source and user-side equivalent harmonic source in the harmonic data to construct a real and imaginary part decoupling matrix.

[0171] A matrix extraction unit is used to construct a first matrix according to the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the harmonic imaginary part determined by the real and imaginary part decoupling matrix, and to construct a second matrix according to the real part of the system side equivalent harmonic source, the imaginary part of the system side equivalent harmonic source, the real part of the user side equivalent harmonic source and the imaginary part of the user side equivalent harmonic source determined by the real and imaginary part decoupling matrix.

[0172] A signal separation unit is used to determine the harmonic source approximation signal according to a preset FastICA algorithm based on a blind source separation model matrix determined by the first matrix and the second matrix.

[0173] The local search unit is used to search and obtain the local signal within a preset local signal length range of the harmonic source approximation signal of the harmonic data.

[0174] Based on the above-mentioned embodiment of the invention, the signal sparse processing module 520 includes:

[0175] A dictionary generating unit is used to generate the sparse dictionary of the local signal according to a preset dictionary learning algorithm and the sparsity of the local signal.

[0176] A sparse processing unit is used to transform a first matrix corresponding to the harmonic voltage and the harmonic current in the harmonic data according to the sparse dictionary.

[0177] Based on the above-mentioned embodiment of the invention, the global signal construction module 530 includes:

[0178] A data extraction unit is used to determine a third matrix in the processed harmonic data, wherein the third matrix is ​​composed of the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the imaginary part of the harmonic current processed by the sparse dictionary.

[0179] A matrix determination unit is used to determine, when the third matrix is ​​a sparse matrix, that the harmonic data passing through the sparse dictionary meets the preset sparse screening criterion, and then use the local signal corresponding to the current sparse dictionary as the target local signal.

[0180] Based on the above-mentioned embodiment of the invention, the global signal construction module 530 also includes: a signal construction unit, used to obtain the third matrix corresponding to the target local signal, and determine the clustering straight line slope of the third matrix; generate the target global signal according to the clustering straight line slope and the target local signal.

[0181] Based on the above-mentioned embodiment of the invention, the source signal evaluation module 540 includes:

[0182] The preprocessing unit is used to preprocess the target global signal so that the target global signal has zero mean and unit variance.

[0183] The negative entropy determination unit is used to determine the negative entropy of the target global signal according to a preset negative entropy approximate calculation formula, and extract the real maximum negative entropy, the real minimum negative entropy, the imaginary maximum negative entropy and the imaginary minimum negative entropy from the negative entropy.

[0184] The normalization unit is used to normalize the negative entropy for each target global signal according to the respective maximum negative entropy of the real part, the minimum negative entropy of the real part, the maximum negative entropy of the imaginary part and the minimum negative entropy of the imaginary part.

[0185] The total negative entropy unit is used to determine the comprehensive negative entropy value corresponding to each of the negative entropies that have undergone the normalization processing, and to find the maximum comprehensive negative entropy value among the comprehensive negative entropy values.

[0186] An evaluation execution unit is used to use the target global signal of the maximum comprehensive negative entropy value as the real harmonic source signal, and determine the harmonic impedance of the target global signal as the harmonic transmission level evaluation information.

[0187] Based on the above-mentioned embodiment of the invention, a negative entropy approximate calculation formula is preset in the device, including:

[0188] J(s)≈[E{G(s)}-E{G(s Gauss)}] 2

[0189] Where s represents the target global signal, G(s Gauss ) represents a Gaussian random signal, E{} represents the expectation, G(s) = -exp (-s 2 / 2), and exp represents a natural exponential function.

[0190] The flexible DC converter station harmonic assessment device provided in the embodiment of the present invention can execute the flexible DC converter station harmonic assessment method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0191] Embodiment 7

[0192] Figure 8 : is a schematic diagram of the structure of an electronic device for implementing the method for harmonic evaluation of a flexible DC converter station according to an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein. Figure 8 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0193] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0194] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the flexible DC converter station harmonic evaluation method.

[0195] In some embodiments, the flexible DC converter station harmonic evaluation method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the flexible DC converter station harmonic evaluation method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the flexible DC converter station harmonic evaluation method in any other appropriate manner (e.g., by means of firmware).

[0196] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0197] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0198] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0199] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0200] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0201] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0202] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0203] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating harmonics in a flexible DC converter station, characterized in that: The method comprises: Determine a harmonic source approximate signal according to the acquired harmonic data, and perform a local search within the harmonic data based on the harmonic source approximate signal to acquire a local signal; Constructing a sparse dictionary according to the local signal, and applying the sparse dictionary to the harmonic data; Screening the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal, and constructing a target global signal according to the target local signal; Selecting the target global signal with the strongest non-Gaussianity as a true harmonic source signal, and generating harmonic transmission level assessment information according to the true harmonic source signal; The step of determining a harmonic source approximate signal according to the acquired harmonic data, and performing a local search in the harmonic data based on the harmonic source approximate signal to acquire a local signal comprises: Decoupling the real and imaginary parts of the harmonic voltage, harmonic current, system-side equivalent harmonic source, and user-side equivalent harmonic source in the harmonic data to construct a real and imaginary part decoupling matrix; A first matrix is ​​constructed according to the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the imaginary part of the harmonic determined by the real and imaginary part decoupling matrix, and a second matrix is ​​constructed according to the real part of the system side equivalent harmonic source, the imaginary part of the system side equivalent harmonic source, the real part of the user side equivalent harmonic source and the imaginary part of the user side equivalent harmonic source determined by the real and imaginary part decoupling matrix; Based on the blind source separation model matrix determined by the first matrix and the second matrix, determining the harmonic source approximation signal according to a preset FastICA algorithm; Searching and acquiring the local signal within a preset local signal length range of the harmonic source approximation signal of the harmonic data; The step of selecting the target global signal with the strongest non-Gaussian property as the real harmonic source signal and generating harmonic transmission level assessment information according to the real harmonic source signal comprises: Preprocessing the target global signal so that the target global signal has zero mean and unit variance; Determine the negative entropy of the target global signal according to a preset negative entropy approximate calculation formula, and extract the maximum negative entropy of the real part, the minimum negative entropy of the real part, the maximum negative entropy of the imaginary part, and the minimum negative entropy of the imaginary part from the negative entropy; For each of the target global signals, normalizing the negative entropy according to the respective maximum negative entropy of the real part, the minimum negative entropy of the real part, the maximum negative entropy of the imaginary part, and the minimum negative entropy of the imaginary part; Determine the comprehensive negative entropy value corresponding to each of the negative entropies subjected to the normalization process, and find the maximum comprehensive negative entropy value among the comprehensive negative entropy values; The target global signal of the maximum comprehensive negative entropy value is used as the real harmonic source signal, and the harmonic impedance of the target global signal is determined as the harmonic transmission level evaluation information.

