Self-controllable gateway power meter interharmonic adaptive cancellation frequency estimation method

Abstract

The application discloses a self-adaptive frequency estimation method for inter-harmonic of self-controllable gateway electric energy meter, and belongs to the technical field of electric variable measurement. The application continuously samples a voltage signal, and the sampling frequency is three times of an integer multiple of a fundamental frequency; after band-pass filtering, the voltage signal is continuously grouped and sampled at intervals of one-third of a fundamental period; intermediate variables are calculated for each group of sampling data; then, a frequency domain compression factor reflecting only the fundamental frequency is directly extracted by using the mechanism that 1.5 times of inter-harmonic components are cancelled in factor operation; and finally, a convergence terminal value is obtained through adaptive iterative estimation, and the frequency is obtained through inverse cosine operation. The application integrates inter-harmonic cancellation and frequency information generation into the same link, breaks through the response speed bottleneck of the series structure of pre-filtering and phase-locked loop, and realizes fast and accurate estimation of the fundamental frequency under the interference of 1.5 times of inter-harmonic and dynamic load.

Description

Technical Field

[0001] This invention belongs to the field of measuring electrical variables, and specifically relates to a frequency estimation method. Background Technology

[0002] Accurate frequency measurement is a fundamental parameter for the stable operation of power systems and a core prerequisite for high-precision energy metering at key power outlets. In modern electricity market trading and settlement systems, the metering accuracy of key power outlets directly affects the fairness and economic benefits of transactions, and they must rely on real-time, accurate frequency information to complete error compensation, power calculation, and synchronous sampling control. However, with the widespread integration of nonlinear and fluctuating loads such as renewable energy grid-connected equipment and frequency converters into new power systems, complex interharmonic interference frequently appears in grid voltage and current signals. These components with frequencies other than integer multiples of the fundamental frequency severely impair the accuracy of frequency measurement, leading to inaccurate energy metering and affecting the reliability and fairness of the electricity market.

[0003] This problem is particularly prominent in real-world industrial scenarios. For example, in the steelmaking process of the metallurgical industry, the drastic nonlinear changes in the load of the electric arc furnace, and the operation of the cyclic converter in rail transit, both inject significant 1.5th interharmonic waves (75Hz) into the power grid. This frequency component lies precisely between the fundamental frequency (50Hz) and the second harmonic (100Hz), belonging to non-integer harmonics. Due to the physical characteristics of the transition band and the phase distortion introduced thereafter, conventional digital filters cannot completely filter out this non-integer frequency component while fully preserving the fundamental frequency information. As a result, the 1.5th interharmonic wave is easily aliased in the sampled signal, causing continuous and severe interference to frequency estimation algorithms such as zero-crossing detection and Fourier transform, making it difficult for the frequency-locked loop to converge. The newly released national standard "GB / T 19862-2025 General Requirements for Power Quality Monitoring Equipment" and power grid enterprise standards have both clearly listed the 1.5th interharmonic wave as a key power quality monitoring indicator, and put forward strict test and evaluation requirements for its measurement accuracy, spectral resolution, and immunity. This signifies that the anti-interference capability and measurement performance of existing electricity meters and monitoring equipment in this special frequency band have become important technical indicators for measuring whether they meet the needs of high-end metering.

[0004] To suppress harmonic and interharmonic components in grid voltage, various pre-filtering methods have been proposed by those skilled in the art. Among them, the time-delay signal cancellation technique utilizes the periodic characteristics of electrical quantities, constructing a cancellation signal by delaying half of the harmonic period to eliminate the target harmonics. This method can be applied to both single-phase and three-phase systems. However, the ability of a single time-delay operator to suppress harmonics is limited, and multiple time-delay operators need to be cascaded in practical applications. In their article "A Fast Time-Delay Signal Cancellation Filtering Method Based on the Vector Triangle Principle," published in Volume 41 of the *Proceedings of the Chinese Society for Electrical Engineering* in 2021, Li Xinian, Xu Tao, and Niu Decun proposed a fast time-delay signal cancellation filtering method to eliminate the main 5th, 7th, 11th, and 13th harmonics in grid voltage. By sampling three times at equal intervals and using the vector sum of two sampled values ​​to cancel the harmonic components in the third sampled value, the filtering delay for eliminating the same set of harmonics is shortened to a quarter of the fundamental frequency period, increasing the degree of freedom in parameter configuration. However, this method still follows the serial approach of "pre-filter + phase-locked loop" in its system architecture. That is, the filter first performs signal cancellation independently, and then the subsequent stages perform frequency estimation. The filtering process itself does not generate frequency information, and its inherent delay still constitutes a bottleneck for the system's fast response, failing to fundamentally overcome the speed limitations imposed by the traditional architecture. Therefore, a new method is urgently needed that can achieve fast and accurate frequency estimation under 1.5th interharmonic interference. Summary of the Invention

