CN117686757B - Intelligent early warning method and system for outdoor power metering box - Google Patents
Intelligent early warning method and system for outdoor power metering box Download PDFInfo
- Publication number
- CN117686757B CN117686757B CN202311672942.XA CN202311672942A CN117686757B CN 117686757 B CN117686757 B CN 117686757B CN 202311672942 A CN202311672942 A CN 202311672942A CN 117686757 B CN117686757 B CN 117686757B
- Authority
- CN
- China
- Prior art keywords
- false touch
- data set
- feature
- probability
- matrix
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000002159 abnormal effect Effects 0.000 claims abstract description 86
- 238000000605 extraction Methods 0.000 claims abstract description 24
- 238000002372 labelling Methods 0.000 claims abstract description 24
- 238000012544 monitoring process Methods 0.000 claims abstract description 23
- 230000001960 triggered effect Effects 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 78
- 238000003909 pattern recognition Methods 0.000 claims description 49
- 238000004364 calculation method Methods 0.000 claims description 28
- 238000004458 analytical method Methods 0.000 claims description 16
- 230000005856 abnormality Effects 0.000 claims description 4
- 230000003213 activating effect Effects 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
Landscapes
- Testing And Monitoring For Control Systems (AREA)
- Emergency Alarm Devices (AREA)
Abstract
The invention provides an intelligent early warning method and system for an outdoor power metering box, which relate to the technical field of data processing, identify historical overheat protection monitoring data, generate a power protection triggered sample data set and label based on an overheat protection mode, obtain a labeling sample data set, perform feature extraction to obtain an overheat feature set, if an overheat protection instruction is triggered to obtain a real-time abnormal data set, input an overheat mode identification model to calculate first miss-touch probability, and generate first early warning information when the first miss-touch probability reaches a preset miss-touch probability, so that the technical problems that the judging mode of overheat protection miss-touch in the prior art is technically limited and not strict enough, the judging accuracy is insufficient, and the operation state of equipment is influenced are solved.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to an intelligent early warning method and system of an outdoor power metering box.
Background
The power metering box is used as basic equipment for electric energy metering, is widely applied to a power system, is influenced by factors such as overload current, internal and external environments and the like in the operation process of the power metering box, can cause excessive temperature to trigger overheat protection, and is easier to generate false triggering of overheat protection due to outdoor high instability. At present, overheat protection false touch early warning of the power metering box is mainly carried out in modes of external equipment monitoring and the like, and basic defects exist.
In the prior art, the judging mode of overheat protection false touch has technical limitation and is not strict enough, so that the judging accuracy is insufficient, and the operation and running state of equipment are affected.
Disclosure of Invention
The application provides an intelligent early warning method and system for an outdoor power metering box, which are used for solving the technical problems that in the prior art, the judgment mode for overheat protection false touch is technically limited and not strict enough, so that the judgment accuracy is insufficient and the operation and running states of equipment are influenced.
In view of the above problems, the application provides an intelligent early warning method and system for an outdoor power metering box.
In a first aspect, the application provides an intelligent early warning method of an outdoor power metering box, which comprises the following steps:
acquiring historical overheat protection monitoring data of a first power metering box;
identifying the historical overheat protection monitoring data to generate a sample data set triggered by power protection;
Labeling the sample data set to obtain a labeled sample data set, wherein the labeled sample data set is a sample data set in a false electric shock protection mode;
Extracting features of the labeling sample data set to obtain a false touch feature set;
when the first power metering box triggers an overheat protection instruction, acquiring a real-time abnormal data set of the first power metering box;
establishing a false touch pattern recognition model by the false touch feature set, inputting the real-time abnormal data set into the false touch pattern recognition model to calculate false touch probability, and acquiring a first false touch probability;
and when the first false touch probability reaches a preset false touch probability, generating first early warning information.
