CN118033300A - Intelligent identification method for load power failure abnormality of dynamic voltage regulator - Google Patents
Intelligent identification method for load power failure abnormality of dynamic voltage regulator Download PDFInfo
- Publication number
- CN118033300A CN118033300A CN202410425326.2A CN202410425326A CN118033300A CN 118033300 A CN118033300 A CN 118033300A CN 202410425326 A CN202410425326 A CN 202410425326A CN 118033300 A CN118033300 A CN 118033300A
- Authority
- CN
- China
- Prior art keywords
- fitting
- data point
- data
- point
- measured
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000005856 abnormality Effects 0.000 title claims abstract description 22
- 238000005070 sampling Methods 0.000 claims abstract description 144
- 230000002159 abnormal effect Effects 0.000 claims abstract description 26
- 238000010606 normalization Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 13
- 238000004364 calculation method Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/165—Indicating that current or voltage is either above or below a predetermined value or within or outside a predetermined range of values
- G01R19/16533—Indicating that current or voltage is either above or below a predetermined value or within or outside a predetermined range of values characterised by the application
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to the technical field of electrical equipment anomaly detection, in particular to an intelligent identification method for load power failure anomalies of a dynamic voltage regulator. According to the method, according to the data values of a data point to be measured and a later undetermined number of data points to be measured, the fitting tolerance of the data point to be measured under the undetermined number is obtained according to the interval between the data point to be measured and the sampling time of the fitting demarcation point under the undetermined number, whether the fitting tolerance of the data point to be measured under the undetermined number at the sampling time meets undetermined conditions is judged, the target fitting number at the sampling time is obtained, and then the fitting time is screened out; and adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points at the fitting time to obtain a corrected reachable distance between the fitting data points, and detecting the load power failure abnormal condition of the dynamic regulator based on the corrected reachable distance. The invention screens the fitting time to construct fitting data points, and utilizes the fitting data points to detect the load power failure abnormality of the dynamic voltage regulator.
Description
Technical Field
The invention relates to the technical field of electrical equipment anomaly detection, in particular to an intelligent identification method for load power failure anomalies of a dynamic voltage regulator.
Background
A dynamic voltage regulator is a device for improving the voltage quality of an electrical power system, the main function of which is to provide fast voltage regulation in case of grid voltage fluctuations or transient faults. During the operation of the dynamic voltage regulator, a situation that the load of the dynamic voltage regulator loses power and is abnormal due to the fault of the power system may be faced, so that the power loss abnormality of the dynamic voltage regulator needs to be monitored.
The abnormal power failure detection of the dynamic voltage regulator mainly comprises the steps of identifying voltage and current, acquiring huge data volume due to high frequency of data acquisition, and preserving data trend while reducing the data volume by using a Fabry-Perot algorithm. The current and the voltage are respectively subjected to data fitting by utilizing a Target-Purchase algorithm to reduce data points, and only one current value and one voltage value at the same moment after data fitting can be caused due to the fact that the fluctuation of the current and the voltage of the dynamic voltage regulator is different, so that the data points at the moment can not be built to detect abnormal load power failure of the dynamic voltage regulator.
Disclosure of Invention
In order to solve the problem that the fluctuation of the current and the voltage of the dynamic voltage regulator is different, only one current value and one voltage value exist at the same moment after data fitting, and the technical problem of influencing the detection of the load power failure abnormality of the dynamic voltage regulator is solved, the invention aims to provide an intelligent identification method for the load power failure abnormality of the dynamic voltage regulator, and the adopted technical scheme is as follows:
the invention provides a load power failure abnormality intelligent identification method of a dynamic voltage regulator, which comprises the following steps:
acquiring a current data point of a current data value at each sampling time and a voltage data point of a voltage data value at each sampling time of a dynamic voltage regulator in a preset time period; recording the current data point and the voltage data point as data points to be measured;
acquiring fitting demarcation points of each data point to be measured under the undetermined number; combining the data value of each data point to be measured and a number of data points to be measured after the data points to be measured, and the time interval between each data point to be measured and the corresponding sampling time of the fitting demarcation point of the data points to be measured under the number to be measured, so as to obtain the fitting tolerance of each data point to be measured under the number to be measured; when the fitting tolerance of the current data point and the voltage data point at each sampling moment under the undetermined quantity respectively does not meet the to-be-selected condition, updating the to-be-selected quantity until the fitting tolerance of the current data point and the voltage data point at each sampling moment under the updated to-be-selected quantity respectively meets the to-be-selected condition; when the condition to be selected is met, taking the updated undetermined number as the target fitting number of each sampling moment;
Selecting different fitting moments from sampling moments in a preset time period based on the target fitting quantity;
Acquiring fitting data points of a current data value and a voltage data value at each fitting moment; adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points corresponding to the fitting time to obtain a corrected reachable distance between the fitting data points;
and detecting the load power failure abnormal condition of the dynamic regulator based on the corrected reachable distance.
Further, the method for obtaining the fitting demarcation point of each data point to be measured under the undetermined quantity comprises the following steps:
selecting any one data point to be detected as a characteristic data point, and establishing a two-dimensional coordinate system by taking time as a horizontal axis and a data value of the characteristic data point as a vertical axis; labeling a feature data point and a number of data points to be measured after the feature data point in the two-dimensional coordinate system to obtain a coordinate point to be measured corresponding to the data point to be measured;
Connecting a coordinate point to be measured with the minimum abscissa and a coordinate point to be measured with the maximum abscissa in the two-dimensional coordinate system to obtain a characteristic line segment; and obtaining the distance between each coordinate point to be measured and the characteristic line segment, and taking the data point to be measured of the coordinate point to be measured corresponding to the maximum distance as the fitting demarcation point of the characteristic data point under the undetermined quantity.