[0005] This invention proposes an autonomous and controllable adaptive cancellation frequency estimation method for interharmonics in gated energy meters. Its purpose is to overcome the bottleneck of the existing "pre-filter + phase-locked loop" series architecture, which does not generate frequency information and has inherent delay that restricts response speed, in the case of 1.5th interharmonic interference in the grid voltage, and to achieve direct, fast and accurate estimation of the fundamental frequency.

[0006] The technical solution of this invention is as follows:

[0007] An autonomous and controllable adaptive destructive frequency estimation method for inter-meter harmonics includes:

[0008] Step 1: Continuously sample the voltage sampling frequency to obtain a voltage signal sequence;

[0009] Step 2: Pass the voltage signal sequence obtained in Step 1 through a bandpass filter to filter out high-order harmonic signal interference;

[0010] In step 1, the sampling frequency is an integer multiple of three times the fundamental frequency of the sampled signal;

[0011] This method also includes:

[0012] Step 3: Calculate the sampling interval based on the sampling frequency determined in Step 1, and perform interval sampling on the voltage signal sequence after filtering in Step 2 to obtain multiple sets of continuous sampling data, each containing four interval sampling points.

[0013] Step 4: Calculate the front window mean factor, back window mean factor, and center summation factor for each group of sampled data, and construct a relationship model between the front window mean factor, back window mean factor, center summation factor, and frequency domain compression factor.

[0014] Step 5: Based on the front window mean factor, back window mean factor, and center summation factor of multiple sets of sampled data, as well as the constructed relational model, adaptively estimate the frequency domain compression factor to obtain the convergent final value;

[0015] Step 6: Convergence final value of the frequency domain compression factor Perform inverse cosine function calculations to obtain an estimated frequency value.

[0016] As a further improvement to the autonomous and controllable gate energy meter interharmonic adaptive cancellation frequency estimation method, the sampling frequency is more than 10 times the theoretical value of the fundamental frequency.

[0017] As a further improvement to the autonomous and controllable gate energy meter inter-harmonic adaptive destructive frequency estimation method, in step 3, the sampling interval is:

[0018]

[0019] In the above formula, Sampling frequency, This is the theoretical value of the fundamental frequency of the voltage signal.

[0020] As a further improvement to the autonomous and controllable gate energy meter inter-harmonic adaptive destructive frequency estimation method, in step 3, the interval sampling method is as follows: according to the sampling interval... Four sampling points are sequentially extracted from the filtered voltage signal sequence and grouped into groups of four, as shown below:

[0021]

[0022]

[0023]

[0024]

[0025]

[0026] This represents the first voltage signal in the filtered voltage signal sequence. One sampling point.

[0027] As a further improvement to the autonomous and controllable gate energy meter inter-harmonic adaptive cancellation frequency estimation method, after sampling at intervals in step 3, the validity of each set of sampled data should be judged and invalid sampled data should be removed.

[0028] As a further improvement to the autonomous and controllable gate energy meter interharmonic adaptive cancellation frequency estimation method, the specific method for judging its effectiveness is as follows: any set of sampled data is denoted as... The subscript represents the relative time order of the point within the group; if If the sampled data is invalid, then the sampled data in this group is considered invalid; among them, This indicates the absolute value operation. Indicates the threshold.

[0029] As a further improvement to the autonomous and controllable gate energy meter inter-harmonic adaptive destructive frequency estimation method: In step 4, for any set of sampled data, it is represented as Then the front window mean factor Back window mean factor summation factor They are respectively:

[0030] ;

[0031] ;

[0032] .

[0033] As a further improvement to the autonomous and controllable inter-meter harmonic adaptive destructive frequency estimation method: In step 4, the relationship model between the front window mean factor, the back window mean factor, the center summation factor, and the frequency domain compression factor is as follows: , This is the calculated value of the frequency domain compression factor corresponding to this set of sampled data.