In a second aspect, the present application provides an intelligent early warning system for an outdoor power metering box, the system comprising:
the data acquisition module is used for acquiring historical overheat protection monitoring data of the first power metering box;
the data identification module is used for identifying the historical overheat protection monitoring data and generating a sample data set triggered by power protection;
The data labeling module is used for labeling the sample data set to obtain a labeled sample data set, wherein the labeled sample data set is a sample data set in an electric shock protection mode;
The feature extraction module is used for extracting features of the labeling sample data set to obtain a false touch feature set;
The triggering acquisition module is used for acquiring a real-time abnormal data set of the first power metering box when the first power metering box triggers an overheat protection instruction;
The false touch probability calculation module is used for establishing a false touch pattern recognition model by the false touch feature set, inputting the real-time abnormal data set into the false touch pattern recognition model to perform false touch probability calculation, and acquiring a first false touch probability;
the early warning information generation module is used for generating first early warning information when the first false touch probability reaches a preset false touch probability.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
According to the intelligent early warning method for the outdoor power metering box, historical overheat protection monitoring data of a first power metering box are obtained, a sample data set triggered by power protection is generated through data identification, the sample data set is marked, a marked sample data set is obtained, the marked sample data set is a sample data set in an electric shock error protection mode, feature extraction is carried out on the marked sample data set, an error touch feature set is obtained, and when the first power metering box triggers an overheat protection instruction, a real-time abnormal data set is obtained; the false touch feature set is used for establishing a false touch pattern recognition model, false touch probability calculation is carried out on the input real-time abnormal data set, first false touch probability is obtained, when the first false touch probability reaches preset false touch probability, first early warning information is generated, the technical problem that the judging mode of overheat protection false touch in the prior art is technically limited and not strict enough, judgment accuracy is insufficient, the operation running state of equipment is affected is solved, the false touch feature recognition matching and comprehensive evaluation are carried out according to real-time abnormal data, analysis accuracy and processing efficiency are guaranteed, false touch probability is accurately determined, and invalid protection operation is avoided.
Drawings
FIG. 1 is a schematic flow diagram of an intelligent early warning method of an outdoor power metering box;
FIG. 2 is a schematic diagram of an activation flow of a false touch pattern recognition model in an intelligent early warning method of an outdoor power metering box;
fig. 3 is a schematic diagram of a first false touch probability obtaining flow chart in an intelligent early warning method of an outdoor power metering box;
fig. 4 is a schematic structural diagram of an intelligent early warning system of an outdoor power metering box.
Reference numerals illustrate: the system comprises a data acquisition module 11, a data identification module 12, a data labeling module 13, a feature extraction module 14, a trigger acquisition module 15, a false touch probability calculation module 16 and an early warning information generation module 17.
Detailed Description
According to the intelligent early warning method and system for the outdoor power metering box, historical overheat protection monitoring data are identified, a sample data set triggered by power protection is generated and marked based on an overheat protection mode, a marked sample data set is obtained, feature extraction is carried out to obtain an overheat feature set, if an overheat protection instruction is triggered to obtain a real-time abnormal data set, a first miss-touch probability is calculated by inputting an overheat pattern recognition model, and when the first miss-touch probability reaches a preset miss-touch probability, first early warning information is generated, so that the technical problem that the overheat protection miss-touch judgment mode in the prior art is technically limited and not strict enough, and the judgment accuracy is insufficient and the operation running state of equipment is influenced is solved.
Example 1
As shown in fig. 1, the application provides an intelligent early warning method of an outdoor power metering box, which comprises the following steps:
Step S100: acquiring historical overheat protection monitoring data of a first power metering box;
Specifically, the electric power metering box is used as basic equipment for electric energy metering, is widely applied to an electric power system, is influenced by factors such as overload current, internal and external environments and the like in the operation process of the electric power metering box, can cause excessive temperature to trigger overheat protection, and is easier to generate false triggering of overheat protection due to outdoor high instability. According to the intelligent early warning method for the outdoor power metering box, disclosed by the application, the collected trigger sample is combined to conduct false touch feature extraction, modeling is conducted to conduct false touch probability analysis of real-time abnormal data, and overheat protection trigger starting is conducted accurately, so that the influence of abnormal triggering on the running state of the power metering box is avoided.
Specifically, the first power metering box is a target power metering box for overheat protection false touch judgment, the first power metering box is subjected to retrieval and calling of a historical monitoring record triggered by overheat protection based on a preset time interval, namely, a data retrieval time period bordering on a current time node, time sequence integration is performed on the call data, and the call data is used as historical overheat protection monitoring data of the first power metering box, wherein the historical overheat protection monitoring data is a reference data source for overheat protection false touch judgment.
Step S200: identifying the historical overheat protection monitoring data to generate a sample data set triggered by power protection;
Step S300: labeling the sample data set to obtain a labeled sample data set, wherein the labeled sample data set is a sample data set in a false electric shock protection mode;
Specifically, the historical overheat protection monitoring data are identified, data extraction association and integration are performed, trigger factors, trigger position, measurement values, trigger duration, trigger modes, such as overload current, winding, specific current values, overheat protection enabling duration, reverse blocking triggering and the like, of each piece of monitoring data are determined, the historical overheat monitoring data are extracted regularly based on the data sequences, a plurality of data sequences are determined and used as the sample data set of the power protection triggering, and false touch judgment identification is further performed on the sample data set.