Further, the method for obtaining the fitting tolerance of each data point to be measured under the undetermined number by combining the data value of the undetermined number of data points to be measured after each data point to be measured and the time interval between the undetermined number of fitting demarcation points of each data point to be measured and the corresponding sampling time of each data point to be measured, includes:
Respectively acquiring a first coordinate point and a second coordinate point from the feature data point and a plurality of data points to be measured after the feature data point in the two-dimensional coordinate system; the ordinate of each first coordinate point is larger than the ordinate of the corresponding point of the first coordinate point on the characteristic line segment, and the ordinate of each second coordinate point is smaller than the ordinate of the corresponding point of the second coordinate point on the characteristic line segment;
Respectively counting the total number of the first coordinate points to be used as a first value of the feature data points under the undetermined number, and the total number of the second coordinate points to be used as a second value of the feature data points under the undetermined number; taking the distance between the coordinate point to be measured corresponding to the fitting demarcation point of the characteristic data point under the undetermined number and the characteristic line segment in the two-dimensional coordinate system as the target distance of the characteristic data point under the undetermined number;
And combining the time interval between the fitting demarcation points of the characteristic data points and the characteristic data points under the undetermined quantity and the corresponding sampling moments of the fitting demarcation points of the characteristic data points, and obtaining the fitting tolerance of the characteristic data points under the undetermined quantity by the distance between the second value and the target.
Further, the fit tolerance of the characteristic data points at the undetermined number is calculated as follows:
; in the/> The fit tolerance for a feature data point at a pending number; n is a pending number; a is a preset reference positive number; /(I)The first value at a pending number for a feature data point; /(I)The second value at the pending number for a feature data point; The target distance for a feature data point at a pending number; t is the sampling time corresponding to the characteristic data point; sampling time corresponding to the fitting demarcation points of the feature data points under the undetermined quantity; /(I) As a function of absolute value; max is a maximum function; exp is an exponential function based on a natural constant e; norms are normalization functions.
Further, the fitting tolerance of the current data point and the voltage data point at each sampling time under the undetermined number does not meet the condition to be selected, which is that:
and the maximum value of the fitting tolerance of the current data point and the voltage data point at each sampling moment under the undetermined quantity is smaller than a preset tolerance threshold.
Further, the method for updating the to-be-counted number comprises the following steps:
The value obtained by subtracting the preset positive integer from the undetermined number is used as the undetermined number after updating.
Further, the method for selecting different fitting moments from sampling moments of a preset time period includes:
Starting from a first sampling time in a preset time period, taking the first sampling time as a fitting time, taking the sampling time which is spaced from each fitting time by the target fitting number of sampling times as the next fitting time of each fitting time, and until the number of sampling times after each fitting time is smaller than the target fitting number of fitting times.
Further, the method for adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points corresponding to the fitting time to obtain the corrected reachable distance between the fitting data points includes:
Selecting any one fitting data point as a core data point, selecting any one data point from a k-th distance neighborhood of the core data point as an example data point, and taking the product of Euclidean distance between the core data point and the example data point and the target fitting quantity of the fitting time corresponding to the core data point as a correction distance between the core data point and the example data point; taking the ratio of the kth distance of the core data point to the target fitting quantity of the fitting time corresponding to the example data point as the corrected kth distance of the core data point; the maximum of the modified distance and the modified kth distance is taken as a modified reachable distance between a core data point and an example data point.
Further, the method for detecting the load power failure abnormal condition of the dynamic regulator based on the corrected reachable distance comprises the following steps:
Based on the corrected reachable distance, LOF values of each fitting data point are obtained by utilizing a local anomaly factor algorithm;
Taking the fitting data points with the LOF values larger than a preset abnormal threshold value as abnormal data points; and each abnormal data point corresponds to a sampling time and load power failure abnormality occurs to the dynamic voltage regulator at the target fitting number of sampling times after the sampling time.
Further, the initial value of the undetermined number is 99.
The invention has the following beneficial effects:
In the embodiment of the invention, a current data point and a voltage data point at each sampling moment are taken as data points to be measured, the important step of the Douglas-Pocke algorithm in the fitting process of the data points is to obtain data points with changed trend, namely fitting boundary points, the data value distribution of the data points to be measured and a plurality of data points to be measured after the data points to be measured reflects the difference condition between the numerical trend of the data points to be measured and the numerical trend of the data points to be measured before the fitting, the interval between the data points to be measured and the fitting boundary points of the data points to be measured at the corresponding sampling moments under the undetermined quantity shows whether the fitting boundary points of the data points to be measured are close to termination points, and the two factors influence the influence of the undetermined quantity of the data points to be measured after the data points to be measured on the trend of the whole data to be measured, so that the fitting tolerance of the data points to be measured under the undetermined quantity is more accurate; the method comprises the steps that even though a data point to be measured with larger fitting tolerance is in a larger turning but does not affect the whole data, according to whether the fitting tolerance of the data point to be measured at the sampling moment under the undetermined quantity meets the undetermined condition, the target fitting quantity at the sampling moment is obtained, the data point to be measured after the data point to be measured at the sampling moment is fitted out, the whole data trend is not affected, the fitting moment is selected based on the target fitting quantity, after the current data point and the voltage data point are fitted respectively, the current data value and the voltage data value are all available at the same fitting moment, and an abnormal detection space can be constructed to obtain the fitting data point at the fitting moment; and the data points with the same trend and the same data performance characteristic after the fitting are erased, the fitting data points are in an outlier state when the abnormal points of the fitting data points are detected, the distance between the fitting data points does not have a reference value, the distance between the fitting data points is adjusted by utilizing the target fitting quantity of the fitting data points corresponding to the fitting time, under the condition of simplifying the calculation of local outlier factors, errors of the result are avoided, and the load power failure abnormal condition of the dynamic regulator is detected based on the corrected reachable distance.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for intelligently identifying abnormal load power loss of a dynamic voltage regulator according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of specific implementation, structure, characteristics and effects of the intelligent identification method for load power failure abnormality of the dynamic voltage regulator according to the invention, which is provided by the invention, with reference to the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the load power failure abnormality intelligent identification method of the dynamic voltage regulator provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a method flowchart of a method for intelligently identifying load power loss anomalies of a dynamic voltage regulator according to an embodiment of the present invention is shown, where the method includes:
Step S1: acquiring a current data point of a current data value at each sampling time and a voltage data point of a voltage data value at each sampling time of a dynamic voltage regulator in a preset time period; the current data point and the voltage data point are recorded as data points to be measured.