[0034] As a further improvement to the autonomous and controllable gate energy meter interharmonic adaptive cancellation frequency estimation method, step 5 uses the steepest gradient principle for adaptive estimation, specifically including:

[0035] Step 5-1: Set the initial estimate of the frequency domain compression factor. Number of iterations ; Indicates the first The estimated value of the frequency domain compression factor corresponding to the group of sampled data. , , They respectively represent based on the first The front window mean factor, back window mean factor, and central summation factor are calculated from the sampled data.

[0036] Step 5-2: Calculate the first... Estimation error corresponding to group sampling data :

[0037] ;

[0038] Step 5-3: Based on the steepest gradient principle, calculate the gradient value for this round using the following formula:

[0039]

[0040] in, These are preset parameters;

[0041] Step 5-4: Based on the result obtained in Step 5-3... gradient value of the first step Calculate the first The estimated value of the next step:

[0042]

[0043] Step 5-5: If the standard deviation of the estimated value is less than the preset threshold for a series of preset times, then determine that the estimated value obtained at this time is... Let be the convergent final value of the frequency domain compression factor; otherwise, let Then proceed to step 5-2.

[0044] As a further improvement to the autonomous and controllable gate energy meter inter-harmonic adaptive destructive frequency estimation method, the estimated frequency value is calculated in step 6 as follows:

[0045]

[0046] In the above formula, This is the convergent final value of the frequency domain compression factor; The time corresponding to the sampling interval. denoted as the fundamental frequency of the signal.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. This invention simultaneously solves two key technical problems: the speed of frequency estimation and the ability to resist interharmonic interference. Based on the setting in step 1 that the sampling frequency is an integer multiple of three times the fundamental frequency, the sampling interval determined in step 3 ensures that the time span between four consecutive sampling points is exactly one-third of the fundamental period. Under this specific sampling interval, the phase difference of the 1.5th interharmonic component between adjacent sampling points is exactly... This component is canceled to zero in the summation operations of the front window mean factor, the back window mean factor, and the center summation factor, while the fundamental component is completely preserved and compressed into the frequency domain compression factor. Therefore, the relational model constructed in step 4 can directly extract the frequency domain compression factor reflecting only the fundamental frequency from the original sampled data containing the 1.5th interharmonic, integrating interharmonic cancellation and frequency information generation into the same computational stage. This fundamentally overcomes the bottleneck of the existing "pre-filter + phase-locked loop" series architecture, where the filtering stage does not generate frequency information and its inherent delay restricts response speed.

[0049] 2. Thanks to the structural design in step 3 where each set of sampled data independently constitutes a complete frequency measurement unit, a complete frequency estimation can be performed on each set of data. When the power grid frequency changes abruptly, from the moment of change, only a maximum of two-thirds of the fundamental frequency cycle length of new sampling points is needed to complete an effective estimation with a new set of data, without waiting for the accumulation of a whole cycle or a longer period of data, thus significantly shortening the response delay of frequency estimation.

[0050] 3. This invention, through the steepest gradient adaptive estimation mechanism in step 5, enables the estimation process of the frequency domain compression factor to possess continuous dynamic correction capabilities. When load fluctuations or grid frequency offsets cause changes in signal characteristics, each set of sampled data independently calculates the estimation error, which drives the gradient-direction-based update, ensuring that the estimated value continuously approaches the true value along the error descent direction without the need for re-acquisition or initiation of the phase-locked loop process. Step 5 also uses the standard deviation of multiple consecutive estimates as a convergence criterion to quantitatively determine the stability of the estimation results, ensuring that the output convergence final value in dynamic scenarios possesses both accuracy and stability, thus enabling the method to maintain high estimation accuracy even under dynamic load conditions.

[0051] 4. To further ensure the reliability of the estimation results, this invention also introduces a validity judgment mechanism after constructing the grouped sampling data, utilizing... The condition eliminates invalid data sets caused by transient disturbances or sampling anomalies, preventing them from participating in the subsequent adaptive estimation process. This judgment condition directly corresponds to the physical constraint that the central summation factor should not be zero under the interharmonic destructive structure, which not only improves the robustness of the estimation method in complex electromagnetic environments, but also prevents invalid data from disturbing the gradient iteration convergence process. Detailed Implementation

[0052] The technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0053] An autonomous and controllable adaptive destructive frequency estimation method for inter-meter harmonics includes:

[0054] Step 1: Continuously sample the voltage sampling frequency to obtain the voltage signal sequence and save it to the buffer. The sampling frequency is an integer multiple of three times the fundamental frequency of the sampled signal.