Specifically, the sample data set includes valid trigger data and false touch data, the sample data set is classified and attributed, for example, false touch judgment is performed on the sample data set based on an expert group, classification is performed based on a judgment result, the false touch data attributed class in the sample data set is identified, specific identification information is customized and set for data identification and distinction, preferably, false touch data aiming at different trigger types and the like can be marked based on different identification information, for example, different colors, different codes and the like, so that visual distinction is facilitated. And labeling and integrating the sample data set belonging to the false electric shock protection mode, and further performing false touch feature recognition and extraction as the labeled sample data set.
Step S400: extracting features of the labeling sample data set to obtain a false touch feature set;
step S500: when the first power metering box triggers an overheat protection instruction, acquiring a real-time abnormal data set of the first power metering box;
Specifically, the labeling sample data set is source data to be subjected to false touch analysis, and false touch characteristics of each labeling sample data, such as external interference, for example electromagnetic interference, environmental factor interference and the like, are analyzed based on the labeling sample data set; power supply effects such as contact anomalies, etc.; component quality problems and the like, the same false touch event can be caused by single factor or multi-factor cooperation, false touch tracing is performed by combining operation records, specific characteristics causing overheat protection false touch are determined, false touch characteristics of each sample data are determined to be integrated regularly, and the false touch characteristics are used as the false touch characteristics set which is a standard for false touch judgment.
And carrying out operation monitoring on the first power metering box, if the first power metering box triggers the overheat protection instruction, carrying out false touch detection judgment before instruction execution is carried out, and determining whether to execute the overheat protection instruction based on a false touch detection result so as to avoid that false touch operation affects normal operation of equipment. When the overheat protection instruction is triggered, the abnormal operation data of the first power metering box are monitored, data acquisition is carried out to obtain the real-time abnormal data set, the real-time abnormal data is a data source for triggering overheat protection, and false touch judgment is needed.
Step S600: establishing a false touch pattern recognition model by the false touch feature set, inputting the real-time abnormal data set into the false touch pattern recognition model to calculate false touch probability, and acquiring a first false touch probability;
step S700: and when the first false touch probability reaches a preset false touch probability, generating first early warning information.
Specifically, the false touch pattern recognition model is built, the real-time abnormal data set is input into the false touch pattern recognition model, the data processing is gradually executed based on a feature extraction layer, a feature matching layer and a probability output layer in the model, and the false touch probability representing overheat protection, namely the first false touch probability, is determined. The higher the first false touch probability is, the lower the feasibility of overheat protection operation is, and the overheat protection operation belongs to invalid operation; the higher the first false touch probability, the more necessary the execution of overheat protection is indicated. Acquiring the preset false touch probability, namely, the critical false touch probability for false touch judgment determined according to the operation and maintenance standard of the first power metering box, checking the first false touch probability and the preset false touch probability, and if the first false touch probability reaches the preset false touch probability, namely, the false touch probability is higher, cutting off overheat protection operation to be executed, generating first early warning information, and carrying out adjustment control based on the warning information.
Further, as shown in fig. 2, step S600 of the present application further includes:
Step S610-1: carrying out abnormality accompanying risk identification on the real-time abnormality data set to obtain a first accompanying index;
step S620-1: judging whether the first accompanying index is smaller than or equal to a preset accompanying index;
Step S630-1: and when the first accompanying index is smaller than or equal to the preset accompanying index, generating a false touch judging instruction, and activating the false touch pattern recognition model according to the false touch judging instruction.
Further, the step S630-1 of the present application further comprises:
Step S631-1: classifying the marked sample data set according to the type of the error electric shock protection mode, and outputting a marked sample classification result;
step S632-1: extracting features according to the labeling sample classification result to obtain multiple classes of false touch feature sets;
Step S633-1: and establishing a multi-type false touch mode identification model according to the multi-type false touch feature set, and carrying out probability calculation under a corresponding false touch mode by using the multi-type false touch mode identification model.
Further, the step S633-1 of the present application further comprises:
step S6331-1: analyzing the real-time abnormal data set to obtain a first matching type;
step S6332-1: performing pattern matching on the multiple types of false touch pattern recognition models according to the first matching type to obtain a first matching recognition pattern;
Step S6333-1: and carrying out false touch probability calculation according to the first matching recognition mode to obtain a first false touch probability.
Specifically, the real-time abnormal data set is subjected to abnormality accompanying risk recognition, that is, the risk level of the current value, for example, the real-time abnormal data set may be subjected to judgment recognition by analogy with the reference data which is similar to and different from the real-time abnormal data set, the accompanying risk corresponding to the reference data is determined, the accompanying index corresponding to the reference data is determined, and fine adjustment of the accompanying index of the reference data is performed based on the recognition accuracy, as the first accompanying index, that is, the index for measuring the risk level of the real-time abnormal data set. And acquiring the preset accompanying index, namely, combining a numerical value which is self-defined and set by expert experience and is used for measuring critical accompanying dangers of abnormal data, checking the first accompanying index and the preset accompanying index, if the first accompanying index is smaller than or equal to the preset accompanying index, indicating that triggering execution under the critical accompanying dangers is performed, further performing false touch judgment analysis, generating the false touch judgment instruction, namely, performing a starting instruction of false touch analysis, and activating the false touch pattern recognition model along with receiving of the false touch judgment instruction.