In particular, dynamic voltage regulators are commonly used to control the voltage level and stability of the power supply output, and typically have a digital display screen that can display the magnitude of the output voltage and current so that a user can monitor and adjust the performance of the power supply. Collecting an initial current value and an initial voltage value of the dynamic voltage regulator at each sampling time within a preset time period; because of the large difference between the current and voltage fluctuation ranges of the dynamic voltage regulator, normalization processing is performed on the initial current value and the initial voltage value to obtain a current data value and a voltage data value at each sampling time for convenient operation.
In the embodiment of the invention, the preset time period is one day, the sampling frequency of the initial current value and the initial voltage value is once per second, namely, the time interval between adjacent sampling moments is 1 second, each sampling moment corresponds to the initial current value and the initial voltage value, and an implementer can set the sampling time according to specific conditions; and respectively carrying out normalization processing on the initial current value and the initial voltage value by using a maximum and minimum normalization method.
Acquiring a current data point of a current data value at each sampling time and a voltage data point of a voltage data value at each sampling time of a dynamic voltage regulator in a preset time period; for convenience of the following description, the current data point and the voltage data point are recorded as data points to be measured.
Step S2: acquiring fitting demarcation points of each data point to be measured under the undetermined number; combining the data value of each data point to be measured and a number of data points to be measured after the data points to be measured, and the time interval between the fitting demarcation points of each data point to be measured and the corresponding sampling moments of the data points to be measured under the number to be measured, so as to obtain the fitting tolerance of each data point to be measured under the number to be measured; when the fitting tolerance of the current data point and the voltage data point at each sampling moment respectively under the undetermined quantity does not meet the to-be-selected condition, updating the to-be-selected quantity until the fitting tolerance of the current data point and the voltage data point at each sampling moment respectively under the updated undetermined quantity meets the to-be-selected condition; and when the condition to be selected is met, taking the updated undetermined number as the target fitting number of each sampling moment.
Specifically, because the acquisition frequency of the dynamic voltage regulator is too high, a higher load demand is put forward on the analysis system due to the fact that the data volume is too large when the data anomaly detection is carried out, so that the acquired current data value and the acquired voltage data value are required to be subjected to time sequence data fitting operation by using a Fabry-Perot algorithm, and important data are eliminated from tiny turns. However, in the fitting process, the current data value and the voltage data value do not necessarily have one-to-one correspondence, so that the fitted data may correspond to a null value at the sampling time, that is, only one current data value and only one voltage data value exist at the same sampling time, and therefore, a detection space of an abnormal detection model cannot be constructed. Therefore, the embodiment of the invention judges the fitting tolerance of each data point in the fitting process by improving the Target Laplace-Prak algorithm, and further determines the target fitting quantity of each sampling moment. The douglas-pock algorithm is a well-known technique, and will not be described herein.
(1) And obtaining the fitting tolerance of the data points to be measured under the undetermined quantity.
When the data are fitted by the traditional Douglas-Praeck algorithm, a straight line is constructed for the fitting starting point and the fitting ending point, and the distance from each data point in the range of the starting point and the ending point to the straight line and the size of the limit difference are judged, so that the trend of the data cannot be changed greatly after the fitting is determined. If the current data value and the voltage data value in the scheme are respectively and directly subjected to data fitting by utilizing the dawster-pock algorithm, the situation that the current data value and the voltage data value cannot exist at the same sampling moment may occur.
Therefore, in the embodiment of the invention, fitting judgment is carried out on the current data and the voltage data values at the same sampling time, the fitting tolerance degree of the data points with the largest straight line distance between the data points in the range of the starting point and the ending point and the fitting starting point and the ending point, namely the data points with the changed trend is obtained, and the data points with the changed trend are taken as fitting boundary points, so that the current data values and the voltage data values exist at the same moment after fitting under the condition of discarding a certain trend, and the construction of an anomaly detection model by keeping the change of an important trend is realized. Firstly, acquiring fitting demarcation points of each data point to be measured under the undetermined number:
Preferably, the specific acquisition method of the fitting demarcation point of the data points to be measured under the undetermined quantity is as follows: selecting any one data point to be detected as a characteristic data point, and establishing a two-dimensional coordinate system by taking time as a horizontal axis and a data value of the characteristic data point as a vertical axis; labeling the feature data points and the number of to-be-measured data points after the feature data points in a two-dimensional coordinate system to obtain to-be-measured coordinate points corresponding to the to-be-measured data points; connecting a coordinate point to be measured with the minimum abscissa and a coordinate point to be measured with the maximum abscissa in a two-dimensional coordinate system to obtain a characteristic line segment; and obtaining the distance from each coordinate point to be measured to the characteristic line segment, and taking the data point to be measured of the coordinate point to be measured corresponding to the maximum distance as a fitting demarcation point of the characteristic data point under the undetermined number.
It should be noted that, if the characteristic data point is a current data point, the data value of the characteristic data point is a current data value of the current data point when the two-dimensional coordinate system is established; if the characteristic data point is a voltage data point, the data value of the characteristic data point is the voltage data value of the voltage data point when the two-dimensional coordinate system is established.
The position distribution of the coordinate points to be measured corresponding to the number of data points to be measured in the two-dimensional coordinate system after the characteristic data points and the characteristic data points relative to the characteristic line segments reflects the difference condition between the numerical trend of the data after fitting and the numerical trend of the data before fitting, and the smaller the difference, the smaller the influence of the data fitting on the overall trend of the data, and the larger the tolerance degree; because the trend change of the dagger-pock algorithm approaching the termination point is more similar to the trend change of the termination point during fitting, the influence on the slope of the characteristic straight line is smaller, the time interval between the data point to be measured and the sampling time corresponding to the fitting boundary points of the data point to be measured under the undetermined quantity is presented, whether the fitting boundary points of the data point to be measured approach the termination point is judged, the influence of the fitting boundary points approaching the termination point on the overall trend of the data is smaller, and the tolerance degree is larger. Therefore, the fitting tolerance of each data point to be measured under the undetermined quantity is more accurate by combining the data value of the undetermined quantity of the data points to be measured after each data point to be measured and the time interval between the fitting demarcation point of each data point to be measured under the undetermined quantity and the corresponding sampling time of each data point to be measured under the undetermined quantity.