[0055] Under the premise of satisfying the Nyquist sampling theorem, the sampling frequency Should meet:

[0056]

[0057] in, It is an integer. This is the theoretical value of the fundamental frequency of the voltage signal, for example, 50Hz. In modern sampling devices, the sampling frequency is mostly adjustable, therefore for The constraints are relatively easy to satisfy. Only by satisfying these constraints can complete 75Hz interharmonic cancellation be achieved.

[0058] Based on experience, the sampling frequency is preferably more than 10 times the fundamental frequency. For a 50Hz power frequency signal, a sampling frequency of 6kHz or 12kHz can be considered.

[0059] Step 2: Pass the voltage signal sequence obtained in Step 1 through a bandpass filter to filter out high-order harmonic signal interference.

[0060] In practical engineering applications, the bandpass filter used in this step is generally an IIR digital filter. Due to computational limitations, the order of the digital filter is typically 6th to 8th. In this case, the 75Hz interharmonic is located in the transition band of the digital filter, and the attenuation of the 75Hz component is generally -10 to -20dB, which is not negligible. The presence of the 75Hz component will seriously interfere with the accuracy of frequency estimation methods such as zero-crossing estimation and PLL.

[0061] In summary, although a bandpass filter was used in step 2, it could only attenuate the harmonic content. For the 75Hz content, only a portion could be attenuated, and the remainder would continue to be passed on to subsequent stages.

[0062] Step 3: Calculate the sampling interval based on the sampling frequency determined in Step 1, and perform interval sampling on the voltage signal sequence after filtering in Step 2 to obtain multiple sets of continuous sampling data, each containing four interval sampling points.

[0063] Step 3-1: Calculate the sampling interval using the following formula:

[0064]

[0065] Step 3-2: Take four sampling points from the filtered voltage signal sequence according to the above sampling interval, and group them into groups of four, as shown below:

[0066]

[0067]

[0068]

[0069]

[0070]

[0071] This represents the first voltage signal in the filtered voltage signal sequence. One sampling point.

[0072] Step 3-3: Determine the validity of each set of sampled data and remove invalid sampled data.

[0073] To facilitate formula derivation, any set of sampled data is used. Unified notation The subscript represents the relative time order of that point within the group. If If the sampled data does not meet the condition, then the sampled data in this group is considered invalid. This indicates the absolute value operation. Indicates the threshold.

[0074] Because subsequent steps require similar differentiation operations. Located at the molecule position, its calculated value cannot be 0. Considering that most practical calculations are floating-point operations, and floating-point subtraction generally cannot be strictly equal to 0, the final threshold is... .

[0075] Step 4: Calculate the front window mean factor, back window mean factor, and center summation factor for each group of sampled data, and construct a relationship model between the front window mean factor, back window mean factor, center summation factor, and frequency domain compression factor.

[0076] For any set of sampled data, it is represented as Then the front window mean factor Back window mean factor summation factor They are respectively:

[0077]

[0078]

[0079]

[0080] The relationship model between the front window mean factor, the back window mean factor, the center summation factor, and the frequency domain compression factor is as follows: , This is the calculated value of the frequency domain compression factor corresponding to this set of sampled data.

[0081] The following explains why the above formula yields the desired result when both the fundamental frequency (50Hz) and the 1.5th interharmonic frequency (75Hz) are present simultaneously. That is, the frequency domain compression factor equivalent to the fundamental frequency. .

[0082] (1) Relationship under single-frequency signal

[0083] First, consider the ideal case where the signal contains only a single frequency component. Let the signal angular frequency be... The amplitude is The initial phase is Then the time-domain expression for four equally spaced sampling points in any set of sampled data is:

[0084]

[0085] in The time corresponding to the sampling interval. This is the earliest sampling point in the group. This is the latest sampling point.

[0086] For the four sampling points mentioned above, calculate the front window mean factor according to the definition in step 4. Back window mean factor summation factor :

[0087]

[0088] First calculate and :

[0089] right Apply the sum-to-product formula :

[0090]

[0091]

[0092] Similarly, for Application of sum-to-product:

[0093]

[0094]

[0095] Then calculate After extracting the common factor, apply the sum-to-product formula again:

[0096]

[0097]

[0098]

[0099] Recalculate ,right Application of sum-to-product:

[0100]

[0101] Comparison yields:

[0102]

[0103] Right now:

[0104]

[0105] make This is called the frequency domain compression factor. It can be seen that, for a single-frequency signal, the signal frequency can be directly calculated from the sampled values ​​using this formula.