Specifically, the false touch pattern recognition model is built, namely a self-built auxiliary model for false touch probability analysis. The type of the error electric shock protection mode, namely, the triggered protection modes of various types, such as open circuit protection, equipment self protection, regulation protection, overcurrent protection, short circuit protection and the like, are obtained, and equipment overheat protection is executed based on different operations. And taking the type of the electric error power protection mode as a division standard, carrying out type division attribution on the marked sample data set, and taking the marked sample classification result as the marked sample classification result, wherein the marked sample classification result corresponds to the type of the electric error power protection mode one by one. And further, extracting features of the labeling sample classification result, namely attributing classification results to the false touch feature set obtained in the earlier stage of the embodiment, and obtaining the multiple classes of false touch feature sets corresponding to the labeling sample classification result.
Further, the multiple types of false touch feature sets are respectively embedded into the false touch analysis spaces correspondingly configured to generate multiple false touch pattern recognition units, the multiple false touch pattern recognition units are generated by training a neural network based on sample data, and are parallel execution units, each false touch pattern recognition unit respectively comprises a feature extraction layer, a feature matching layer and a probability output layer, specific training modes are the same and training samples are different, the multiple false touch pattern recognition units are integrated to form the multiple types of false touch pattern recognition models, the multiple types of false touch pattern recognition models are input into the adaptive false touch pattern recognition units for targeted recognition analysis through input attribution analysis, probability calculation under the corresponding false touch pattern is performed, and calculation efficiency is improved on the basis of guaranteeing output precision.
Further, performing type analysis of a false touch mode on the real-time abnormal data set to serve as the first matching type. And performing mode matching on the multiple types of false touch pattern recognition models based on the first matching type, namely determining a model execution mode corresponding to the first matching type as the first matching recognition mode, wherein the first matching recognition mode corresponds to one false touch pattern recognition unit in the multiple types of false touch pattern recognition models, namely a suitability processing unit for performing false touch analysis on the real-time abnormal data set, and performing false touch probability calculation on the real-time abnormal data set based on the suitability processing unit to serve as the first false touch probability, wherein the first false touch probability is a measurement value for representing the false touch probability and has actual fitness and high accuracy.
Further, as shown in fig. 3, the real-time abnormal data set is input into the false touch pattern recognition model to perform false touch probability calculation, so as to obtain a first false touch probability, and step S600 of the present application further includes:
step S610-2: inputting the real-time abnormal data set into the false touch pattern recognition model, wherein the false touch pattern recognition model comprises a feature extraction layer, a feature matching layer and a probability output layer;
Step S620-2: performing feature extraction on the real-time abnormal data set according to the feature extraction layer, and outputting an abnormal feature set;
Step S630-2: matching the abnormal feature set with the false touch feature set according to the feature matching layer to obtain a feature matching matrix;
step S640-2: and carrying out probability calculation on the feature matching matrix according to the probability output layer, and outputting a first false touch probability.
Further, the step S630-2 of the present application further includes matching the abnormal feature set with the false touch feature set according to the feature matching layer to obtain a feature matching matrix:
step S631-2: generating an abnormal initial matrix according to the abnormal feature set;
Step S632-2: generating a false touch comparison matrix by using the false touch feature set;
Step S633-2: and identifying the false touch comparison matrix according to the abnormal initial matrix, obtaining a coefficient matrix of the coincident characteristic, and outputting the coefficient matrix of the coincident characteristic as the characteristic matching matrix.
Further, the probability calculation is performed on the feature matching matrix according to the probability output layer, and the step S640-2 of the present application further includes:
Step S641-2: disassembling the feature matching matrix, and outputting a first disassembly matrix based on the number of features and a second disassembly matrix based on feature errors;
Step S642-2: and carrying out probability factorization on the first dismantling matrix and the second dismantling matrix, and outputting a first false touch probability.
Specifically, the false touch probability analysis of the real-time abnormal data is performed based on the false touch pattern recognition model. And inputting the real-time abnormal data set into the false touch pattern recognition model, and determining a false touch pattern recognition unit matched with the real-time abnormal data set by pattern matching, wherein the false touch pattern recognition unit comprises the feature extraction layer, the feature matching layer and the probability output layer. Inputting the real-time abnormal data set into the feature extraction layer, and extracting false touch features, such as environmental features, such as over-high environmental temperature, of the real-time abnormal data set; and the power supply characteristic, abnormal on-off and the like, namely, the factor causing false touch, are taken as the abnormal characteristic set. And inputting the abnormal feature set into the feature matching layer to perform feature matching analysis, and generating the feature matching matrix.