2. The method according to claim 1, characterized in that: The step of constructing a sparse dictionary according to the local signal and applying the sparse dictionary to the harmonic data includes: Generating the sparse dictionary of the local signal according to a preset dictionary learning algorithm and the sparsity of the local signal; The first matrix corresponding to the harmonic voltage and the harmonic current in the harmonic data is transformed according to the sparse dictionary.

3. The method according to claim 1, characterized in that: The step of screening the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal comprises: Determine a third matrix in the processed harmonic data, wherein the third matrix is ​​composed of the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current, and the imaginary part of the harmonic current processed by the sparse dictionary; When the third matrix is ​​a sparse matrix, it is determined that the harmonic data passing through the sparse dictionary meets the preset sparse screening criterion, and the local signal corresponding to the current sparse dictionary is used as the target local signal.

4. The method according to claim 3, characterized in that: The step of constructing a target global signal according to the target local signal comprises: Acquire the third matrix corresponding to the target local signal, and determine the clustering straight line slope of the third matrix; The target global signal is generated according to the clustering straight line slope and the target local signal.

5. The method according to claim 1, characterized in that: The preset negative entropy approximate calculation formula includes: J(s)≈[E{G(s)}-E{G(s Gauss )}] 2 Where s represents the target global signal, G(s Gauss ) represents a Gaussian random signal, E{} represents the expectation, G(s) = -exp(-s 2 / 2), and exp represents a natural exponential function.

6. A harmonic evaluation device for a flexible DC converter station, characterized in that: The device comprises: An approximate local signal module, used to determine a harmonic source approximate signal according to the acquired harmonic data, and perform a local search within the harmonic data based on the harmonic source approximate signal to acquire a local signal; A signal sparse processing module, used for constructing a sparse dictionary according to the local signal, and applying the sparse dictionary to the harmonic data; A global signal construction module, configured to screen the processed harmonic data according to a preset sparse screening criterion to determine a target local signal within the local signal, and construct a target global signal according to the target local signal; A source signal evaluation module, used for selecting the target global signal with the strongest non-Gaussianity as a true harmonic source signal, and generating harmonic transmission level evaluation information according to the true harmonic source signal; The approximate local signal module includes: A decoupling processing unit, used for decoupling the real and imaginary parts of the harmonic voltage, harmonic current, system-side equivalent harmonic source, and user-side equivalent harmonic source in the harmonic data to construct a real and imaginary part decoupling matrix; A matrix extraction unit, configured to construct a first matrix according to the real part of the harmonic voltage, the imaginary part of the harmonic voltage, the real part of the harmonic current and the imaginary part of the harmonic determined by the real and imaginary part decoupling matrix, and to construct a second matrix according to the real part of the system side equivalent harmonic source, the imaginary part of the system side equivalent harmonic source, the real part of the user side equivalent harmonic source and the imaginary part of the user side equivalent harmonic source determined by the real and imaginary part decoupling matrix; A signal separation unit, configured to determine the harmonic source approximation signal according to a preset FastICA algorithm based on a blind source separation model matrix determined by the first matrix and the second matrix; A local search unit, configured to search and obtain the local signal within a preset local signal length range of the harmonic source approximation signal of the harmonic data; The source signal evaluation module includes: A preprocessing unit, configured to preprocess the target global signal so that the target global signal has a zero mean and a unit variance; A negative entropy determination unit, used to determine the negative entropy of the target global signal according to a preset negative entropy approximate calculation formula, and extract the maximum negative entropy of the real part, the minimum negative entropy of the real part, the maximum negative entropy of the imaginary part, and the minimum negative entropy of the imaginary part from the negative entropy; A normalization unit, for normalizing the negative entropy according to the respective real maximum negative entropy, the real minimum negative entropy, the imaginary maximum negative entropy and the imaginary minimum negative entropy for each target global signal; A total negative entropy unit, used to determine the comprehensive negative entropy value corresponding to each of the negative entropies subjected to the normalization process, and to find the maximum comprehensive negative entropy value among the comprehensive negative entropy values; An evaluation execution unit is used to use the target global signal of the maximum comprehensive negative entropy value as the real harmonic source signal, and determine the harmonic impedance of the target global signal as the harmonic transmission level evaluation information.

7. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the flexible DC converter station harmonic evaluation method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the flexible DC converter station harmonic evaluation method according to any one of claims 1 to 5 when executed.

Images