[0106] (2) Relationship under dual-frequency signal and interharmonic cancellation principle

[0107] In actual operating conditions, the voltage signal simultaneously contains the fundamental frequency (angular frequency). ) and 1.5th interharmonic (angular frequency) Let their amplitudes be respectively. , The initial phases are respectively , Then the time-domain expression for the four sampling points is:

[0108]

[0109] Four sampling points , , , Both are composed of the superposition of the fundamental component and the 1.5th interharmonic component, that is:

[0110]

[0111]

[0112]

[0113]

[0114] The time-domain expression for each component is as follows:

[0115] Fundamental component:

[0116]

[0117]

[0118]

[0119]

[0120] 1.5th interharmonic component:

[0121]

[0122]

[0123]

[0124]

[0125] Based on intermediate variables , , The definitions are as follows: they are composed of the corresponding fundamental component and the 1.5th interharmonic component respectively:

[0126] Intermediate variables corresponding to the fundamental component:

[0127]

[0128]

[0129]

[0130] Intermediate variables corresponding to the 1.5th interharmonic component:

[0131]

[0132]

[0133]

[0134] Due to the property of linear superposition, the intermediate variable in a dual-frequency signal is equal to the sum of the intermediate variables of each component:

[0135]

[0136]

[0137]

[0138] In this context, the superscript 1 indicates the intermediate variable corresponding to the fundamental component, and the superscript 1.5 indicates the intermediate variable corresponding to the 1.5th interharmonic component.

[0139] From the conclusions drawn from single-frequency signals, we know that the intermediate variables of the fundamental component satisfy the following:

[0140]

[0141] The intermediate variables of the 1.5th interharmonic component also satisfy:

[0142]

[0143] Therefore, under dual-frequency signals, direct calculation is possible. At that time, what was actually obtained was:

[0144]

[0145] Obviously, in general, It is not possible to directly apply the relationship under single frequency to obtain .

[0146] However, the key to this invention lies in the special design of the sampling interval. This is due to the constraints imposed by the sampling frequency. It can be seen that the sampling interval The corresponding time is:

[0147]

[0148] Therefore, for the fundamental frequency: .

[0149] For the 1.5th interharmonic, angular frequency ,therefore:

[0150]

[0151] because The phase difference of the 1.5th interharmonic component between adjacent sampling points is exactly half a period, according to ,have:

[0152]

[0153] Substituting into the definition of intermediate variables, we find that all intermediate variables corresponding to the 1.5th interharmonic component are zero:

[0154]

[0155] therefore:

[0156]

[0157] Right now:

[0158]

[0159] This demonstrates that, under dual-frequency signal conditions where the fundamental frequency and the 1.5th interharmonic coexist, a specially designed sampling interval time can achieve the desired signal strength. The 1.5th interharmonic component is adaptively canceled during the calculation of intermediate variables. The original sampled data containing dual-frequency components is used directly, according to... The calculation yielded a result that was completely equivalent to the frequency domain compression factor when only the fundamental wave was present, thus achieving complete suppression of the 1.5th interharmonic interference.

[0160] Step 5: Based on the front window mean factor, back window mean factor, and center summation factor of multiple sets of sampled data, as well as the constructed relational model, the frequency domain compression factor is adaptively estimated to obtain the convergent final value.

[0161] In this embodiment, the steepest gradient principle is used for adaptive estimation.

[0162] Step 5-1: Let the initial estimate be... Number of iterations ; Indicates the first The estimated value of the frequency domain compression factor corresponding to the group of sampled data. , , They respectively represent based on the first The front window mean factor, back window mean factor, and central summation factor are calculated from the sampled data.

[0163] Step 5-2: Calculate the first... Estimation error corresponding to group sampling data :

[0164]

[0165] Step 5-3: Based on the steepest gradient principle, calculate the gradient value for this round using the following formula:

[0166]

[0167] in, For experience values, excessively large values ​​are not advisable. The value may lead to overshooting or undershooting of the estimated value. A low value may lead to slow convergence of the estimated value. For a fundamental frequency of 50Hz... The value can be selected within the range of 0.5-2.

[0168] Step 5-4: Based on the result obtained in Step 5-3... gradient value of the first step Calculate the first The estimated value of the next step:

[0169]

[0170] Step 5-5: If the standard deviation of the estimated value is less than the preset threshold for several consecutive times (e.g., 10 times), then the estimated value is considered valid. Let be the convergent final value of the frequency domain compression factor; otherwise, let Then proceed to step 5-2.

