CN112213561A - Measurement data preprocessing method and device for leading load parameter noise identification - Google Patents
Measurement data preprocessing method and device for leading load parameter noise identification Download PDFInfo
- Publication number
- CN112213561A CN112213561A CN202011025550.0A CN202011025550A CN112213561A CN 112213561 A CN112213561 A CN 112213561A CN 202011025550 A CN202011025550 A CN 202011025550A CN 112213561 A CN112213561 A CN 112213561A
- Authority
- CN
- China
- Prior art keywords
- time sequence
- measurement data
- pmu measurement
- value
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R25/00—Arrangements for measuring phase angle between a voltage and a current or between voltages or currents
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The embodiment of the invention provides a method and a device for preprocessing measurement data for leading load parameter noise identification, wherein the method comprises the following steps: collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, and if any time sequence in the i group of PMU measurement data is determined to pass serious error troubleshooting and serious abnormity troubleshooting, respectively removing low-frequency fluctuation trends of an active power time sequence and a reactive power time sequence injected into the i group of PMU measurement data by adopting empirical mode decomposition to obtain trend removing stable signals corresponding to any time sequence, wherein i is more than or equal to 1 and is less than or equal to K; and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence. The method and the device provided by the embodiment of the invention realize the acquisition of more accurate and reliable dominant load parameters and improve the identification availability of the data to be identified.
Description
Technical Field
The invention relates to the technical field of dominant load parameter noise identification, in particular to a measured data preprocessing method and device for dominant load parameter noise identification.
Background
The wide application of the synchronized phasor measurement technology (PMU) in the regional power grid provides a rich and accurate data source for the load modeling research of the power system. By using the measurement data of the noise-like voltage and the power noise-like collected at the high-voltage side of the transformer at the normal operation state of each station, the load leading parameters represented by the Z + M model can be obtained on line by a load model parameter identification method based on the noise-like. However, since the disturbance amplitude of the noise-like signal is small, the influence of the high-frequency noise and the abnormal value in the PMU measurement data on the model parameter identification becomes more prominent, which directly causes the model parameter optimization fitting effect to be poor or converges to an unreliable local optimal solution, and finally causes the actual requirement of online identification to be difficult to meet due to the low identification availability of the actually measured PMU data.
Therefore, how to avoid the actual requirement that the recognition availability of the actually measured PMU data is low and is difficult to meet online recognition, the recognition availability and the overall fitting effect of the data to be recognized are improved, and more accurate and reliable dominant load parameters are obtained, which is still a problem to be solved by those skilled in the art.
Disclosure of Invention
The embodiment of the invention provides a method for guaranteeing consistency of a telecommunication service contract based on a block chain, which is used for solving the problems that the existing telecommunication service contract based on the block chain cannot guarantee the verification activation capability of an offline business hall and cannot guarantee the online-offline consistency of the telecommunication service contract.
In a first aspect, an embodiment of the present invention provides a method for preprocessing measurement data for dominant load parameter noise identification, including:
collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, wherein any group of PMU measurement data comprises a time sequence with N/K total voltage amplitude time points, N/K total time sequences of active power injection time points and N/K total time sequences of reactive power injection time points on a main transformer high-voltage side of a transformer substation, and N and K are positive integers;
if any time sequence in the ith set of PMU measurement data is determined to pass serious error troubleshooting and serious abnormality troubleshooting, respectively removing low-frequency fluctuation trends of an active power time sequence and a reactive power time sequence which are injected into the ith set of PMU measurement data by adopting empirical mode decomposition to obtain trend-removed stable signals corresponding to any time sequence, wherein i is more than or equal to 1 and less than or equal to K;
and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
Preferably, the method further comprises:
and if the fact that the time sequence in the ith PMU measurement data does not pass the serious error investigation or the serious abnormal investigation is determined, rejecting the ith PMU measurement data.
Preferably, in the method, the performing a fatal error check on any time sequence in the ith set of PMU measurement data specifically includes:
and if any condition or combination of any condition that the average voltage of the voltage amplitude time sequence deviates from a preset range, the voltage value fluctuation of the voltage amplitude time sequence is smaller than a first threshold, a negative value appears in a power value injected into the active power time sequence, a negative value appears in a power value injected into the reactive power time sequence, the power value fluctuation of the power value injected into the active power time sequence is smaller than a second threshold, and the power value fluctuation of the power value injected into the reactive power time sequence is smaller than a third threshold is detected in the i-th set of PMU measurement data, it is determined that any time sequence in the i-th set of PMU measurement data passes serious error checking, and if not, the time sequence in the i-th set of PMU measurement data does.
Preferably, in the method, the performing an abnormal error check on any time sequence in the ith PMU measurement data includes:
determining the mean value of time series in the ith set of PMU measurement data
And standard deviation of
Wherein j is 1,2,3, which is used to identify a voltage amplitude time sequence, an injection active power time sequence and an injection reactive power time sequence, respectively;
for any time sequence in the ith set of PMU measurement data
Determining that the value of the sequence element is not
The elements in the range are outliers of the time series identified by j;
if the abnormal values in any time sequence are determined not to continuously have the fourth threshold value and more than one abnormal value, sequentially correcting each abnormal value method based on the previous average difference value and judging that any time sequence in the ith group of PMU measurement data passes serious error troubleshooting; otherwise, the time sequence in the ith PMU measurement data does not pass the abnormal error checking.
Preferably, in the method, the sequentially correcting the abnormal value methods based on the previous mean-difference value specifically includes:
correcting abnormal values sequentially according to the following formula in time sequence:
wherein the content of the first and second substances,
t in time series identified by j in ith set of PMU measurement datamThe abnormal value at the time point, j ═ 1,2,3, is used to identify the voltage amplitude time series, the injected active power time series, and the injected reactive power time series, respectively.