Specifically, based on the abnormal feature set, the abnormal feature is taken as a matrix row, and feature values corresponding to the abnormal features are taken as a matrix column, so that the abnormal feature set is converted, and the abnormal initial matrix is generated. The false touch feature set is embedded in the feature matching layer and is a completeness false touch feature corresponding to the current mode, the completeness false touch feature set is used for reference judgment, the false touch feature is taken as a matrix row, feature values corresponding to the false touch features are taken as a matrix array, the false touch feature set is converted, and the false touch comparison matrix is generated. And further performing overlapping correction on the abnormal initial matrix and the false touch comparison matrix, determining a coincidence coefficient, namely identifying the abnormal initial matrix in the false touch comparison matrix, judging the abnormal initial matrix as a coincidence characteristic, configuring the coincidence coefficient, for example, taking 1 as the coincidence coefficient of a matched coincidence characteristic, taking 0 as the coincidence coefficient of a non-coincidence characteristic, generating the coefficient matrix, namely taking the abnormal characteristic set as a matrix row, taking the false touch characteristic set as a matrix column, taking 1 and 0 as coefficient matrixes of intermediate characterization coefficients, and outputting the coefficient matrix of the coincidence characteristic as the characteristic matching matrix.
And further, inputting the feature matching matrix into the probability output layer for false touch probability analysis. Specifically, identifying the coincidence coefficient of the feature matching matrix, extracting matrix items with the coincidence coefficient of 1, and combining the matrix items as the feature number to form a first disassembly matrix for characterizing the feature number; and aiming at the first disassembly matrix, carrying out feature difference value calculation of the abnormal features and the false touch features corresponding to mapping, and constructing the second disassembly matrix taking the abnormal features and the false touch features as matrix rows and columns and taking the feature difference value as an intermediate item, wherein the feature difference value is used as a feature error. And traversing the second disassembly matrix as a factorial term based on the feature errors by taking the feature number in the first disassembly matrix as the factorial term, and determining the false touch probability based on the feature errors, wherein the feature errors are inversely related to the corresponding false touch probability, and taking a calculation result as the first false touch probability determined based on feature fusion by probability factorization, wherein the higher the false touch probability is, the more invalid triggering is indicated, and the lower the executable probability of overheat protection is. The first false touch probability is the false touch probability determined based on the completeness existence characteristic, and has actual fitness and measurement accuracy.
Example two
Based on the same inventive concept as the intelligent early warning method of an outdoor power metering box in the foregoing embodiment, as shown in fig. 4, the present application provides an intelligent early warning system of an outdoor power metering box, where the system includes:
The data acquisition module 11 is used for acquiring historical overheat protection monitoring data of the first power metering box;
the data identification module 12 is used for identifying the historical overheat protection monitoring data and generating a sample data set triggered by power protection;
The data labeling module 13 is configured to label the sample data set to obtain a labeled sample data set, where the labeled sample data set is a sample data set in a protection mode of error electric power;
The feature extraction module 14 is configured to obtain a false touch feature set by performing feature extraction on the labeled sample data set by the feature extraction module 14;
The trigger acquisition module 15 is used for acquiring a real-time abnormal data set of the first power metering box when the first power metering box triggers an overheat protection instruction;
The false touch probability calculation module 16, where the false touch probability calculation module 16 is configured to establish a false touch pattern recognition model with the false touch feature set, input the real-time abnormal data set into the false touch pattern recognition model to perform false touch probability calculation, and obtain a first false touch probability;
the early warning information generation module 17 is configured to generate first early warning information when the first false touch probability reaches a preset false touch probability.
Further, the system further comprises:
the accompanying index acquisition module is used for identifying abnormal accompanying dangers of the real-time abnormal data set and acquiring a first accompanying index;
The accompanying index judging module is used for judging whether the first accompanying index is smaller than or equal to a preset accompanying index or not;
The command activation module is used for generating a false touch discrimination command when the first accompanying index is smaller than or equal to the preset accompanying index, and activating the false touch pattern recognition model according to the false touch discrimination command.
Further, the system further comprises:
the data classification module is used for classifying the marked sample data set according to the type of the error electric power protection mode and outputting a marked sample classification result;
The feature set acquisition module is used for extracting features according to the classification result of the labeling sample to acquire multiple types of false touch feature sets;
The model construction and calculation module is used for establishing a multi-class false touch mode identification model according to the multi-class false touch feature set, and carrying out probability calculation under a corresponding false touch mode by using the multi-class false touch mode identification model.
Further, the system further comprises:
The first matching type acquisition module is used for analyzing the real-time abnormal data set to acquire a first matching type;
The pattern matching module is used for performing pattern matching on the multi-class false touch pattern recognition model according to the first matching type to obtain a first matching recognition pattern;
and the probability calculation module is used for carrying out false touch probability calculation according to the first matching recognition mode to obtain first false touch probability.
Further, the system further comprises:
The data input module is used for inputting the real-time abnormal data set into the false touch pattern recognition model, wherein the false touch pattern recognition model comprises a feature extraction layer, a feature matching layer and a probability output layer;
the abnormal feature set output module is used for carrying out feature extraction on the real-time abnormal data set according to the feature extraction layer and outputting an abnormal feature set;
The feature matching matrix acquisition module is used for matching the abnormal feature set with the false touch feature set according to the feature matching layer to acquire a feature matching matrix;
The first false touch probability output module is used for carrying out probability calculation on the feature matching matrix according to the probability output layer and outputting first false touch probability.
Further, the system further comprises:
The abnormal initial matrix generation module is used for generating an abnormal initial matrix according to the abnormal characteristic set;
The false touch comparison matrix generation module is used for generating a false touch comparison matrix by the false touch feature set;
the matrix identification module is used for identifying the false touch comparison matrix according to the abnormal initial matrix, acquiring a coefficient matrix of the coincident characteristic, and outputting the coefficient matrix of the coincident characteristic as the characteristic matching matrix.
Further, the system further comprises:
the matrix disassembling module is used for disassembling the feature matching matrix and outputting a first disassembling matrix based on the number of features and a second disassembling matrix based on the feature errors;
and the probability factorization module is used for performing probability factorization on the first disassembly matrix and the second disassembly matrix and outputting first false touch probability.
Through the foregoing detailed description of an intelligent early warning method of an outdoor power metering box, those skilled in the art can clearly know an intelligent early warning method and an intelligent early warning system of an outdoor power metering box in this embodiment, and for the device disclosed in the embodiment, the description is relatively simple because the device corresponds to the method disclosed in the embodiment, and relevant places refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (5)
1. An intelligent early warning method of an outdoor power metering box is characterized by comprising the following steps:
acquiring historical overheat protection monitoring data of a first power metering box;
identifying the historical overheat protection monitoring data to generate a sample data set triggered by power protection;
Labeling the sample data set to obtain a labeled sample data set, wherein the labeled sample data set is a sample data set in a false electric shock protection mode;
Extracting features of the labeling sample data set to obtain a false touch feature set;
when the first power metering box triggers an overheat protection instruction, acquiring a real-time abnormal data set of the first power metering box;
establishing a false touch pattern recognition model by the false touch feature set, inputting the real-time abnormal data set into the false touch pattern recognition model to calculate false touch probability, and acquiring a first false touch probability;
when the first false touch probability reaches a preset false touch probability, generating first early warning information;
The real-time abnormal data set is input into the false touch pattern recognition model to calculate false touch probability, and a first false touch probability is obtained, and the method further comprises the following steps:
Inputting the real-time abnormal data set into the false touch pattern recognition model, wherein the false touch pattern recognition model comprises a feature extraction layer, a feature matching layer and a probability output layer;
Performing feature extraction on the real-time abnormal data set according to the feature extraction layer, and outputting an abnormal feature set;
Matching the abnormal feature set with the false touch feature set according to the feature matching layer to obtain a feature matching matrix;
performing probability calculation on the feature matching matrix according to the probability output layer, and outputting a first false touch probability;
The method comprises the steps of matching the abnormal feature set with the false touch feature set according to the feature matching layer to obtain a feature matching matrix, wherein the method comprises the following steps:
generating an abnormal initial matrix according to the abnormal feature set;
Generating a false touch comparison matrix by using the false touch feature set;
Identifying the false touch comparison matrix according to the abnormal initial matrix, obtaining a coefficient matrix of the coincident characteristic, and outputting the coefficient matrix of the coincident characteristic as the characteristic matching matrix;
the probability calculation is carried out on the feature matching matrix according to the probability output layer, and the method comprises the following steps:
disassembling the feature matching matrix, and outputting a first disassembly matrix based on the number of features and a second disassembly matrix based on feature errors;
and carrying out probability factorization on the first dismantling matrix and the second dismantling matrix, and outputting a first false touch probability.
2. The method of claim 1, wherein the method further comprises:
Carrying out abnormality accompanying risk identification on the real-time abnormal data set, and acquiring a first accompanying index, wherein the first accompanying index refers to an index for measuring the risk degree of the real-time abnormal data set;
judging whether the first accompanying index is smaller than or equal to a preset accompanying index;
And when the first accompanying index is smaller than or equal to the preset accompanying index, generating a false touch judging instruction, and activating the false touch pattern recognition model according to the false touch judging instruction.
3. The method of claim 1, wherein the method further comprises:
Classifying the marked sample data set according to the type of the error electric shock protection mode, and outputting a marked sample classification result;
extracting features according to the labeling sample classification result to obtain multiple classes of false touch feature sets;
And establishing a multi-type false touch mode identification model according to the multi-type false touch feature set, and carrying out probability calculation under a corresponding false touch mode by using the multi-type false touch mode identification model.
4. A method as claimed in claim 3, wherein the method further comprises:
Analyzing the real-time abnormal data set to obtain a first matching type, wherein the first matching type refers to type analysis of a false touch mode of the real-time abnormal data set;
Performing pattern matching on the multiple types of false touch pattern recognition models according to the first matching type to obtain a first matching recognition pattern, wherein the first matching recognition pattern corresponds to one false touch pattern recognition unit in the multiple types of false touch pattern recognition models, namely an adaptability processing unit for performing false touch analysis on the real-time abnormal data set;
And carrying out false touch probability calculation according to the first matching recognition mode to obtain a first false touch probability.
5. An intelligent early warning system for an outdoor power metering box, applied to the method of any one of claims 1 to 4, the system comprising:
the data acquisition module is used for acquiring historical overheat protection monitoring data of the first power metering box;
the data identification module is used for identifying the historical overheat protection monitoring data and generating a sample data set triggered by power protection;
The data labeling module is used for labeling the sample data set to obtain a labeled sample data set, wherein the labeled sample data set is a sample data set in an electric shock protection mode;
The feature extraction module is used for extracting features of the labeling sample data set to obtain a false touch feature set;
The triggering acquisition module is used for acquiring a real-time abnormal data set of the first power metering box when the first power metering box triggers an overheat protection instruction;
The false touch probability calculation module is used for establishing a false touch pattern recognition model by the false touch feature set, inputting the real-time abnormal data set into the false touch pattern recognition model to perform false touch probability calculation, and acquiring a first false touch probability;
the early warning information generation module is used for generating first early warning information when the first false touch probability reaches a preset false touch probability.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311672942.XA CN117686757B (en) | 2023-12-07 | 2023-12-07 | Intelligent early warning method and system for outdoor power metering box |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311672942.XA CN117686757B (en) | 2023-12-07 | 2023-12-07 | Intelligent early warning method and system for outdoor power metering box |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117686757A CN117686757A (en) | 2024-03-12 |
| CN117686757B true CN117686757B (en) | 2024-06-11 |
Family
ID=90129449
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311672942.XA Active CN117686757B (en) | 2023-12-07 | 2023-12-07 | Intelligent early warning method and system for outdoor power metering box |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117686757B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108108909A (en) * | 2018-01-05 | 2018-06-01 | 广东电网有限责任公司中山供电局 | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations |
| CN109064060A (en) * | 2018-09-10 | 2018-12-21 | 国网江苏省电力有限公司 | A kind of excitation surge current leads to the appraisal procedure of protective relaying maloperation risk |
| CN109558032A (en) * | 2018-12-05 | 2019-04-02 | 北京三快在线科技有限公司 | Operation processing method, device and computer equipment |
| CN110807172A (en) * | 2019-10-12 | 2020-02-18 | 许昌许继软件技术有限公司 | Risk prediction method and device for relay protection of power system |
| CN111475804A (en) * | 2020-03-05 | 2020-07-31 | 浙江省北大信息技术高等研究院 | Alarm prediction method and system |
| CN112015279A (en) * | 2019-05-28 | 2020-12-01 | 北京搜狗科技发展有限公司 | Key mis-touch error correction method and device |
| CN113176847A (en) * | 2021-04-07 | 2021-07-27 | 上海墨案智能科技有限公司 | Method and equipment for preventing mistaken touch of ink screen |
| CN113449946A (en) * | 2020-03-27 | 2021-09-28 | 广西电网有限责任公司 | Risk assessment method and device for relay protection setting system |
| CN115407259A (en) * | 2022-09-20 | 2022-11-29 | 南方电网科学研究院有限责任公司 | Suspected coefficient-considered out-of-tolerance electric meter online detection method and device and storage medium |
| CN116467667A (en) * | 2023-06-20 | 2023-07-21 | 图观(天津)数字科技有限公司 | Power failure monitoring and early warning method and system based on data fusion |