[0171] Step 6: Convergence final value of the frequency domain compression factor By performing an inverse cosine function operation, the estimated frequency is obtained:

[0172]

[0173] In the above formula, The time corresponding to the sampling interval. denoted as the fundamental frequency of the signal.

[0174] Furthermore, the frequency estimates obtained in step 6 There may be slight fluctuations. You can consider averaging the multiple frequency estimates obtained consecutively to filter out the AC components and use the result as the final estimate.

[0175] It should be noted that, as will be apparent to those skilled in the art, the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics thereof. The scope of the present invention is defined by the claims rather than the foregoing description.

Claims

1. A self-controllable adaptive destructive frequency estimation method for inter-meter harmonics, comprising: Step 1: Continuously sample the voltage sampling frequency to obtain a voltage signal sequence; Step 2: Pass the voltage signal sequence obtained in Step 1 through a bandpass filter to filter out high-order harmonic signal interference; Its features are: In step 1, the sampling frequency is an integer multiple of three times the fundamental frequency of the sampled signal; This method also includes: Step 3: Calculate the sampling interval based on the sampling frequency determined in Step 1, and perform interval sampling on the voltage signal sequence after filtering in Step 2 to obtain multiple sets of continuous sampling data, each containing four interval sampling points. Step 4: Calculate the front window mean factor, back window mean factor, and center summation factor for each group of sampled data, and construct a relationship model between the front window mean factor, back window mean factor, center summation factor, and frequency domain compression factor. Step 5: Based on the front window mean factor, back window mean factor, and center summation factor of multiple sets of sampled data, as well as the constructed relational model, adaptively estimate the frequency domain compression factor to obtain the convergent final value; Step 6: Convergence final value of the frequency domain compression factor Perform inverse cosine function calculations to obtain an estimated frequency value.

2. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that: The sampling frequency is more than 10 times the theoretical value of the fundamental frequency.

3. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that, In step 3, the sampling interval is: ； In the above formula, Sampling frequency, This is the theoretical value of the fundamental frequency of the voltage signal.

4. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that, In step 3, the interval sampling method is as follows: according to the sampling interval Four sampling points are sequentially extracted from the filtered voltage signal sequence and grouped into sets of four, as follows: ； ； ； ； ； This represents the first voltage signal in the filtered voltage signal sequence. One sampling point.

5. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that, After sampling at intervals in step 3, the validity of each set of sampled data must be judged, and invalid sampled data must be removed.

6. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 5, characterized in that, The specific method for judging the validity is as follows: Any set of sampled data is denoted as... The subscript represents the relative time order of the point within the group; if If the sampled data is invalid, then the sampled data in this group is considered invalid; among them, This indicates the absolute value operation. Indicates the threshold.

7. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that: In step 4, for any set of sampled data, it is represented as Then the front window mean factor Back window mean factor summation factor They are respectively: ； ； 。 8. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 7, characterized in that: In step 4, the relationship model between the front window mean factor, the back window mean factor, the center summation factor, and the frequency domain compression factor is as follows: , This is the calculated value of the frequency domain compression factor corresponding to this set of sampled data.

9. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that, In step 5, the steepest gradient principle is used for adaptive estimation, specifically including: Step 5-1: Set the initial estimate of the frequency domain compression factor. Number of iterations ; Indicates the first The estimated value of the frequency domain compression factor corresponding to the group of sampled data. , , They respectively represent based on the first The front window mean factor, back window mean factor, and central summation factor are calculated from the sampled data. Step 5-2: Calculate the first... Estimation error corresponding to group sampling data : ； Step 5-3: Based on the steepest gradient principle, calculate the gradient value for this round using the following formula: ； in, These are preset parameters; Step 5-4: Based on the result obtained in Step 5-3... gradient value of the first step Calculate the first The estimated value of the next step: ； Step 5-5: If the standard deviation of the estimated value is less than the preset threshold for a series of preset times, then determine that the estimated value obtained at this time is... Let be the convergent final value of the frequency domain compression factor; otherwise, let Then proceed to step 5-2.

10. The autonomous and controllable inter-meter harmonic adaptive cancellation frequency estimation method as described in claim 1, characterized in that, The method for calculating the frequency estimate in step 6 is as follows: ； In the above formula, This is the convergent final value of the frequency domain compression factor; The time corresponding to the sampling interval. denoted as the fundamental frequency of the signal.

Images