Preferably, in the method, the removing, by using empirical mode decomposition, the low-frequency fluctuation trend of the injection active power time sequence and the injection reactive power time sequence in the ith set of PMU measurement data to obtain a trend-removed stationary signal corresponding to any one of the time sequences specifically includes:
injecting an active power time sequence and an injected reactive power time sequence into the ith set of PMU measurement data, respectively performing multiple rounds of iterative EMD decomposition, removing the last IMF component and the residual error of a fifth threshold value of any time sequence corresponding to the iteration in each round of iterative EMD decomposition, and calculating a signal residual error energy ratio;
and stopping iteration until the energy ratio of the number residual error obtained by the current iteration is greater than a sixth threshold, and eliminating the IMF component and the residual error of the last fifth threshold from the time sequence corresponding to the iteration to obtain the trend-removing stable signal corresponding to the injection active power time sequence and the injection reactive power time sequence.
Preferably, in the method, the performing high-frequency noise filtering based on wavelet threshold denoising on the trend-removed stationary signal corresponding to any time sequence to obtain a denoised estimated signal corresponding to any time sequence specifically includes:
performing wavelet transform decomposition on the trend-free stationary signal corresponding to any time sequence to obtain a wavelet coefficient value corresponding to any time sequence;
and selecting a heuristic threshold adaptive to noise levels of different types of noise signals to filter small-amplitude noise components in wavelet coefficient values corresponding to any time sequence, and then performing inverse wavelet transform processing to obtain a denoised estimation signal corresponding to any time sequence.
In a second aspect, an embodiment of the present invention provides a measured data preprocessing apparatus for dominant load parameter noise identification, including:
the system comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, wherein any group of PMU measurement data comprises a time sequence with N/K total voltage amplitude time points, N/K total time sequences with N/K total active power injection time points and N/K total reactive power injection time points on the high-voltage side of a main transformer of a transformer substation, and N and K are positive integers;
the EMD unit is used for respectively removing low-frequency fluctuation trends of an active power injection time sequence and a reactive power injection time sequence in the ith PMU measurement data by adopting empirical mode decomposition to obtain trend removing stable signals corresponding to any time sequence if any time sequence in the ith PMU measurement data is determined to pass serious error troubleshooting and serious anomaly troubleshooting, wherein i is more than or equal to 1 and less than or equal to K;
and the wavelet threshold unit is used for respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
In a third aspect, an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the steps of the measurement data preprocessing method for dominant load parameter noise identification as provided in the first aspect.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the steps of the metrology data preprocessing method for dominant loading parameter noise-like identification as provided in the first aspect.
The method and the device provided by the embodiment of the invention collect K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, and if any time sequence in the ith group of PMU measurement data is determined to pass serious error investigation and serious abnormity investigation, an empirical mode decomposition is adopted to respectively remove low-frequency fluctuation trends of an active power time sequence and a reactive power time sequence injected into the ith group of PMU measurement data, so as to obtain trend removing stable signals corresponding to any time sequence; and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence. Therefore, the serious bad data with obvious problems are removed through proper data serious error elimination and abnormal data detection, and then the abnormal values and the high-frequency noise of the available measured data are corrected and filtered by a trend removing processing method based on EMD decomposition and a high-frequency noise filtering method based on wavelet threshold denoising. Therefore, the method and the device provided by the embodiment of the invention realize effective data preprocessing on PMU voltage and power measurement data for load identification, improve the identification availability and the overall fitting effect of the data to be identified, and obtain more accurate and reliable dominant load parameters.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief description will be given below of the drawings required for the embodiments or the technical solutions in the prior art, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a measurement data preprocessing method for noise identification of a dominant load parameter according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a measurement data preprocessing device for noise identification of a dominant load parameter according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a preprocessing flow for noise identification of the dominant load parameter class according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without inventive effort based on the embodiments of the present invention, are within the scope of the present invention.
The existing actual measurement PMU data generally has the problems that the identification availability is low, the actual requirement of online identification is difficult to meet, the identification availability of the data to be identified is low, the overall fitting effect is poor, and accurate and reliable dominant load parameters are difficult to obtain. In view of the above, the embodiment of the present invention provides a method for preprocessing measurement data for noise identification of a dominant load parameter. Fig. 1 is a method for preprocessing measurement data of noise-like identification of a dominant load parameter according to an embodiment of the present invention, as shown in fig. 1, the method includes:
and step 110, collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, wherein any group of PMU measurement data comprises a time sequence with the total number of voltage amplitude time points of a main transformer high-voltage side of a transformer substation being N/K, a time sequence with the total number of active power injection time points being N/K and a time sequence with the total number of reactive power injection time points being N/K, and N and K are positive integers.
Specifically, the PMU measurement data includes three physical parameters, which are a voltage amplitude, an injected active power and an injected reactive power of a high-voltage side of the transformer substation, where the transformer substation may be a transformer substation with different voltage levels, and the collected PMU measurement data is only three physical parameters for the high-voltage side of the same transformer substation. During collection, continuous N unit time is collected, for example, a second is taken as a time unit, and then a group of PMU measurement data is obtained every N/K time, so that each group of PMU measurement data comprises a time sequence with the total number of voltage amplitude points of the main transformer high-voltage side of the transformer substation being N/K, a time sequence with the total number of active power injection points being N/K, and a time sequence with the total number of reactive power injection points being N/K, wherein N and K are positive integers.
And 120, if any time sequence in the ith set of PMU measurement data is determined to pass serious error troubleshooting and serious abnormality troubleshooting, respectively removing low-frequency fluctuation trends of an active power injection time sequence and a reactive power injection time sequence in the ith set of PMU measurement data by adopting empirical mode decomposition to obtain trend-removed stable signals corresponding to any time sequence, wherein i is more than or equal to 1 and less than or equal to K.