| CN116643110A (en) * | 2023-06-29 | 2023-08-25 | 山东兴楚电气工程有限公司 | High-voltage frequency converter operation monitoring system |
| JP2023139900A (en) * | 2022-03-22 | 2023-10-04 | 三菱電機株式会社 | Monitoring control device, monitoring control system, monitoring control method, and monitoring control program |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105186463B (en) * | 2015-08-31 | 2018-03-16 | 许继集团有限公司 | A kind of method for preventing relay protection device of intelligent substation maloperation |
| US20230237897A1 (en) * | 2022-01-21 | 2023-07-27 | Syber Sense IP Holdings, LLC | Method of reducing a false trigger alarm on a security ecosystem |
-
2023
- 2023-12-07 CN CN202311672942.XA patent/CN117686757B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108108909A (en) * | 2018-01-05 | 2018-06-01 | 广东电网有限责任公司中山供电局 | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations |
| CN109064060A (en) * | 2018-09-10 | 2018-12-21 | 国网江苏省电力有限公司 | A kind of excitation surge current leads to the appraisal procedure of protective relaying maloperation risk |
| CN109558032A (en) * | 2018-12-05 | 2019-04-02 | 北京三快在线科技有限公司 | Operation processing method, device and computer equipment |
| CN112015279A (en) * | 2019-05-28 | 2020-12-01 | 北京搜狗科技发展有限公司 | Key mis-touch error correction method and device |
| CN110807172A (en) * | 2019-10-12 | 2020-02-18 | 许昌许继软件技术有限公司 | Risk prediction method and device for relay protection of power system |
| CN111475804A (en) * | 2020-03-05 | 2020-07-31 | 浙江省北大信息技术高等研究院 | Alarm prediction method and system |
| CN113449946A (en) * | 2020-03-27 | 2021-09-28 | 广西电网有限责任公司 | Risk assessment method and device for relay protection setting system |
| CN113176847A (en) * | 2021-04-07 | 2021-07-27 | 上海墨案智能科技有限公司 | Method and equipment for preventing mistaken touch of ink screen |
| JP2023139900A (en) * | 2022-03-22 | 2023-10-04 | 三菱電機株式会社 | Monitoring control device, monitoring control system, monitoring control method, and monitoring control program |
| CN115407259A (en) * | 2022-09-20 | 2022-11-29 | 南方电网科学研究院有限责任公司 | Suspected coefficient-considered out-of-tolerance electric meter online detection method and device and storage medium |
| CN116467667A (en) * | 2023-06-20 | 2023-07-21 | 图观(天津)数字科技有限公司 | Power failure monitoring and early warning method and system based on data fusion |
| CN116643110A (en) * | 2023-06-29 | 2023-08-25 | 山东兴楚电气工程有限公司 | High-voltage frequency converter operation monitoring system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117686757A (en) | 2024-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3454289B1 (en) | Plant abnormality detection method and system | |
| CN106407589B (en) | Fan state evaluation and prediction method and system | |
| CN104390657B (en) | A kind of Generator Unit Operating Parameters measurement sensor fault diagnosis method and system | |
| CN116204842B (en) | Abnormality monitoring method and system for electrical equipment | |
| CN114266944B (en) | Rapid model training result checking system | |
| CN115373370A (en) | Method and system for monitoring running state of programmable controller | |
| CN109284886B (en) | Electricity utilization safety management method and device based on artificial intelligence | |
| CN116704733B (en) | Aging early warning method and system for aluminum alloy cable | |
| CN116304909A (en) | Abnormality detection model training method, fault scene positioning method and device | |
| CN117057644A (en) | Equipment production quality detection method and system based on characteristic matching | |
| CN116822115A (en) | Environment management method and system for intelligent park based on digital twin technology | |
| CN117686757B (en) | Intelligent early warning method and system for outdoor power metering box | |
| CN113868948A (en) | User-oriented dynamic threshold model training system and method | |
| CN112734977B (en) | Equipment risk early warning system and algorithm based on Internet of things | |
| CN116796261B (en) | Closed switch equipment mechanical characteristic prediction method based on artificial intelligence | |
| CN117590159A (en) | Tunnel cable power supply state monitoring method and system based on deep learning | |
| CN116520068B (en) | Diagnostic method, device, equipment and storage medium for electric power data | |
| CN115034094B (en) | Prediction method and system for operation state of metal processing machine tool | |
| CN114037010A (en) | Method and device for identifying abnormal electric quantity data | |
| CN117908522B (en) | Fault positioning method for initiating and controlling equipment based on simulation analysis | |
| CN117330507B (en) | Remote test control method for handheld laser instrument | |
| CN117011597A (en) | Hard pressing plate state identification method and system based on primary equipment operation mode | |
| CN116663875B (en) | Intelligent installation engineering monitoring system and method based on big data | |
| CN117648649B (en) | State detection and analysis method and device for intelligent identification | |
| KR102201303B1 (en) | Method of Acquiring Plant Logic Model Using Visual Device |
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 |