In the embodiment of the invention, the initial value of the undetermined number takes an empirical value of 99, and an implementer can set the initial value according to specific situations. It should be noted that, if the data point to be measured is a current data point, a preset number of data points to be measured after the data point to be measured are all current data points; if the data point to be measured is a voltage data point, the preset number of data points to be measured after the data point to be measured are all voltage data points.
Preferably, the method for acquiring the fitting tolerance of the data points to be measured under the undetermined quantity comprises the following steps: respectively acquiring a first coordinate point and a second coordinate point from a to-be-measured coordinate point corresponding to a to-be-measured number of to-be-measured data points in a two-dimensional coordinate system after the characteristic data points and the characteristic data points; the ordinate of each first coordinate point is larger than the ordinate of the corresponding point of the first coordinate point on the characteristic line segment, and the ordinate of each second coordinate point is smaller than the ordinate of the corresponding point of the second coordinate point on the characteristic line segment; respectively counting the total number of the first coordinate points to be used as a first value of the feature data points under the undetermined number, and the total number of the second coordinate points to be used as a second value of the feature data points under the undetermined number; the distance between the fitting demarcation point of the characteristic data point under the undetermined quantity and the corresponding coordinate point to be tested and the characteristic line segment in the two-dimensional coordinate system is used as the target distance of the characteristic data point under the undetermined quantity; and combining the time interval between the fitting demarcation points of the characteristic data points and the characteristic data points under the undetermined quantity and the corresponding sampling time, and the first value, the second value and the target distance of the characteristic data points under the undetermined quantity to obtain the fitting tolerance of the characteristic data points under the undetermined quantity.
The fit tolerance of the characteristic data points under the undetermined quantity is calculated as follows:
In the method, in the process of the invention, Fitting tolerance of the feature data points under the undetermined quantity; n is a pending number; a is a preset reference positive number, and the value range of the preset reference positive number a is/>；/>A first value at the pending number for the feature data point; /(I)A second value for the feature data point at the pending number; /(I)Target distance of the feature data points under the undetermined quantity; t is the sampling time corresponding to the characteristic data point; /(I)Sampling time corresponding to fitting demarcation points of the feature data points under the undetermined quantity; /(I)As a function of absolute value; max is a maximum function; exp is an exponential function based on a natural constant e; norms are normalization functions.
It should be noted that the number of the substrates,Presenting the trend distribution of the characteristic data points and the data values of the number of data points to be measured after the characteristic data points. If the feature data points and the data points to be measured under the undetermined quantity are uniformly distributed on the two sides of the feature line segment in the two-dimensional coordinate system corresponding to the coordinate points to be measured, namely/>The closer/>The trend of the data value of the number of data points to be measured after the characteristic data points is similar to the trend of the characteristic line segments, the trend of the data after fitting is consistent with the actual situation, the influence of the data fitting on the overall data trend is smaller, and the greater the tolerance degree of the data fitting is, the greater the data fitting isThe larger. If the coordinate points to be measured are intensively distributed on one side of the characteristic line segment, namely/>And/>The larger the difference between the characteristic data points is, the data value trend of the undetermined number of the data points to be measured after the characteristic data points is different from the trend of the characteristic line segments, the larger the difference exists between the data trend after direct fitting and the actual situation, and the larger the influence of the data fitting on the overall data trend is, the smaller the tolerance degree of the data fitting is, and the number of the data points to be measured is/areThe smaller. In the embodiment of the invention, a reference threshold value is preset to take an experience value/>The implementer can set up by himself according to the specific circumstances.
When (when)Smaller the time,/>The greater the probability of being smaller than the difference of the limits of the Douglas-Pocke algorithm, the smaller the influence of the undetermined number of data points to be measured on the overall data trend after the characteristic data points are removed, the greater the tolerance degree of fitting, and the greater the tolerance degree of the fitting isThe larger. As the trend of the dagger-pock algorithm approaches to the end point in fitting is changed, the influence on the slope of the straight line is smaller; when/>When the distance between the fitting boundary point and the feature straight line is larger, the position of the point farthest from the feature straight line, namely the fitting boundary point, is closer to the ending point of the feature straight line is described, the influence of the fitting boundary point on the trend of the data value of the undetermined number of data points to be measured after the feature data point is smaller, the data trend is consistent with the actual situation after the direct fitting, and the tolerance degree of the fitting is larger, namely/>The larger; by/>The value is reduced.
According to the calculation method of the fitting tolerance of the characteristic data points under the undetermined quantity, the fitting tolerance of each data point to be tested under the undetermined quantity at each sampling time in a preset time period is obtained.
(2) And obtaining the target fitting quantity at the sampling moment.
The data points with larger fitting tolerance can be fitted even in larger turning but have no influence on the whole data, so that the fitting quantity of each sampling moment is determined according to the fitting tolerance of the data points to be measured at each sampling moment under the undetermined quantity.
Preferably, the specific acquisition method of the target fitting quantity is as follows: when the maximum value in the fitting tolerance of the current data point and the voltage data point at each sampling moment under the undetermined quantity is smaller than a preset tolerance threshold, taking the numerical value obtained by subtracting the preset positive integer from the undetermined quantity as the updated undetermined quantity until the maximum value in the fitting tolerance of the current data point and the voltage data point at each sampling moment under the updated undetermined quantity is larger than or equal to the preset tolerance threshold, and taking the updated undetermined quantity as the target fitting quantity at each sampling moment.
It should be noted that, there are two data points to be measured at each sampling time, including a current data point and a voltage data point at each sampling time. In the embodiment of the invention, the preset tolerance threshold takes the empirical value of 0.68, the preset positive integer takes the empirical value of 1, and the implementer can set the preset tolerance threshold according to specific situations.
As an example, taking the ith sampling time as an example for analysis, the initial value of the undetermined number is 99, and the fitting tolerance of the current data point and the voltage data point at the ith sampling time under the undetermined number 99 is sequentiallyAnd/>If/>It is indicated that fitting of 99 data points to be measured after each data point to be measured at the ith sampling moment causes a change in overall trend of the data, namely, data distortion is caused after fitting, so that the data points to be measured at the (i+99) th sampling moment cannot be fitted, namely, the data points to be measured at the sampling moments between the ith sampling moment and the (i+99) th sampling moment cannot be directly removed, that the data points to be measured at the sampling moments between the (i) th sampling moment and the (i+99) th sampling moment have data points with larger turning data values due to an excessive number of data points to be measured, the fitting number of the data points needs to be reduced, the number to be measured is updated, and the updated number to be 98. If/>The fitting of the data points to be measured at the sampling time between the ith sampling time and the (i+99) th sampling time meets the tolerance requirement, the influence of removing the data points to be measured between the ith sampling time and the (i+99) th sampling time on the overall data trend is small, and the undetermined number 99 is used as the target fitting number of the ith sampling time.