Specifically, all data are divided into K groups for processing, for the ith (i is more than or equal to 1 and less than or equal to K) group data, firstly, data with serious errors and abnormal data need to be checked, the ith group data are directly rejected when serious errors occur, the ith group data are directly skipped to the processing of the (i + 1) th group data, if the serious errors pass the checking, serious abnormal checking is performed, data which are determined to be abnormal in any sequence continuously occur (the continuous number exceeds a certain threshold value), the ith group data are also directly rejected, the processing of the (i + 1) th group data is skipped, otherwise, the existing sporadic abnormal data are corrected, and any time sequence in the ith group of PMU measurement data is determined to pass the serious error checking and the serious abnormal checking. Then, removing the low-frequency fluctuation trend in the injection active power time sequence and the injection reactive power time sequence by using a trend removing processing method based on EMD decomposition, separating the low-frequency fluctuation trend from the steady noise signals, and obtaining trend removing steady signals corresponding to any time sequence, wherein i is more than or equal to 1 and less than or equal to K.
And step 130, respectively performing high-frequency noise filtering based on wavelet threshold denoising on the trend-removed stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
Specifically, a wavelet threshold denoising method is adopted to perform adaptive denoising on noise-like signals with different effective frequency component distributions and noise levels, and the general operation includes the following three steps: 1. wavelet transform decomposition, in order to avoid signal distortion interference and subsequent identification, Symlets wavelet basis functions with orthogonal symmetrical tight support characteristics are selected for wavelet decomposition; 2. threshold processing, namely selecting a heuristic threshold criterion adaptive to noise levels of different types of noise signals after obtaining a wavelet coefficient value, processing the wavelet coefficient threshold obtained by conversion by adopting a smoother soft threshold function after denoising, and filtering a noise component with a smaller wavelet coefficient amplitude; 3. and (4) signal reconstruction, namely signal reconstruction is carried out through wavelet inverse transformation.
The method provided by the embodiment of the invention comprises the steps of collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, and if any time sequence in the ith group of PMU measurement data is determined to pass serious error investigation and serious anomaly investigation, respectively removing the low-frequency fluctuation trend of the time sequence by adopting empirical mode decomposition to obtain a trend removing stable signal corresponding to the time sequence; and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence. Therefore, the serious bad data with obvious problems are removed through proper data serious error elimination and abnormal data detection, and then the abnormal values and the high-frequency noise of the available measured data are corrected and filtered by a trend removing processing method based on EMD decomposition and a high-frequency noise filtering method based on wavelet threshold denoising. Therefore, the method provided by the embodiment of the invention realizes effective data preprocessing on the PMU voltage and power measurement data for load identification, improves the identification availability and the overall fitting effect of the data to be identified, and obtains more accurate and reliable dominant load parameters.
Based on the above embodiment, the method further includes:
and if the fact that the time sequence in the ith PMU measurement data does not pass the serious error investigation or the serious abnormal investigation is determined, rejecting the ith PMU measurement data.
Specifically, for a time sequence which does not pass serious error investigation or serious abnormal investigation, deleting the whole PMU measurement data group to which the time sequence belongs, and directly preprocessing the next group of PMU measurement data until all the K groups of PMU measurement data are preprocessed.
Based on any of the above embodiments, in the method, performing a fatal error checking on any time sequence in the ith set of PMU measurement data specifically includes:
and if any condition or combination of any condition that the average voltage of the voltage amplitude time sequence deviates from a preset range, the voltage value fluctuation of the voltage amplitude time sequence is smaller than a first threshold, a negative value appears in a power value injected into the active power time sequence, a negative value appears in a power value injected into the reactive power time sequence, the power value fluctuation of the power value injected into the active power time sequence is smaller than a second threshold, and the power value fluctuation of the power value injected into the reactive power time sequence is smaller than a third threshold is detected in the i-th set of PMU measurement data, it is determined that any time sequence in the i-th set of PMU measurement data passes serious error checking, and if not, the time sequence in the i-th set of PMU measurement data does.
Specifically, the determination conditions of the serious error check of different time series (voltage amplitude time series, active power injection time series and reactive power injection time series on the main transformer high-voltage side of the transformer substation) of each set of PMU measurement data are different, and for the voltage amplitude time series, when the average voltage of the voltage amplitude time series deviates from a preset range or the voltage value fluctuation of the voltage amplitude time series is detected to be less than a first threshold, the voltage amplitude time series is determined not to pass the serious error check, the PMU measurement data set to which the voltage amplitude time series belongs needs to be rejected, preferably, the preset range is a normal allowable range of 0.8p.u. -1.2p.u., and the voltage value fluctuation is not less than the first threshold, wherein the voltage value fluctuation is defined as the number of different values of data in the time series sample data and the number of data variation (t-time data minus t-1 time data) which is not zero, based on the length of the data segment to be identified (N/K in the embodiment of the invention), time windows with proper quantity and sliding intervals can be set to carry out sliding coverage type detection according to the volatility definition, and if the number of different values and the number of data variable quantities which are not zero in a certain sliding time window are smaller than a set first threshold value, the volatility is judged to be too small; further, if the number of time windows with too small volatility exceeds 10% of the total number of windows of the data segment to be identified, the data segment cannot pass volatility test and should be screened, and the first threshold is set according to an actual application scenario, for example, when the transformer substation to be tested belongs to transformer substations with different voltage classes, the corresponding first thresholds are also different; when detecting that a negative value occurs to a power value in the active power injection time sequence or the fluctuation of the power value in the active power injection time sequence is smaller than a second threshold, determining that the active power injection time sequence cannot pass a serious error check, wherein a PMU measurement data group to which the active power injection time sequence belongs needs to be rejected, and the fluctuation of the power value cannot be smaller than the second threshold, wherein the fluctuation of the voltage value is defined as the number of different values of data in time sequence sample data and the number of data variation (data at time t minus data at time t-1) which is not zero, based on the length of a data segment to be identified (N/K in the embodiment of the invention), a time window with proper number and sliding intervals can be set for performing sliding overlay detection according to the fluctuation definition, if the number of different values and data variation which are not zero in a certain sliding time window is smaller than the set second threshold, determining that the volatility is too small; further, if the number of the time windows with too small volatility exceeds 10% of the total number of the windows of the data segment to be identified, the data segment cannot pass volatility test and should be screened out, and the second threshold value is set according to an actual application scenario, for example, when the transformer substation to be tested belongs to transformer substations with different voltage classes, the corresponding second threshold values are also different. When detecting that a negative value occurs to a power value in the reactive power injection time sequence or the fluctuation of the power value in the reactive power injection time sequence is smaller than a third threshold, determining that the reactive power injection time sequence cannot pass serious error checking, wherein PMU measurement data groups to which the reactive power injection time sequence belongs need to be removed, and the fluctuation of the power value cannot be smaller than the third threshold, wherein voltage value fluctuation is defined as the number of different values of data in time sequence sample data and the number of data variation (data at time t minus data at time t-1) which are not zero, based on the length of a data segment to be identified (in the embodiment of the invention, N/K), a time window with proper number and sliding intervals can be set for sliding coverage detection according to the fluctuation definition, if the number of different values and data variation which are not zero in a certain sliding time window is smaller than the set third threshold, determining that the volatility is too small; further, if the number of the time windows with too small volatility exceeds 10% of the total number of the windows of the data segment to be identified, the data segment cannot pass the volatility test and should be screened out, and the third threshold is set according to an actual application scenario, for example, when the transformer substation to be tested belongs to transformer substations with different voltage classes, the corresponding third thresholds are also different.
Based on any of the above embodiments, in the method, performing abnormal error checking on any time sequence in the ith set of PMU measurement data specifically includes:
determining the mean value of time series in the ith set of PMU measurement data
And standard deviation of
Wherein j is 1,2,3, which is used to identify a voltage amplitude time sequence, an injection active power time sequence and an injection reactive power time sequence, respectively;
for any time sequence in the ith set of PMU measurement data
Determining that the value of the sequence element is not
The elements in the range are outliers of the time series identified by j;
if the abnormal values in any time sequence are determined not to continuously have the fourth threshold value and more than one abnormal value, sequentially correcting each abnormal value method based on the previous average difference value and judging that any time sequence in the ith group of PMU measurement data passes serious error troubleshooting; otherwise, the time sequence in the ith PMU measurement data does not pass the abnormal error checking.
Specifically, data points in the metrology sample data that are not within the range of [ μ -3 σ, μ +3 σ ] (μ is the sample mean, σ is the sample standard deviation) are determined to be outliers using the "3 σ" criterion. And then, correcting the sequence abnormal value by adopting a method of previous mean value interpolation. Particularly, data samples with a fourth threshold value and above judgment abnormal values are directly eliminated, and the preprocessing flow of the next group of PMU measurement data is entered. Preferably, the fourth threshold value is 5.
Based on any one of the embodiments, in the method, the sequentially correcting each abnormal value method based on the previous mean-difference value specifically includes:
correcting abnormal values sequentially according to the following formula in time sequence:
wherein the content of the first and second substances,
t in time series identified by j in ith set of PMU measurement datamThe abnormal value at the time point, j ═ 1,2,3, is used to identify the voltage amplitude time series, the injected active power time series, and the injected reactive power time series, respectively.
Specifically, the above formula provides a method for calculating the mean difference of the antecedents.
Based on any of the above embodiments, in the method, the removing, by using empirical mode decomposition, the low-frequency fluctuation trend of the injection active power time sequence and the injection reactive power time sequence in the ith set of PMU measurement data to obtain a trend-removed stationary signal corresponding to any time sequence specifically includes:
injecting an active power time sequence and an injected reactive power time sequence into the ith set of PMU measurement data, respectively performing multiple rounds of iterative EMD decomposition, removing the last IMF component and the residual error of a fifth threshold value of any time sequence corresponding to the iteration in each round of iterative EMD decomposition, and calculating a signal residual error energy ratio;
and stopping iteration until the energy ratio of the number residual error obtained by the current iteration is greater than a sixth threshold, and eliminating the IMF component and the residual error of the last fifth threshold from the time sequence corresponding to the iteration to obtain the trend-removing stable signal corresponding to the injection active power time sequence and the injection reactive power time sequence.
Specifically, after rough screening and abnormal value processing are carried out on PMU measurement data, the low-frequency fluctuation trend of a power signal in the PMU measurement data is removed by Empirical Mode Decomposition (EMD), and a stable signal for subsequent identification is obtained. The basic idea of EMD decomposition is as follows: and eliminating the last fifth threshold Intrinsic Mode Function (IMF) component and residual error representing the low-frequency fluctuation mode through proper times of decomposition, and reconstructing the retained IMF component to achieve the purpose of data trend elimination. Further, the signal residual energy ratio is selected as a stopping condition for EMD decomposition. And stopping decomposition when the signal residual energy ratio is larger than a sixth threshold, removing IMF components and residual errors of the last fifth threshold to obtain a trend-removing stable signal, and reconstructing the retained IMF components to achieve the aim of removing the trend of the data. Preferably, the fifth threshold value is 1, and the sixth threshold value is 20 dB.
Based on any of the above embodiments, in the method, the performing high-frequency noise filtering based on wavelet threshold denoising on the detrending stationary signal corresponding to any time sequence to obtain a denoised estimated signal corresponding to any time sequence specifically includes:
performing wavelet transform decomposition on the trend-free stationary signal corresponding to any time sequence to obtain a wavelet coefficient value corresponding to any time sequence;
and selecting a heuristic threshold adaptive to noise levels of different types of noise signals to filter small-amplitude noise components in wavelet coefficient values corresponding to any time sequence, and then performing inverse wavelet transform processing to obtain a denoised estimation signal corresponding to any time sequence.
Specifically, wavelet threshold denoising comprises the following steps:
1. wavelet transform decomposition, in order to avoid signal distortion interference and subsequent identification, Symlets wavelet basis functions with orthogonal symmetrical tight support characteristics are selected for multi-scale wavelet transform, preferably, Sym8 wavelets are taken as basis functions, and the number of wavelet decomposition layers is set to be 5.