When (when)When the fitting tolerance of the current data point and the voltage data point at the ith sampling time under the undetermined quantity 98 is obtained as/>And/>If/>The fitting cannot be performed at the (i+98) th sampling time, and the to-be-measured number is updated to obtain an updated to-be-measured number of 97; if it isThe fitting of the data points to be measured at the sampling time between the ith sampling time and the (i+98) th sampling time meets the tolerance requirement, the fitting can be directly carried out, and the undetermined number 98 is used as the target fitting number of the ith sampling time.
And obtaining the target fitting quantity of each sampling moment in a preset time period according to the method for obtaining the target fitting quantity of the ith sampling moment.
Step S3: and selecting different fitting moments from sampling moments in a preset time period based on the target fitting quantity.
The target fitting quantity of the sampling moment is equivalent to that the current data points and the voltage data points at the sampling moment are respectively consistent with the trend of the data points of the target fitting quantity after the current data points and the voltage data points at the sampling moment, the data points are removed, the overall trend of the data can be ensured under the condition of reducing the data points, and the moment corresponding to the retained data points is determined.
Preferably, the screening method of the fitting time is as follows: starting from a first sampling time in a preset time period, taking the first sampling time as a fitting time, taking the sampling time which is spaced from each fitting time by a target fitting number of sampling times as the next fitting time of each fitting time, and until the number of sampling times after each fitting time is smaller than the target fitting number of fitting times.
As an example, the first sampling time in the preset history period is the fitting time; assuming that 20 sampling moments exist in a preset time period, and the target fitting quantity of the 1 st sampling moment in the preset time period is 4, the 6 th sampling moment is the 2 nd fitting moment; if the target fitting number of the 6 th sampling time is 9, the 16 th sampling time is the 3 rd fitting time. If the target fitting quantity of the 16 th sampling time is 3, the 20 th sampling time is the 4 th fitting time, and the fitting time in the preset time period is sequentially the 1 st sampling time, the 6 th sampling time, the 16 th sampling time and the 20 th sampling time; if the target fitting quantity of the 16 th sampling time is 4, the fitting time in the preset time period is obtained, and the fitting time in the preset time period is sequentially the 1 st sampling time, the 6 th sampling time and the 16 th sampling time.
The current data points at the fitting time in the preset time period are equivalent to the data points obtained by fitting the current data points at all sampling time in the preset time period by utilizing the Target-Puck algorithm; the voltage data points at the fitting time within the preset time period are equivalent to the data points obtained by fitting the voltage data points at all sampling time within the preset time period by utilizing the Target-Puck algorithm; the method and the device can keep the change of the important trend under the condition of discarding a certain trend, and realize the purpose that the current data value and the voltage data value exist at the same moment after fitting.
Step S4: acquiring fitting data points of a current data value and a voltage data value at each fitting moment; and adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points corresponding to the fitting time, and obtaining the corrected reachable distance between the fitting data points.
In the embodiment of the invention, the local anomaly factor algorithm is selected to detect the anomaly points of the fitting data points. The assumption of anomaly detection of the local anomaly factor algorithm is that: the density around the non-outlier object is similar to the density around its neighborhood, while the density around the outlier object is significantly different from the density around its neighborhood. When the space is built for all data points, the data points with the same trend and the same data performance characteristics are gathered together in the anomaly detection space, namely the local densities are larger and similar; however, after the optimized Douglas-Prak algorithm is fitted, the data points with the same trend and the same data performance characteristics are erased, namely a plurality of data points with the same characteristics are reflected to one data point after the fitting; in this case, if conventional local anomaly factors are used to detect outliers, each of the fitted data points is outlier compared to the other fitted data points. Therefore, the embodiment uses the number of data points fitted by each fitted data point, namely the target fitting number of the fitting time corresponding to the fitted data point, as the density of each fitted data point, thereby avoiding the error of the result under the condition of simplifying the calculation of the local outlier factor. The local anomaly factor algorithm is a well known technique for those skilled in the art, and will not be described herein.
The k-th distance neighborhood needs to be defined in LOF anomaly detection, and when the number of fitting data points in anomaly detection space is reduced, the k-th distance neighborhood is defined for each fitting data pointThe range of the distance neighborhood will become larger, so that the traditional k-th reachable distance calculation method is not provided with reference value. The improved method for correcting the reachable distance between the fitting data points in the embodiment of the invention comprises the following steps:
Preferably, the method for obtaining the corrected reachable distance comprises the following steps: selecting any one fitting data point as a core data point, selecting any one data point from a k-th distance neighborhood of the core data point as an example data point, and taking the product of Euclidean distance between the core data point and the example data point and the target fitting quantity of fitting time corresponding to the core data point as a correction distance between the core data point and the example data point; taking the ratio of the kth distance of the core data point to the target fitting quantity of the fitting time corresponding to the example data point as the corrected kth distance of the core data point; the maximum of the corrected distance and the corrected kth distance is taken as the corrected reachable distance between the core data point and the example data point.
It should be noted that, when the core data point o corresponds to the target fitting number of the sampling timeThe larger the core data point o contains, the more data points, the smaller the degree to which the core data point appears to be outlier, and the final reachable distance may be the Euclidean distance/>, between the core data point and the example data pointAnd the kth distance of core data points/>Is selected fromThe greater the likelihood of/>Representing the density of the example data point p, the higher the concentration of the example data point p with surrounding fitted data points, the more likely the reachable distance of each fitted data point is the smaller the respective kth distance, such that the final outlier factor is smaller; target number of fits at sampling instants corresponding to example data point p/>The larger the probability that an illustrative data point is an outlier is less likely, again making it more likely to choose/>The smaller the final outlier factor is made. The kth distance of the data points is a well-known content in the local anomaly factor algorithm, and is not described herein.