2. And (4) threshold processing, namely selecting a heuristic threshold criterion for self-adapting to the noise levels of different types of noise signals to define the noise and useful signals after obtaining the wavelet coefficient value. The heuristic threshold is a compromise between two criteria of unbiased likelihood estimation and fixed threshold estimation, and if the signal-to-noise ratio is very small, the fixed threshold form is selected:
in the above formula, λ is a threshold estimation value, M is a length of a wavelet coefficient vector, otherwise, unbiased likelihood estimation is adopted:
in the above formula, f (-) is a new sequence obtained by squaring absolute values of wavelet coefficient vectors and then sorting the wavelet coefficient vectors from small to large, kminThe subscript corresponding to the minimum point of the Risk vector Risk, wherein,
then, a wavelet coefficient threshold lambda obtained by processing and transforming the denoised smoother soft threshold function is adopted to filter a noise component with a smaller wavelet coefficient amplitude, and the processing expression of the soft threshold function is as follows:
in the above formula, w is the original wavelet coefficient, wλFor the processed wavelet coefficients, sgn (w) is a sign function.
3. And (3) signal reconstruction, namely obtaining a denoised time domain estimation signal through wavelet inverse transformation, completing a preprocessing process of the noise-like signal, and finally using the preprocessed signal for the identification of the Z + M model leading load model parameters based on the actually measured noise-like.
Based on any of the above embodiments, an embodiment of the present invention provides a measured data preprocessing device for dominant load parameter noise identification, and fig. 2 is a schematic structural diagram of the measured data preprocessing device for dominant load parameter noise identification according to the embodiment of the present invention. As shown in fig. 2, the apparatus includes an acquisition unit 210, an EMD unit 220, and a wavelet threshold unit 230, wherein,
the acquisition unit 210 is configured to acquire K sets of PMU measurement data for leading load parameter noise identification for N consecutive unit times, where any set of PMU measurement data includes a time sequence with N/K total voltage amplitude time points at a main transformer high-voltage side of the transformer substation, a time sequence with N/K total active power injection time points, and a time sequence with N/K total reactive power injection time points, where N and K are positive integers;
the EMD unit 220 is configured to, if it is determined that any time sequence in the ith set of PMU measurement data passes through serious error troubleshooting and serious anomaly troubleshooting, respectively remove low-frequency fluctuation trends of an active power injection time sequence and a reactive power injection time sequence in the ith set of PMU measurement data by using empirical mode decomposition, and obtain trend-removed stationary signals corresponding to the time sequence sequences, where i is greater than or equal to 1 and less than or equal to K;
the wavelet threshold unit 230 is configured to perform high-frequency noise filtering based on wavelet threshold denoising on the trend-removed stationary signal corresponding to any time sequence, so as to obtain a denoised estimated signal corresponding to any time sequence.
The device provided by the embodiment of the invention collects K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, and if any time sequence in the ith group of PMU measurement data is determined to pass serious error investigation and serious anomaly investigation, the low-frequency fluctuation trend of any time sequence is respectively removed by adopting empirical mode decomposition, so as to obtain a trend removing stable signal corresponding to any time sequence; and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence. Therefore, the serious bad data with obvious problems are removed through proper data serious error elimination and abnormal data detection, and then the abnormal values and the high-frequency noise of the available measured data are corrected and filtered by a trend removing processing method based on EMD decomposition and a high-frequency noise filtering method based on wavelet threshold denoising. Therefore, the device provided by the embodiment of the invention realizes effective data preprocessing on PMU voltage and power measurement data for load identification, improves the identification availability and the overall fitting effect of the data to be identified, and obtains more accurate and reliable dominant load parameters.
Based on any embodiment, the device also comprises a rejecting unit,
and if the time sequence in the ith PMU measurement data is determined to fail to pass the serious error investigation or the serious abnormal investigation, the ith PMU measurement data is removed.
Based on any of the above embodiments, in the apparatus, performing a fatal error checking on any time sequence in the ith set of PMU measurement data specifically includes:
and if any condition or combination of any condition that the average voltage of the voltage amplitude time sequence deviates from a preset range, the voltage value fluctuation of the voltage amplitude time sequence is smaller than a first threshold, a negative value appears in a power value injected into the active power time sequence, a negative value appears in a power value injected into the reactive power time sequence, the power value fluctuation of the power value injected into the active power time sequence is smaller than a second threshold, and the power value fluctuation of the power value injected into the reactive power time sequence is smaller than a third threshold is detected in the i-th set of PMU measurement data, it is determined that any time sequence in the i-th set of PMU measurement data passes serious error checking, and if not, the time sequence in the i-th set of PMU measurement data does.
Based on any of the above embodiments, in the apparatus, performing abnormal error checking on any time sequence in the ith set of PMU measurement data specifically includes:
determining the mean value of time series in the ith set of PMU measurement data
And standard deviation of
Wherein j is 1,2,3, which is used to identify a voltage amplitude time sequence, an injection active power time sequence and an injection reactive power time sequence, respectively;
for any time sequence in the ith set of PMU measurement data
Determining that the value of the sequence element is not
The elements in the range are outliers of the time series identified by j;
if the abnormal values in any time sequence are determined not to continuously have the fourth threshold value and more than one abnormal value, sequentially correcting each abnormal value method based on the previous average difference value and judging that any time sequence in the ith group of PMU measurement data passes serious error troubleshooting; otherwise, the time sequence in the ith PMU measurement data does not pass the abnormal error checking.
Based on any one of the above embodiments, in the apparatus, the sequentially correcting each abnormal value method based on the previous mean-difference value specifically includes:
correcting abnormal values sequentially according to the following formula in time sequence:
wherein the content of the first and second substances,
t in time series identified by j in ith set of PMU measurement datamThe abnormal value at the time point, j ═ 1,2,3, is used to identify the voltage amplitude time series, the injected active power time series, and the injected reactive power time series, respectively.
Based on any of the above embodiments, in the apparatus, the removing, by using empirical mode decomposition, the low-frequency fluctuation trend of the injection active power time sequence and the injection reactive power time sequence in the i-th set of PMU measurement data to obtain a trend-removed stationary signal corresponding to any time sequence specifically includes:
injecting an active power time sequence and an injected reactive power time sequence into the ith set of PMU measurement data, respectively performing multiple rounds of iterative EMD decomposition, removing the last IMF component and the residual error of a fifth threshold value of any time sequence corresponding to the iteration in each round of iterative EMD decomposition, and calculating a signal residual error energy ratio;
and stopping iteration until the energy ratio of the number residual error obtained by the current iteration is greater than a sixth threshold, and eliminating the IMF component and the residual error of the last fifth threshold from the time sequence corresponding to the iteration to obtain the trend-removing stable signal corresponding to the injection active power time sequence and the injection reactive power time sequence.
Based on any of the above embodiments, in the apparatus, the performing high-frequency noise filtering based on wavelet threshold denoising on the trend-removed stationary signal corresponding to any time sequence to obtain a denoised estimated signal corresponding to any time sequence specifically includes:
performing wavelet transform decomposition on the trend-free stationary signal corresponding to any time sequence to obtain a wavelet coefficient value corresponding to any time sequence;
and selecting a heuristic threshold adaptive to noise levels of different types of noise signals to filter small-amplitude noise components in wavelet coefficient values corresponding to any time sequence, and then performing inverse wavelet transform processing to obtain a denoised estimation signal corresponding to any time sequence.
Based on any of the above embodiments, an embodiment of the present invention provides a preprocessing flow for identifying noise of a dominant load parameter class, and fig. 3 is a schematic diagram of the preprocessing flow for identifying noise of the dominant load parameter class provided by the embodiment of the present invention. As shown in fig. 3, firstly, grouping PMU measurement data in time sequence, for the k-th PMU measurement data in the preprocessing process, firstly, performing data rough screening to determine whether a serious data problem exists, if so, skipping the next preprocessing, returning to the first step to start preprocessing of the k + 1-th PMU measurement data, if not, performing the next detection and processing of abnormal values to determine whether a plurality of continuous abnormalities exist, if so, skipping the next preprocessing, returning to the first step to start preprocessing of the k + 1-th PMU measurement data, if not, performing the next trend removing of the power signal EMD, denoising with a wavelet threshold, reconstructing the measurement data identified as the noise-like dominant load parameter, determining whether k reaches the maximum group number (i.e., whether k is the last group of PMU measurement data), and if not, continuing preprocessing of the next group of data, if so, finishing the preprocessing of all the data and finishing the preprocessing.
Fig. 4 is a schematic entity structure diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 4, the electronic device may include: a processor (processor)401, a communication Interface (communication Interface)402, a memory (memory)403 and a communication bus 404, wherein the processor 401, the communication Interface 402 and the memory 403 complete communication with each other through the communication bus 404. The processor 401 may invoke a computer program stored in the memory 403 and executable on the processor 401 to perform the metrology data preprocessing method for dominant load parameter noise-like identification provided by the embodiments described above, including, for example: collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, wherein any group of PMU measurement data comprises a time sequence with N/K total voltage amplitude time points, a time sequence with N/K total active power injection time points and a time sequence with N/K total reactive power injection time points on a main transformer high-voltage side of a transformer substation, and N and K are both positive integers; if any time sequence in the ith set of PMU measurement data is determined to pass serious error troubleshooting and serious abnormality troubleshooting, respectively removing low-frequency fluctuation trends of an active power time sequence and a reactive power time sequence which are injected into the ith set of PMU measurement data by adopting empirical mode decomposition to obtain trend-removed stable signals corresponding to any time sequence, wherein i is more than or equal to 1 and less than or equal to K; and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
In addition, the logic instructions in the memory 403 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solutions of the embodiments of the present invention may be essentially implemented or make a contribution to the prior art, or may be implemented in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
An embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, is implemented to perform the method for preprocessing measurement data for noise identification of a dominant load parameter, which includes: collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, wherein any group of PMU measurement data comprises a time sequence with N/K total voltage amplitude time points, a time sequence with N/K total active power injection time points and a time sequence with N/K total reactive power injection time points on a main transformer high-voltage side of a transformer substation, and N and K are both positive integers; if any time sequence in the ith set of PMU measurement data is determined to pass serious error troubleshooting and serious abnormality troubleshooting, respectively removing low-frequency fluctuation trends of an active power time sequence and a reactive power time sequence which are injected into the ith set of PMU measurement data by adopting empirical mode decomposition to obtain trend-removed stable signals corresponding to any time sequence, wherein i is more than or equal to 1 and less than or equal to K; and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for preprocessing measurement data for noise identification of a dominant load parameter class, comprising:
collecting K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, wherein any group of PMU measurement data comprises a time sequence with N/K total voltage amplitude time points, N/K total time sequences of active power injection time points and N/K total time sequences of reactive power injection time points on a main transformer high-voltage side of a transformer substation, and N and K are positive integers;
if any time sequence in the ith set of PMU measurement data is determined to pass serious error troubleshooting and serious abnormality troubleshooting, respectively removing low-frequency fluctuation trends of an active power time sequence and a reactive power time sequence which are injected into the ith set of PMU measurement data by adopting empirical mode decomposition to obtain trend-removed stable signals corresponding to any time sequence, wherein i is more than or equal to 1 and less than or equal to K;
and respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
2. The method of claim 1, further comprising:
and if the fact that the time sequence in the ith PMU measurement data does not pass the serious error investigation or the serious abnormal investigation is determined, rejecting the ith PMU measurement data.
3. The method as claimed in claim 1 or 2, wherein the performing a fatal error check on any time sequence of the ith set of PMU measurement data includes:
and if any condition or combination of any condition that the average voltage of the voltage amplitude time sequence deviates from a preset range, the voltage value fluctuation of the voltage amplitude time sequence is smaller than a first threshold, a negative value appears in a power value injected into the active power time sequence, a negative value appears in a power value injected into the reactive power time sequence, the power value fluctuation of the power value injected into the active power time sequence is smaller than a second threshold, and the power value fluctuation of the power value injected into the reactive power time sequence is smaller than a third threshold is detected in the i-th set of PMU measurement data, it is determined that any time sequence in the i-th set of PMU measurement data passes serious error checking, and if not, the time sequence in the i-th set of PMU measurement data does.