Step S5: and detecting the load power failure abnormal condition of the dynamic regulator based on the corrected reachable distance.
Based on the corrected reachable distance, LOF values of each fitting data point are obtained by utilizing a local anomaly factor algorithm; taking the fitting data points with LOF values larger than a preset abnormal threshold value as abnormal data points; and the load power failure abnormality occurs to the dynamic voltage regulator under the condition that the sampling time corresponds to each abnormal data point and the target fitting number is a plurality of sampling times after the sampling time.
It should be noted that, the reachable distance between the fitting data points is replaced by the corrected reachable distance, other contents of the local anomaly factor algorithm are kept unchanged, and the LOF value of each fitting data point is obtained. In the embodiment of the invention, the preset abnormal threshold value takes an empirical value of 1, and an implementer can set the abnormal threshold value according to specific conditions.
The present invention has been completed.
In summary, in the embodiment of the present invention, according to the data values of the data point to be measured and the number of data points to be measured after the data point to be measured, the fitting tolerance of the data point to be measured under the number to be measured is obtained according to the interval between the data points to be measured and the sampling time of the fitting boundary point under the number to be measured, whether the fitting tolerance of the data point to be measured at the sampling time under the number to be measured meets the condition to be measured is judged, the target fitting number of the sampling time is obtained, and then the fitting time is screened out; and adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points at the fitting time to obtain a corrected reachable distance between the fitting data points, and detecting the load power failure abnormal condition of the dynamic regulator based on the corrected reachable distance. The invention screens the fitting time to construct fitting data points, and utilizes the fitting data points to detect the load power failure abnormality of the dynamic voltage regulator.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (9)
1. The intelligent identification method for the load power failure abnormality of the dynamic voltage regulator is characterized by comprising the following steps of:
acquiring a current data point of a current data value at each sampling time and a voltage data point of a voltage data value at each sampling time of a dynamic voltage regulator in a preset time period; recording the current data point and the voltage data point as data points to be measured;
acquiring fitting demarcation points of each data point to be measured under the undetermined number; combining the data value of each data point to be measured and a number of data points to be measured after the data points to be measured, and the time interval between each data point to be measured and the corresponding sampling time of the fitting demarcation point of the data points to be measured under the number to be measured, so as to obtain the fitting tolerance of each data point to be measured under the number to be measured; when the fitting tolerance of the current data point and the voltage data point at each sampling moment under the undetermined quantity respectively does not meet the to-be-selected condition, updating the to-be-selected quantity until the fitting tolerance of the current data point and the voltage data point at each sampling moment under the updated to-be-selected quantity respectively meets the to-be-selected condition; when the condition to be selected is met, taking the updated undetermined number as the target fitting number of each sampling moment;
Selecting different fitting moments from sampling moments in a preset time period based on the target fitting quantity;
Acquiring fitting data points of a current data value and a voltage data value at each fitting moment; adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points corresponding to the fitting time to obtain a corrected reachable distance between the fitting data points;
Detecting the load power failure abnormal condition of the dynamic regulator based on the corrected reachable distance;
the method for acquiring the fitting demarcation point of each data point to be measured under the undetermined number comprises the following steps:
selecting any one data point to be detected as a characteristic data point, and establishing a two-dimensional coordinate system by taking time as a horizontal axis and a data value of the characteristic data point as a vertical axis; labeling a feature data point and a number of data points to be measured after the feature data point in the two-dimensional coordinate system to obtain a coordinate point to be measured corresponding to the data point to be measured;
Connecting a coordinate point to be measured with the minimum abscissa and a coordinate point to be measured with the maximum abscissa in the two-dimensional coordinate system to obtain a characteristic line segment; and obtaining the distance between each coordinate point to be measured and the characteristic line segment, and taking the data point to be measured of the coordinate point to be measured corresponding to the maximum distance as the fitting demarcation point of the characteristic data point under the undetermined quantity.
2. The method for intelligently identifying abnormal load power loss of a dynamic voltage regulator according to claim 1, wherein the method for combining the data value of a pending number of data points after each data point to be tested and the time interval between the corresponding sampling moments of the fitting demarcation point of each data point to be tested and the data point to be tested under the pending number to obtain the fitting tolerance of each data point to be tested under the pending number comprises the following steps:
Respectively acquiring a first coordinate point and a second coordinate point from the feature data point and a plurality of data points to be measured after the feature data point in the two-dimensional coordinate system; the ordinate of each first coordinate point is larger than the ordinate of the corresponding point of the first coordinate point on the characteristic line segment, and the ordinate of each second coordinate point is smaller than the ordinate of the corresponding point of the second coordinate point on the characteristic line segment;
Respectively counting the total number of the first coordinate points to be used as a first value of the feature data points under the undetermined number, and the total number of the second coordinate points to be used as a second value of the feature data points under the undetermined number; taking the distance between the coordinate point to be measured corresponding to the fitting demarcation point of the characteristic data point under the undetermined number and the characteristic line segment in the two-dimensional coordinate system as the target distance of the characteristic data point under the undetermined number;
And combining the time interval between the fitting demarcation points of the characteristic data points and the characteristic data points under the undetermined quantity and the corresponding sampling moments of the fitting demarcation points of the characteristic data points, and obtaining the fitting tolerance of the characteristic data points under the undetermined quantity by the distance between the second value and the target.
3. The intelligent identification method of load power loss abnormality of a dynamic voltage regulator according to claim 2, wherein the fit tolerance of the characteristic data points under a predetermined number is calculated as follows:
; in the/> The fit tolerance for a feature data point at a pending number; n is a pending number; a is a preset reference positive number; /(I)The first value at a pending number for a feature data point; /(I)The second value at the pending number for a feature data point; /(I)The target distance for a feature data point at a pending number; t is the sampling time corresponding to the characteristic data point; /(I)Sampling time corresponding to the fitting demarcation points of the feature data points under the undetermined quantity; /(I)As a function of absolute value; max is a maximum function; exp is an exponential function based on a natural constant e; norms are normalization functions.
4. The intelligent identification method of load power failure abnormality of a dynamic voltage regulator according to claim 1, wherein the fitting tolerance of the current data point and the voltage data point at each sampling time under a predetermined number does not meet a condition to be selected, which is:
and the maximum value of the fitting tolerance of the current data point and the voltage data point at each sampling moment under the undetermined quantity is smaller than a preset tolerance threshold.
5. The intelligent identification method for load power loss abnormality of dynamic voltage regulator according to claim 1, wherein the method for updating the to-be-counted number comprises the steps of:
The value obtained by subtracting the preset positive integer from the undetermined number is used as the undetermined number after updating.
6. The intelligent identification method for load power loss abnormality of a dynamic voltage regulator according to claim 1, wherein the method for selecting different fitting moments from sampling moments of a preset time period comprises the following steps:
Starting from a first sampling time in a preset time period, taking the first sampling time as a fitting time, taking the sampling time which is spaced from each fitting time by the target fitting number of sampling times as the next fitting time of each fitting time, and until the number of sampling times after each fitting time is smaller than the target fitting number of fitting times.
7. The intelligent identification method for load power failure abnormality of a dynamic voltage regulator according to claim 1, wherein the method for adjusting the distance between the fitting data points according to the target fitting quantity of the fitting data points corresponding to the fitting time to obtain the corrected reachable distance between the fitting data points comprises the following steps:
Selecting any one fitting data point as a core data point, selecting any one data point from a k-th distance neighborhood of the core data point as an example data point, and taking the product of Euclidean distance between the core data point and the example data point and the target fitting quantity of the fitting time corresponding to the core data point as a correction distance between the core data point and the example data point; taking the ratio of the kth distance of the core data point to the target fitting quantity of the fitting time corresponding to the example data point as the corrected kth distance of the core data point; the maximum of the modified distance and the modified kth distance is taken as a modified reachable distance between a core data point and an example data point.
8. The intelligent identification method for load power loss abnormality of a dynamic voltage regulator according to claim 1, wherein the method for detecting the load power loss abnormality of the dynamic voltage regulator based on the corrected reachable distance comprises the following steps:
Based on the corrected reachable distance, LOF values of each fitting data point are obtained by utilizing a local anomaly factor algorithm;
Taking the fitting data points with the LOF values larger than a preset abnormal threshold value as abnormal data points; and each abnormal data point corresponds to a sampling time and load power failure abnormality occurs to the dynamic voltage regulator at the target fitting number of sampling times after the sampling time.
9. The intelligent identification method for load loss of power abnormality of dynamic voltage regulator according to claim 1, wherein the initial value of the undetermined number is 99.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410425326.2A CN118033300B (en) | 2024-04-10 | 2024-04-10 | Intelligent identification method for load power failure abnormality of dynamic voltage regulator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410425326.2A CN118033300B (en) | 2024-04-10 | 2024-04-10 | Intelligent identification method for load power failure abnormality of dynamic voltage regulator |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118033300A true CN118033300A (en) | 2024-05-14 |
| CN118033300B CN118033300B (en) | 2024-06-14 |
Family
ID=91002552
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410425326.2A Active CN118033300B (en) | 2024-04-10 | 2024-04-10 | Intelligent identification method for load power failure abnormality of dynamic voltage regulator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118033300B (en) |
Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006060518A2 (en) * | 2004-11-30 | 2006-06-08 | Circumnav Networks, Inc. | Methods for deducing road geometry and connectivity |
| CN101561480A (en) * | 2009-05-22 | 2009-10-21 | 哈尔滨工业大学 | Method for measuring parameter of magnetic characteristic of permanent magnet |
| CN202856358U (en) * | 2012-08-26 | 2013-04-03 | 厦门埃锐圣电力科技有限公司 | Double closed loop dynamic voltage regulator |
| CN104316788A (en) * | 2014-10-16 | 2015-01-28 | 国家电网公司 | Method for judging tripping of outgoing line of double-feed wind turbine generator |
| CN104377702A (en) * | 2014-11-19 | 2015-02-25 | 上海电力学院 | Reversible adjustment dynamic voltage adjuster |
| CN110659267A (en) * | 2019-07-30 | 2020-01-07 | 浙江华云信息科技有限公司 | Douglas-Puke-based current three-phase imbalance curve simplification method |
| WO2020019403A1 (en) * | 2018-07-26 | 2020-01-30 | 平安科技（深圳）有限公司 | Electricity consumption abnormality detection method, apparatus and device, and readable storage medium |
| CN111555311A (en) * | 2020-05-20 | 2020-08-18 | 国网陕西省电力公司电力科学研究院 | Plug and play stability analysis and control method for power system |
| CN111562467A (en) * | 2020-03-24 | 2020-08-21 | 中国电力科学研究院有限公司 | Halo-starting judgment method and system based on ground synthetic electric field measurement data |
| KR20210014969A (en) * | 2019-07-31 | 2021-02-10 | 현대엠엔소프트 주식회사 | Road slop information modeling apparatus and method based on ramer-douglas-peuker algorithm |
| DE102019124498A1 (en) * | 2019-09-12 | 2021-03-18 | Valeo Schalter Und Sensoren Gmbh | Method for compressing 3D point data in a 3D point cloud generated by means of a detection system of a vehicle |
| CN116054166A (en) * | 2023-02-07 | 2023-05-02 | 湖南大学 | Power grid inertia coefficient distribution identification method based on disturbance measurement data |
| CN116228921A (en) * | 2023-02-15 | 2023-06-06 | 合肥栈顶信息科技有限公司 | Graph fitting algorithm based on Douglas-Puck algorithm |
| CN116930771A (en) * | 2023-07-31 | 2023-10-24 | 上海理工大学 | Real vehicle hidden electricity quantity estimation method based on DTW |
| CN116957634A (en) * | 2023-09-19 | 2023-10-27 | 贵昌集团有限公司 | Information intelligent acquisition processing method for electronic commerce platform |
| CN117112963A (en) * | 2023-08-11 | 2023-11-24 | 河南大学 | Improved method of maximum depth-based Target Laplace-Praekey algorithm |
| CN117235972A (en) * | 2023-08-07 | 2023-12-15 | 中国大唐集团科学技术研究总院有限公司华东电力试验研究院 | Flow curve piecewise fitting method and system based on secondary thinning technology |
| CN117521905A (en) * | 2023-11-16 | 2024-02-06 | 深圳海辰储能科技有限公司 | Power generation equipment screening method and related device |
| CN117591964A (en) * | 2024-01-12 | 2024-02-23 | 山西思极科技有限公司 | Electric power intelligent analysis method based on artificial intelligence |
-
2024
- 2024-04-10 CN CN202410425326.2A patent/CN118033300B/en active Active
Patent Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006060518A2 (en) * | 2004-11-30 | 2006-06-08 | Circumnav Networks, Inc. | Methods for deducing road geometry and connectivity |
| CN101561480A (en) * | 2009-05-22 | 2009-10-21 | 哈尔滨工业大学 | Method for measuring parameter of magnetic characteristic of permanent magnet |
| CN202856358U (en) * | 2012-08-26 | 2013-04-03 | 厦门埃锐圣电力科技有限公司 | Double closed loop dynamic voltage regulator |
| CN104316788A (en) * | 2014-10-16 | 2015-01-28 | 国家电网公司 | Method for judging tripping of outgoing line of double-feed wind turbine generator |
| CN104377702A (en) * | 2014-11-19 | 2015-02-25 | 上海电力学院 | Reversible adjustment dynamic voltage adjuster |
| WO2020019403A1 (en) * | 2018-07-26 | 2020-01-30 | 平安科技（深圳）有限公司 | Electricity consumption abnormality detection method, apparatus and device, and readable storage medium |
| CN110659267A (en) * | 2019-07-30 | 2020-01-07 | 浙江华云信息科技有限公司 | Douglas-Puke-based current three-phase imbalance curve simplification method |
| KR20210014969A (en) * | 2019-07-31 | 2021-02-10 | 현대엠엔소프트 주식회사 | Road slop information modeling apparatus and method based on ramer-douglas-peuker algorithm |
| DE102019124498A1 (en) * | 2019-09-12 | 2021-03-18 | Valeo Schalter Und Sensoren Gmbh | Method for compressing 3D point data in a 3D point cloud generated by means of a detection system of a vehicle |
| CN111562467A (en) * | 2020-03-24 | 2020-08-21 | 中国电力科学研究院有限公司 | Halo-starting judgment method and system based on ground synthetic electric field measurement data |
| CN111555311A (en) * | 2020-05-20 | 2020-08-18 | 国网陕西省电力公司电力科学研究院 | Plug and play stability analysis and control method for power system |
| CN116054166A (en) * | 2023-02-07 | 2023-05-02 | 湖南大学 | Power grid inertia coefficient distribution identification method based on disturbance measurement data |
| CN116228921A (en) * | 2023-02-15 | 2023-06-06 | 合肥栈顶信息科技有限公司 | Graph fitting algorithm based on Douglas-Puck algorithm |
| CN116930771A (en) * | 2023-07-31 | 2023-10-24 | 上海理工大学 | Real vehicle hidden electricity quantity estimation method based on DTW |
| CN117235972A (en) * | 2023-08-07 | 2023-12-15 | 中国大唐集团科学技术研究总院有限公司华东电力试验研究院 | Flow curve piecewise fitting method and system based on secondary thinning technology |
| CN117112963A (en) * | 2023-08-11 | 2023-11-24 | 河南大学 | Improved method of maximum depth-based Target Laplace-Praekey algorithm |
| CN116957634A (en) * | 2023-09-19 | 2023-10-27 | 贵昌集团有限公司 | Information intelligent acquisition processing method for electronic commerce platform |
| CN117521905A (en) * | 2023-11-16 | 2024-02-06 | 深圳海辰储能科技有限公司 | Power generation equipment screening method and related device |
| CN117591964A (en) * | 2024-01-12 | 2024-02-23 | 山西思极科技有限公司 | Electric power intelligent analysis method based on artificial intelligence |
Non-Patent Citations (1)
| Title |
|---|
| 谢永明: "动态电压调节器的控制策略研究", 《中国电力教育》, 31 December 2010 (2010-12-31), pages 432 - 434 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118033300B (en) | 2024-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2021235867B2 (en) | Method and apparatus for determining operating state of photovoltaic array, device and storage medium | |
| CN117556279B (en) | Method and system for monitoring running state of spin coater based on electrical parameter analysis | |
| CN116523913B (en) | Intelligent detection method for quality of screw rod | |
| CN118091234B (en) | Current transformer for fault diagnosis processing | |
| CN117034177B (en) | Intelligent monitoring method for abnormal data of power load | |
| CN117633695B (en) | Transformer operation monitoring method based on electrical parameter time sequence analysis | |
| CN118033300B (en) | Intelligent identification method for load power failure abnormality of dynamic voltage regulator | |
| CN116363984A (en) | Display screen flicker correction method and correction system | |
| CN118424672A (en) | Transformer mechanical defect monitoring method and device with self-adaptive threshold adjusting function | |
| CN112126907B (en) | Vacuum coating control system and control method thereof, and vacuum coating equipment | |
| CN118034069A (en) | Self-adaptive control method for running state of intelligent controller | |
| CN118034417A (en) | Plant introduction and conservation environment monitoring management system | |
| CN115909932A (en) | Method and device for determining detection result, electronic equipment and storage medium | |
| CN118465521B (en) | Operation abnormality detection method based on three-loop control digital current-sharing parallel circuit | |
| CN115907567B (en) | Load event detection method and system based on robust random segmentation forest algorithm | |
| CN114877998B (en) | Infrared image blind pixel processing method and thermal infrared imager | |
| CN113077357B (en) | Power time sequence data anomaly detection method and filling method thereof | |
| CN116821834B (en) | Vacuum circuit breaker overhauling management system based on embedded sensor | |
| CN113704321B (en) | Power grid abnormal data identification method, device and system | |
| CN118094264B (en) | Intelligent power capacitor partial discharge detection method and system | |
| CN116819426A (en) | Ammeter fault detection method, device, equipment, medium, equipment and product | |
| CN111833287A (en) | System for judging defects of display panel by applying deep learning and judging method thereof | |
| WO2023278102A1 (en) | Service location anomalies | |
| CN118608194A (en) | Power data collection method, system and storage medium | |
| CN118707349A (en) | Battery abnormality detection method, electronic device, storage medium, and program product |
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 |