4. The method as claimed in claim 1 or 2, wherein the performing an abnormal error check on any time sequence in the ith set of PMU measurement data includes:
determining the mean value of time series in the ith set of PMU measurement data
And standard deviation of
Wherein j is 1,2,3, which is used to identify a voltage amplitude time sequence, an injection active power time sequence and an injection reactive power time sequence, respectively;
for any time sequence in the ith set of PMU measurement data
Determining that the value of the sequence element is not
The elements in the range are outliers of the time series identified by j;
if the abnormal values in any time sequence are determined not to continuously have the fourth threshold value and more than one abnormal value, sequentially correcting each abnormal value method based on the previous average difference value and judging that any time sequence in the ith group of PMU measurement data passes serious error troubleshooting; otherwise, the time sequence in the ith PMU measurement data does not pass the abnormal error checking.
5. The method as claimed in claim 4, wherein the step of sequentially modifying the abnormal value methods based on the previous mean-difference value includes:
correcting abnormal values sequentially according to the following formula in time sequence:
wherein the content of the first and second substances,
t in time series identified by j in ith set of PMU measurement datamAbnormal value of time point, j ═1,2 and 3, respectively used for identifying a voltage amplitude time sequence, an injection active power time sequence and an injection reactive power time sequence.
6. The method for preprocessing measurement data for noise identification of a leading load parameter of claim 1 or 2, wherein the step of removing a low-frequency fluctuation trend of an active power time sequence and an injected reactive power time sequence in the i-th set of PMU measurement data by empirical mode decomposition to obtain a trend-removed stationary signal corresponding to any time sequence comprises:
injecting an active power time sequence and an injected reactive power time sequence into the ith set of PMU measurement data, respectively performing multiple rounds of iterative EMD decomposition, removing the last IMF component and the residual error of a fifth threshold value of any time sequence corresponding to the iteration in each round of iterative EMD decomposition, and calculating a signal residual error energy ratio;
and stopping iteration until the energy ratio of the number residual error obtained by the current iteration is greater than a sixth threshold, and eliminating the IMF component and the residual error of the last fifth threshold from the time sequence corresponding to the iteration to obtain the trend-removing stable signal corresponding to the injection active power time sequence and the injection reactive power time sequence.
7. The method for preprocessing measurement data for noise-like identification of a dominant load parameter according to claim 1 or 2, wherein the step of respectively performing high-frequency noise filtering based on wavelet threshold denoising on the detrended stationary signal corresponding to any time sequence to obtain a denoised estimated signal corresponding to any time sequence comprises:
performing wavelet transform decomposition on the trend-free stationary signal corresponding to any time sequence to obtain a wavelet coefficient value corresponding to any time sequence;
and selecting a heuristic threshold adaptive to noise levels of different types of noise signals to filter small-amplitude noise components in wavelet coefficient values corresponding to any time sequence, and then performing inverse wavelet transform processing to obtain a denoised estimation signal corresponding to any time sequence.
8. A pre-processing apparatus for metrology data for noise-like identification of dominant load parameters, comprising:
the system comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring K groups of PMU measurement data for leading load parameter noise identification in N continuous unit time, and any group of PMU measurement data comprises a time sequence with N/K total voltage amplitude time points, N/K total active power injection time points and N/K total reactive power injection time points on the high-voltage side of a main transformer of a transformer substation, wherein N and K are positive integers;
the EMD unit is used for respectively removing low-frequency fluctuation trends of an active power injection time sequence and a reactive power injection time sequence in the ith PMU measurement data by adopting empirical mode decomposition to obtain trend removing stable signals corresponding to any time sequence if any time sequence in the ith PMU measurement data is determined to pass serious error troubleshooting and serious anomaly troubleshooting, wherein i is more than or equal to 1 and less than or equal to K;
and the wavelet threshold unit is used for respectively carrying out high-frequency noise filtering based on wavelet threshold denoising on the trend-removing stationary signals corresponding to any time sequence to obtain denoised estimated signals corresponding to any time sequence.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the method for pre-processing metrology data for dominant load parameter noise-like identification as claimed in any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for preprocessing metrology data for dominant load parameter noise-like identification as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011025550.0A CN112213561B (en) | 2020-09-25 | 2020-09-25 | Measurement data preprocessing method and device for leading load parameter noise identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011025550.0A CN112213561B (en) | 2020-09-25 | 2020-09-25 | Measurement data preprocessing method and device for leading load parameter noise identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112213561A true CN112213561A (en) | 2021-01-12 |
| CN112213561B CN112213561B (en) | 2022-01-18 |
Family
ID=74051147
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011025550.0A Active CN112213561B (en) | 2020-09-25 | 2020-09-25 | Measurement data preprocessing method and device for leading load parameter noise identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112213561B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114705289A (en) * | 2022-04-13 | 2022-07-05 | 中国石油天然气集团有限公司 | Method, system and equipment for measuring vibration of drilling tool while drilling |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007001236A1 (en) * | 2005-06-28 | 2007-01-04 | Agency For Science, Technology And Research | Wavelet shrinkage pre-filtering of mr images for brain tissue segmentation |
| CN103558441A (en) * | 2013-10-25 | 2014-02-05 | 兰州交通大学 | Noise removing method for contact net insulator leaked currents |
| CN104135521A (en) * | 2014-07-29 | 2014-11-05 | 广东省环境监测中心 | Method and system of identifying data abnormal values of environment automatic monitoring network |
| CN104636609A (en) * | 2015-01-30 | 2015-05-20 | 电子科技大学 | Signal combined denoising method based on empirical mode decomposition (EMD) and wavelet analysis |
| CN105975510A (en) * | 2016-04-28 | 2016-09-28 | 广东鹰视能效科技有限公司 | Method and system for carrying out automatic additional recording of data on the basis of intelligent instrument |
| CN106786567A (en) * | 2017-03-02 | 2017-05-31 | 国家电网公司 | A kind of online load modeling method based on PMU noise like data |
| CN107422235A (en) * | 2017-08-30 | 2017-12-01 | 广东电网有限责任公司珠海供电局 | A kind of cable local discharge Signal Pre-Processing Method |
| CN109086928A (en) * | 2018-07-27 | 2018-12-25 | 福州大学 | Photovoltaic plant realtime power prediction technique based on SAGA-FCM-LSSVM model |
| CN110926386A (en) * | 2019-11-25 | 2020-03-27 | 国网吉林省电力有限公司电力科学研究院 | CEEMD-improved wavelet threshold denoising-based transformer winding ultrasonic detection three-dimensional imaging method |
| CN111190890A (en) * | 2019-12-26 | 2020-05-22 | 深圳供电局有限公司 | User daily electricity quantity data cleaning method, device, equipment and medium |
| CN111697574A (en) * | 2020-06-19 | 2020-09-22 | 杭州岑石能源科技有限公司 | Short-term low-voltage distribution network power transfer judgment method based on 6sigma principle |
-
2020
- 2020-09-25 CN CN202011025550.0A patent/CN112213561B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007001236A1 (en) * | 2005-06-28 | 2007-01-04 | Agency For Science, Technology And Research | Wavelet shrinkage pre-filtering of mr images for brain tissue segmentation |
| CN103558441A (en) * | 2013-10-25 | 2014-02-05 | 兰州交通大学 | Noise removing method for contact net insulator leaked currents |
| CN104135521A (en) * | 2014-07-29 | 2014-11-05 | 广东省环境监测中心 | Method and system of identifying data abnormal values of environment automatic monitoring network |
| CN104636609A (en) * | 2015-01-30 | 2015-05-20 | 电子科技大学 | Signal combined denoising method based on empirical mode decomposition (EMD) and wavelet analysis |
| CN105975510A (en) * | 2016-04-28 | 2016-09-28 | 广东鹰视能效科技有限公司 | Method and system for carrying out automatic additional recording of data on the basis of intelligent instrument |
| CN106786567A (en) * | 2017-03-02 | 2017-05-31 | 国家电网公司 | A kind of online load modeling method based on PMU noise like data |
| CN107422235A (en) * | 2017-08-30 | 2017-12-01 | 广东电网有限责任公司珠海供电局 | A kind of cable local discharge Signal Pre-Processing Method |
| CN109086928A (en) * | 2018-07-27 | 2018-12-25 | 福州大学 | Photovoltaic plant realtime power prediction technique based on SAGA-FCM-LSSVM model |
| CN110926386A (en) * | 2019-11-25 | 2020-03-27 | 国网吉林省电力有限公司电力科学研究院 | CEEMD-improved wavelet threshold denoising-based transformer winding ultrasonic detection three-dimensional imaging method |
| CN111190890A (en) * | 2019-12-26 | 2020-05-22 | 深圳供电局有限公司 | User daily electricity quantity data cleaning method, device, equipment and medium |
| CN111697574A (en) * | 2020-06-19 | 2020-09-22 | 杭州岑石能源科技有限公司 | Short-term low-voltage distribution network power transfer judgment method based on 6sigma principle |
Non-Patent Citations (1)
| Title |
|---|
| 王颖 等: "类噪声辨识负荷参数的实际电网校验方法", 《电网技术》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114705289A (en) * | 2022-04-13 | 2022-07-05 | 中国石油天然气集团有限公司 | Method, system and equipment for measuring vibration of drilling tool while drilling |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112213561B (en) | 2022-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111461533A (en) | Fault monitoring method and system of industrial production line based on big data | |
| CN111079710B (en) | Multilayer noise reduction method based on improved CEEMD rolling bearing signal | |
| CN112213561B (en) | Measurement data preprocessing method and device for leading load parameter noise identification | |
| CN116865269B (en) | Wind turbine generator system high harmonic compensation method and system | |
| CN116739829B (en) | Big data-based power data analysis method, system and medium | |
| CN110927497B (en) | Point switch fault detection method and device | |
| CN111239565B (en) | Oil-filled casing partial discharge pulse signal processing method and system based on layered denoising model | |
| CN111783696A (en) | Edge calculation method for low-voltage branch topology real-time analysis based on PV relation | |
| CN115800272A (en) | Power grid fault analysis method, system, terminal and medium based on topology identification | |
| Swaminathan et al. | A CNN-LSTM-based fault classifier and locator for underground cables | |
| CN112202630A (en) | Network quality abnormity detection method and device based on unsupervised model | |
| CN113490857A (en) | Method and test device | |
| CN116482526A (en) | Analysis system for non-fault phase impedance relay | |
| CN114676593B (en) | Abnormality detection method and related device for textile equipment | |
| Nicolae et al. | Comparative study of wavelet based techniques for electromagnetic noise evaluation and removal | |
| CN113986711A (en) | Time series data peak value detection method, device and equipment | |
| CN114065815A (en) | Transformer partial discharge identification method and system based on EEMD and neural network | |
| CN114089218A (en) | Power distribution network high-resistance grounding fault identification method, device, terminal and medium | |
| Nicolae et al. | Improving the efficiency of PQ analysis concerning quasistationary waveforms through special computational techniques | |
| CN113655344B (en) | Power distribution network fault identification method, device, terminal and medium | |
| CN117388637B (en) | AI-based converter station direct current system abnormal signal identification and auxiliary decision-making method | |
| Ming et al. | Partial discharge de-noising based on hybrid particle swarm optimization SWT adaptive threshold | |
| Kavya et al. | Parabolic filter for removal of powerline interference in ecg signal using periodogram estimation technique | |
| Kaszowska et al. | Assessment of Available Measurement Data, Data Breaks and Estimation of Missing Data from AMI Meters | |
| CN113190788A (en) | Method and device for adaptively extracting and reducing bus characteristics of power distribution system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |