CN113391239B - Mutual inductor anomaly monitoring method and system based on edge calculation - Google Patents
Mutual inductor anomaly monitoring method and system based on edge calculation Download PDFInfo
- Publication number
- CN113391239B CN113391239B CN202110648925.7A CN202110648925A CN113391239B CN 113391239 B CN113391239 B CN 113391239B CN 202110648925 A CN202110648925 A CN 202110648925A CN 113391239 B CN113391239 B CN 113391239B
- Authority
- CN
- China
- Prior art keywords
- data
- anomaly
- edge
- abnormal
- primary side
- 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
- 238000004364 calculation method Methods 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000012544 monitoring process Methods 0.000 title claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 95
- 238000013145 classification model Methods 0.000 claims abstract description 87
- 238000000605 extraction Methods 0.000 claims abstract description 62
- 238000012549 training Methods 0.000 claims abstract description 47
- 238000013527 convolutional neural network Methods 0.000 claims description 52
- 238000012545 processing Methods 0.000 claims description 39
- 238000001514 detection method Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 15
- 230000005856 abnormality Effects 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000008447 perception Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 239000002131 composite material Substances 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- WABPQHHGFIMREM-UHFFFAOYSA-N lead(0) Chemical compound [Pb] WABPQHHGFIMREM-UHFFFAOYSA-N 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000002699 waste material Substances 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
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/56—Testing of electric apparatus
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R35/00—Testing or calibrating of apparatus covered by the other groups of this subclass
- G01R35/02—Testing or calibrating of apparatus covered by the other groups of this subclass of auxiliary devices, e.g. of instrument transformers according to prescribed transformation ratio, phase angle, or wattage rating
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing And Monitoring For Control Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a mutual inductor anomaly monitoring method and a system based on edge calculation, which are used for collecting operation data of primary side equipment in normal operation for d days as normal sample data, classifying the operation data, respectively training a CNN (computer numerical network) feature extraction model, an anomaly classification model and a KNN classification model according to the types, calculating anomaly scores of the normal sample data, and obtaining primary side equipment anomaly score distribution parameter estimation in normal operation; real-time operation data of primary side equipment are collected in real time, and classification, feature extraction and anomaly score calculation are carried out by utilizing each trained model; and comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center. Faults possibly occurring in the transformer are predicted in time, so that related maintenance personnel can carry out overhaul work in time.
Description
Technical Field
The invention relates to a transformer abnormality detection technology, in particular to a method and a system for monitoring transformer abnormality based on edge calculation.
Background
With the continuous development of modern industrial technology and information technology, the traditional power grid gradually develops towards intellectualization and digitization. The smart power grid has advanced measurement and sensing technologies, and ensures safe and reliable operation of the modern power grid. The transformer is an indispensable ring in the intelligent power grid, can realize standardization and miniaturization of measuring instruments, protection equipment and automatic control equipment, can isolate a high-voltage system, and ensures personal safety. Therefore, the running condition of the transformer is monitored, so that the transformer can run stably, faults are prevented, and the method has very important significance.
The existing transformer fault diagnosis method is less, the transformer fault is monitored and evaluated, the transformer state is generally measured manually in a certain verification period under the power failure state, but because the high-voltage line has a certain economic loss due to power failure, most transformers in a power grid are in an operation state exceeding the verification period, and state detection is needed under the condition of no power failure. In the prior art, most of the fault diagnosis is performed by using an empirical judgment or simpler machine learning algorithm, the fault monitoring of the transformer is performed by combining the machine learning algorithm according to the secondary side current and voltage information and the temperature information acquired by the acquisition module, and the deep fault effect with a complex analysis structure is not ideal. In addition, the diagnosis methods need to collect the operation parameters of the transformer for analysis and judgment, which can lead to the information acquired by a large number of acquisition modules to be uploaded to the cloud, occupy a large amount of bandwidth and calculation resources, and waste a large amount of bandwidth resources during data transmission.
Disclosure of Invention
The invention aims to provide an edge calculation-based transformer anomaly monitoring method which is used for monitoring the running state of a transformer in real time and predicting faults in time by providing calculation resources at the edge of a network.
The invention is realized by the following technical scheme:
an edge calculation-based transformer anomaly monitoring method comprises the following steps:
step S1, collecting operation data of the primary side transformer equipment in normal operation for d days as normal sample data, classifying the normal sample data, respectively training a convolutional neural network feature extraction model, an abnormal classification model and a KNN classification model according to the classes, and calculating abnormal scores of the normal sample data according to the abnormal classification model to obtain primary side transformer equipment abnormal score distribution parameter estimation in normal operation;
s2, acquiring real-time operation data of primary side transformer equipment in real time, classifying by using a trained KNN classification model to obtain data to be detected, sequentially extracting data features and detecting anomalies by using a convolutional neural network feature extraction model and an anomaly classification model, and calculating anomaly scores according to anomaly detection results to obtain anomaly score calculation results;
and step S3, comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center.
Further, the specific process of step S1 is as follows:
s11, acquiring operation data of the primary side transformer equipment in normal operation by using a sensor, and summarizing data of d days of the primary side transformer equipment in normal operation as normal sample data to be transmitted to a collector, wherein d is more than or equal to 90;
step S12, uploading normal sample data to an edge classification data center through a collector, and forwarding the normal sample data to a centralized data processing center by the edge classification data center;
step S13, preprocessing the normal sample data by the centralized data processing center, and quickly clustering the normal sample data by using a CFSFDP algorithm to obtain training data with a data tag;
step S14, the centralized data processing center classifies training data according to the data labels, respectively trains a convolutional neural network characteristic extraction model and an abnormal classification model under corresponding categories according to the classified training data to obtain convolutional neural network characteristic extraction model parameters and abnormal classification model parameters corresponding to the edge processing data centers under different categories, and issues the obtained model parameters to different edge processing data centers according to the categories; training the KNN classification model according to the clustering data with the data label to obtain model parameters of the KNN classification model, and issuing the model parameters of the KNN classification model to the edge classification data center;
step S15, the centralized data processing center calculates abnormal scores of normal sample data according to the abnormal classification model, obtains primary side transformer equipment abnormal score distribution parameter estimation in normal operation and sends the primary side transformer equipment abnormal score distribution parameter estimation to a corresponding edge processing data center;
taking d days as a primary updating period, taking the collected operation data of the primary side transformer equipment in normal operation as normal sample data every d days, and executing the steps to finish updating the convolutional neural network feature extraction model, the abnormal classification model and the KNN classification model.
Further, the specific process of step S2 is as follows:
s21, acquiring real-time operation data of primary side transformer equipment by using a sensor, and uploading the real-time operation data to an acquisition device;
step S22, uploading real-time operation data to an edge classification data center through a collector, adjusting a KNN classification model in the edge classification data center by the edge classification data center according to KNN classification model parameters issued by a centralized data processing center, classifying the real-time operation data according to the KNN model to obtain data to be detected, and distributing the data to be detected to the edge processing data center for processing the data to be detected according to the classification;
step S23, the edge processing data center sequentially performs data feature extraction and anomaly detection on the data to be detected according to the adjusted convolutional neural network feature extraction model and the anomaly classification model, and performs anomaly score calculation according to the anomaly detection result to obtain an anomaly score calculation result;
and step S24, comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center.
Further, in step S1, the training process of the convolutional neural network feature extraction model specifically includes:
decomposing the training data into one-dimensional time operation data and two-dimensional day operation data;
inputting the one-dimensional time operation data and the two-dimensional day operation data into a convolutional neural network feature extraction model, sequentially carrying out local perception, parameter sharing and convolutional calculation on the input data by the convolutional neural network feature extraction model, extracting data features of normal sample data, and obtaining training data;
the activation function and the optimization function in the convolutional neural network feature extraction model are a ReLU algorithm and a sgd algorithm respectively, and dropout is used in a multi-layer convolutional structure of the convolutional neural network feature extraction model.
Further, the abnormal classification model adopts an isolated forest algorithm, and the specific process of obtaining the abnormal score distribution parameter estimation by using the abnormal classification model is as follows:
sampling from training data to obtain a plurality of sample points, respectively forming a plurality of sub-samples by the plurality of sample points, respectively constructing an isolated tree for each sub-sample, testing each isolated tree in a forest, and recording the path length of each isolated tree;
calculating an anomaly score by utilizing an anomaly score calculation formula according to the path length of each tree to obtain an anomaly score of each sample point;
and summarizing the abnormal score of each sample point to obtain the corresponding primary side transformer equipment abnormal score distribution parameter estimation in normal operation.
Further, the specific process of testing each isolated tree is as follows:
randomly selecting psi data from the training data as sub-samples, and putting the sub-samples into a root node of an isolated tree;
randomly designating a dimension, and randomly generating a cutting point p in the current node data range, wherein the cutting point p is generated between the maximum value and the minimum value of the designated dimension in the current node data;
dividing the data space of the current node by using the cutting point p, and continuously repeating until the height of only one data in the subsamples or the isolated tree in the root node corresponding to the current node reaches log2 (ψ), and recording the path length of the isolated tree at the moment.
Further, calculating an anomaly score, and calculating an anomaly score of each sample point in the training data according to an anomaly score calculation formula, wherein the calculation formula is as follows:
where x represents one sample point in the subsamples, h (x) represents the path length that the sample point x passes to reach the leaf node on the ith isolated tree, and E (h (x)) represents the average path length of x on Nt trees in the forest, expressed as:
c (ψ) represents the average path length required in the binary search tree when a search is unsuccessful, and the calculation formula is:
when the abnormality score s (x, psi) of x is close to 1, the sample point is a discrete point, and the corresponding primary side transformer equipment operation state is abnormal;
when the abnormality score s (x, ψ) of x is close to 0, the sample point is a normal point, and the corresponding primary side transformer device operation state is normal.
The existing monitoring and evaluation of faults of the transformer generally adopts the method that the state of the transformer is manually measured in a power failure state in a certain verification period, but because the power failure of a high-voltage line can cause certain economic loss, most of transformers in a power grid are in an operation state exceeding the verification period, and state detection is required under the condition of non-power failure. In the prior art, the fault classification of the transformer is rarely realized by adopting a machine learning algorithm based on an edge computing framework, and in a transformer fault monitoring method by using a non-manual method, most of the fault classification is realized by adopting secondary side current and voltage information and temperature information acquired by an acquisition module in combination with the machine learning algorithm, so that a large amount of information acquired by the acquisition module needs to be uploaded to a cloud, and a large amount of bandwidth and computing resources are occupied. Simple fault identification is through artificial mode, needs the staff to carry equipment to rush to the scene, and dismouting mutual-inductor one-time lead wire, inefficiency, and can't accurately judge mutual-inductor running state, influence electric power system's safe operation and electric energy measurement's fairness. According to the technical scheme, the bandwidth pressure is reduced by selecting the composite convolutional neural network which can extract the periodicity rule as the characteristic extraction model, selecting the isolated forest algorithm which is suitable for carrying out anomaly discrimination and can score the anomaly state as the fault classification model, collecting the information of the primary physical power grid of the transformer, judging the running state and the running trend of the current transformer, and timely predicting the transformer with the descending running trend and the poorer running state according to the characteristic extraction of the running parameters of the front and rear transformer equipment, so that possible faults of the transformer can be effectively predicted, and related maintenance personnel can carry out overhaul work timely. Most computing tasks are concentrated in an edge computing center, computing results are uploaded to a cloud for early warning, transmitted data are reduced, and bandwidth resources can be saved; in addition, the system model is updated regularly, so that errors in the prediction process can be reduced.
In addition, the invention provides an edge calculation-based transformer anomaly monitoring system, which comprises primary side transformer equipment, a sensor, a collector, an edge data center and a centralized data processing center, wherein,
the primary side transformer device is used for providing normal sample data and real-time operation data of normal operation;
the sensor is used for collecting real-time operation data and normal sample data of the primary side transformer equipment and summarizing and transmitting the real-time operation data and the normal sample data to the collector;
the collector is used for uploading real-time operation data and normal sample data of the primary side transformer equipment to the edge data center;
the edge data center is used for uploading normal sample data to the centralized data processing center, downloading KNN classification model parameters and convolutional neural network characteristics under corresponding categories from the centralized data processing center, extracting model parameters and abnormal classification model parameters, classifying real-time operation data according to the downloaded models, calculating abnormal scores, and giving an alarm to the centralized data processing center according to calculation results;
the centralized data processing center is used for regularly receiving normal sample data in normal operation from the edge data center, clustering all the normal sample data, respectively training a convolutional neural network feature extraction model and an abnormal classification model according to the clustered data to obtain convolutional neural network feature extraction model parameters and abnormal classification model parameters under different categories, calculating abnormal scores of the normal sample data according to the abnormal classification model, obtaining primary transformer equipment abnormal score distribution parameter estimation in normal operation, and issuing each obtained model parameter and primary transformer equipment abnormal score distribution parameter estimation to the edge data center according to the categories, training a KNN classification model according to the normal sample data, obtaining parameters of the KNN classification model and issuing the parameters to the edge data center.
Further, the edge data center comprises an edge classification data center and an edge processing data center,
the edge classification data center is used for uploading normal sample data to the centralized data processing center, classifying the real-time operation data according to the trained KNN classification model to obtain data to be detected, and distributing the data to be detected to the edge processing data center under the corresponding classification;
the edge processing data center comprises a model training module, a data processing module and an early warning module,
the model training module is used for downloading the convolutional neural network characteristic extraction model parameters and the abnormal classification model parameters of the edge processing data center under the corresponding classification from the centralized data processing center, and adjusting the corresponding convolutional neural network characteristic extraction model and the abnormal classification model according to the model parameters;
the data processing module is used for carrying out data feature extraction and anomaly detection on the data to be detected distributed from the edge classification data center according to the adjusted convolutional neural network feature extraction model and the anomaly classification model, and carrying out anomaly score calculation according to the anomaly detection result to obtain an anomaly score calculation result;
the early warning module is used for comparing the obtained abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment, predicting whether the corresponding primary side transformer equipment fails or not, and sending the prediction result to the centralized data processing center.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the method and the system for monitoring the abnormity of the transformer based on the edge calculation, computing resources are provided at the edge of a network, a composite convolutional neural network which can extract a periodicity rule is selected as a feature extraction model, an isolated forest algorithm which is suitable for performing abnormity discrimination and can score abnormal states is selected as a fault classification model to monitor the running state of the transformer in real time, so that on one hand, bandwidth pressure is reduced, on the other hand, the transformer with lower running trend and poorer running state is predicted in time according to the extraction features of the running parameters of the front and rear transformer devices, and possible faults are predicted in time, so that related maintenance personnel can perform overhaul work in time.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiments of the invention. In the drawings:
FIG. 1 is a schematic flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of the overall framework of the system of the present invention;
fig. 3 is a schematic diagram of fault monitoring and acquisition.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present invention, the present invention will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present invention and the descriptions thereof are for illustrating the present invention only and are not to be construed as limiting the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one of ordinary skill in the art that: no such specific details are necessary to practice the invention. In other instances, well-known structures, circuits, materials, or methods have not been described in detail in order not to obscure the invention.
Throughout the specification, references to "one embodiment," "an embodiment," "one example," or "an example" mean: a particular feature, structure, or characteristic described in connection with the embodiment or example is included within at least one embodiment of the invention. Thus, the appearances of the phrases "in one embodiment," "in an example," or "in an example" in various places throughout this specification are not necessarily all referring to the same embodiment or example. Furthermore, the particular features, structures, or characteristics may be combined in any suitable combination and/or sub-combination in one or more embodiments or examples. Moreover, those of ordinary skill in the art will appreciate that the illustrations provided herein are for illustrative purposes and that the illustrations are not necessarily drawn to scale. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
In the description of the present invention, it should be understood that the terms "front", "rear", "left", "right", "upper", "lower", "vertical", "horizontal", "high", "low", "inner", "outer", etc. indicate orientations or positional relationships based on the drawings, are merely for convenience in describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the scope of the present invention.
Example 1
As shown in fig. 1, the method for monitoring the anomaly of the transformer based on edge calculation comprises the following steps:
step S1, collecting operation data of primary side transformer equipment (such as a current transformer) in normal operation for d days as normal sample data, classifying the normal sample data, respectively training a convolutional neural network feature extraction model, an abnormal classification model and a KNN classification model according to the classes, and calculating abnormal scores of the normal sample data according to the abnormal classification model to obtain abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation; and taking d days as a primary updating period, taking the collected operation data of the primary side transformer equipment in normal operation as normal sample data every d days, and executing the following steps of updating the convolutional neural network feature extraction model, the abnormal classification model and the KNN classification model, wherein the specific process is as follows:
s11, acquiring operation data (such as voltage and current actual values of a transformer) of the primary side transformer equipment in normal operation by using a sensor, and summarizing data of d days of the primary side transformer equipment in normal operation as normal sample data and transmitting the normal sample data to a collector, wherein d is more than or equal to 90;
step S12, uploading normal sample data to an edge classification data center through a collector, and forwarding the normal sample data to a centralized data processing center by the edge classification data center;
step S13, preprocessing the normal sample data by the centralized data processing center, and quickly clustering the normal sample data by using a CFSFDP algorithm to obtain training data with a data tag;
step S14, the centralized data processing center classifies training data according to the data labels, respectively trains a convolutional neural network characteristic extraction model and an abnormal classification model under corresponding categories according to the classified training data to obtain convolutional neural network characteristic extraction model parameters and abnormal classification model parameters corresponding to the edge processing data centers under different categories, and issues the obtained model parameters to different edge processing data centers according to the categories; training the KNN classification model according to the clustering data with the data label to obtain model parameters of the KNN classification model, and issuing the model parameters of the KNN classification model to the edge classification data center;
and S15, the centralized data processing center calculates abnormal scores of the normal sample data according to the abnormal classification model, obtains the distribution parameter estimation of the abnormal scores of the primary side transformer equipment in normal operation, and sends the distribution parameter estimation to the corresponding edge processing data center.
S2, acquiring real-time operation data of primary side transformer equipment in real time, classifying by using a trained KNN classification model to obtain data to be detected, sequentially extracting data features and detecting anomalies by using a convolutional neural network feature extraction model and an anomaly classification model, and calculating anomaly scores according to anomaly detection results to obtain anomaly score calculation results;
s21, acquiring real-time operation data of primary side transformer equipment by using a sensor, and uploading the real-time operation data to an acquisition device;
step S22, uploading real-time operation data to an edge classification data center through a collector, adjusting a KNN classification model in the edge classification data center by the edge classification data center according to KNN classification model parameters issued by a centralized data processing center, classifying the real-time operation data according to the KNN model to obtain data to be detected, and distributing the data to be detected to the edge processing data center for processing the data to be detected according to the classification;
step S23, the edge processing data center sequentially performs data feature extraction and anomaly detection on the data to be detected according to the adjusted convolutional neural network feature extraction model and the anomaly classification model, and performs anomaly score calculation according to the anomaly detection result to obtain an anomaly score calculation result;
and step S24, comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center.
And step S3, comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center.
Specifically, clustering is performed in a centralized processing center using a fast search density peak clustering technique (CFSFDP algorithm) into 2 steps, including PCA feature extraction, finding density peak points and classifying according to distance.
The PCA feature extraction obtains a device operation feature matrix V, V in the following form:
wherein: k represents the number of the characteristics of the equipment, namely the running parameters of the transformer such as current, voltage, maximum value, minimum value, standard deviation of current-voltage sequence and the like; v in matrix uifk Representing the value of user i on feature k. Searching density peak points and classifying according to the distance, wherein the density peak points are defined according to the following formula:
there are two methods for the representation of the local density, wherein equation 2 is the cut-off distance method, wherein(i, j) is point i and point jEuclidean distance between->For the cut-off distance, equation 3 is a kernel distance method.
Delta (i) is the peak distance, the CFSFDP algorithm calculates the local density ρ and the higher density distance delta, maps the data set into a two-dimensional map and constructs a decision map for selection; in the decision graph, the point where ρ and δ are both large (the upper right point) is the cluster center. After the cluster center is selected, the rest points are distributed to the cluster center closest to the cluster center to complete the clustering.
The training process using the CNN convolutional neural network feature extraction model in the step S1 specifically includes:
decomposing the training data into one-dimensional time operation data, and converting the one-dimensional time operation data into two-dimensional day operation data; features are conveniently extracted, and the shape of the operation data input in one dimension is as follows:
M 1×d =[P 1 P 2 … P d ] (5)
wherein h is the total operating hours of the primary side equipment;
the two-dimensional daily operation data input shape is as follows:
wherein d is the total operation days of the primary side equipment, and d is more than or equal to 90;
inputting one-dimensional time operation data and two-dimensional day operation data into a convolutional neural network feature extraction model, sequentially carrying out local perception, parameter sharing and convolutional calculation on the input data by the convolutional neural network feature extraction model, and obtaining an effective feature extraction result by utilizing the parameter sharing to achieve a great reduction of the number of parameters, extracting data features of normal sample data and obtaining training data;
the activation function and the optimization function in the convolutional neural network feature extraction model are a ReLU algorithm and a sgd algorithm respectively, and dropout is used in a multilayer convolutional structure of the convolutional neural network feature extraction model; the probability of the occurrence of the fitting problem is reduced; wherein the formula of the ReLU is as follows:
wherein y is j Is the output of the full link layer in the jth neuron, n is the one-dimensional input data length, w i,j Representing the neuron weight between the 1 st input value and the j-th neuron, b 1 Is the deviation.
In this embodiment, the abnormal classification model adopts an isolated forest algorithm, and the specific process of obtaining the abnormal score distribution parameter estimation by using the abnormal classification model is as follows:
sampling from training data to obtain a plurality of sample points, respectively constructing an isolated tree for each sample point, testing each isolated tree in a forest, and recording the path length of each isolated tree;
calculating an anomaly score by utilizing an anomaly score calculation formula according to the path length of each tree to obtain an anomaly score of each sample point;
and summarizing the abnormal score of each sample point to obtain the corresponding primary side transformer equipment abnormal score distribution parameter estimation in normal operation.
Specifically, the specific process of testing each isolated tree is as follows:
randomly selecting psi data from the training data as sub-samples, and putting the sub-samples into a root node of an isolated tree;
randomly designating a dimension, and randomly generating a cutting point p in the current node data range, wherein the cutting point p is generated between the maximum value and the minimum value of the designated dimension in the current node data;
dividing the data space of the current node by using the cutting point p, and continuously repeating until the height of only one data in the subsamples or the isolated tree in the root node corresponding to the current node reaches log2 (ψ), and recording the path length of the isolated tree at the moment.
Calculating an anomaly score, and calculating an anomaly score of each sample point in the training data according to an anomaly score calculation formula, wherein the calculation formula is as follows:
where h (x) represents the path length traversed to reach the leaf node on the ith orphan tree, and E (h (x)) represents the average of the path lengths of x over Nt trees, expressed as:
c (ψ) represents the average path length required in the binary search tree when a search is unsuccessful, and the calculation formula is:
when the abnormality score s (x, psi) of x is close to 1, the sample point is a discrete point, and the corresponding primary side transformer equipment operation state is abnormal;
when the abnormality score s (x, ψ) of x is close to 0, the sample point is a normal point, and the corresponding primary side transformer device operation state is normal.
When the parameters of the transformer are collected, for example, aiming at the current transformer and the voltage transformer, the system needs to collect current signals and voltage signals as the basis for judging whether the transformer operates normally, and the fault monitoring and collecting principle is shown in fig. 3:
in FIG. 3, CT is a current transformer, PT is a voltage transformer, and the current of the phase A of the transmission line is I A The current of the C phase is I C The metering unit is capable of counting the voltage U a And U c Secondary monitoring signals u of CT1 and CT2 a And u c Uploading the counted current and voltage signals to an edge data center;
before model training, data needs to be preprocessed, and normally the input and output of a neural network are within [0,1], but parameters such as system current and voltage are not within [0,1], so that normalization processing is needed for collected data, and the normalization processing method is shown as the following formula:
wherein Z represents specific input data such as current, voltage and the like, so that the efficiency and accuracy of fault judgment can be improved through the current and voltage values after data processing.
Example 2
As shown in fig. 2, the difference between this embodiment and embodiment 1 is that the present invention provides an edge calculation-based transformer anomaly monitoring system, which includes a primary side transformer device, a sensor, a collector, an edge data center and a centralized data processing center, wherein,
the primary side transformer device is used for providing normal sample data and real-time operation data of normal operation;
the sensor is used for collecting real-time operation data and normal sample data of the primary side transformer equipment and summarizing and transmitting the real-time operation data and the normal sample data to the collector;
the collector is used for uploading real-time operation data and normal sample data of the primary side transformer equipment to the edge data center;
the edge data center comprises an edge classification data center and an edge processing data center, the edge classification data center is used for uploading normal sample data to the centralized data processing center, classifying real-time operation data according to a trained KNN classification model to obtain data to be detected, and distributing the data to be detected to the edge processing data center under the corresponding classification;
the edge processing data center comprises a model training module, a data processing module and an early warning module,
the model training module is used for downloading the convolutional neural network characteristic extraction model parameters and the abnormal classification model parameters of the edge processing data center under the corresponding classification from the centralized data processing center, and adjusting the corresponding convolutional neural network characteristic extraction model and the abnormal classification model according to the model parameters;
the data processing module is used for carrying out data feature extraction and anomaly detection on the data to be detected distributed from the edge classification data center according to the adjusted convolutional neural network feature extraction model and the anomaly classification model, and carrying out anomaly score calculation according to the anomaly detection result to obtain an anomaly score calculation result;
the early warning module is used for comparing the obtained abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment, predicting whether the corresponding primary side transformer equipment fails or not, and sending the prediction result to the centralized data processing center;
the centralized data processing center is used for regularly (for example, 90-120 days) receiving normal sample data in normal operation from the edge data center, clustering all the normal sample data, respectively training a convolutional neural network characteristic extraction model and an abnormal classification model according to the clustered data to obtain convolutional neural network characteristic extraction model parameters and abnormal classification model parameters in different categories, calculating abnormal scores of the normal sample data according to the abnormal classification model, obtaining primary transformer equipment abnormal score distribution parameter estimation in normal operation, sending each obtained model parameter and primary transformer equipment abnormal score distribution parameter estimation to the edge processing data center according to the categories, training a KNN classification model according to the normal sample data, obtaining parameters of the KNN classification model and sending the parameters to the edge classification data center.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (8)
1. The method for monitoring the abnormity of the transformer based on the edge calculation is characterized by comprising the following steps of:
s1, collecting operation data of the primary side transformer equipment in normal operation for d days as normal sample data, classifying the normal sample data, respectively training a convolutional neural network feature extraction model, an abnormal classification model and a KNN classification model according to the classes, and calculating abnormal scores of the normal sample data according to the abnormal classification model to obtain primary side transformer equipment abnormal score distribution parameter estimation in normal operation; the specific process of the step S1 is as follows: s11, acquiring operation data of the primary side transformer equipment in normal operation by using a sensor, and summarizing data of d days of the primary side transformer equipment in normal operation as normal sample data to be transmitted to a collector, wherein d is more than or equal to 90; step S12, uploading normal sample data to an edge classification data center through a collector, and forwarding the normal sample data to a centralized data processing center by the edge classification data center; step S13, preprocessing the normal sample data by the centralized data processing center, and quickly clustering the normal sample data by using a CFSFDP algorithm to obtain training data with a data tag; step S14, the centralized data processing center classifies training data according to the data labels, respectively trains a convolutional neural network characteristic extraction model and an abnormal classification model under corresponding categories according to the classified training data to obtain convolutional neural network characteristic extraction model parameters and abnormal classification model parameters corresponding to the edge processing data centers under different categories, and issues the obtained model parameters to different edge processing data centers according to the categories; training the KNN classification model according to the clustering data with the data label to obtain model parameters of the KNN classification model, and issuing the model parameters of the KNN classification model to the edge classification data center; step S15, the centralized data processing center calculates abnormal scores of normal sample data according to the abnormal classification model, obtains primary side transformer equipment abnormal score distribution parameter estimation in normal operation and sends the primary side transformer equipment abnormal score distribution parameter estimation to a corresponding edge processing data center;
s2, acquiring real-time operation data of primary side transformer equipment in real time, classifying by using a trained KNN classification model to obtain data to be detected, sequentially extracting data features and detecting anomalies by using a convolutional neural network feature extraction model and an anomaly classification model, and calculating anomaly scores according to anomaly detection results to obtain anomaly score calculation results;
and step S3, comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center.
2. The method for monitoring the anomaly of the transformer based on the edge calculation according to claim 1, wherein the specific process of the step S2 is as follows:
s21, acquiring real-time operation data of primary side transformer equipment by using a sensor, and uploading the real-time operation data to an acquisition device;
step S22, uploading real-time operation data to an edge classification data center through a collector, adjusting a KNN classification model in the edge classification data center by the edge classification data center according to KNN classification model parameters issued by a centralized data processing center, classifying the real-time operation data according to the KNN model to obtain data to be detected, and distributing the data to be detected to the edge processing data center for processing the data to be detected according to the classification;
step S23, the edge processing data center sequentially performs data feature extraction and anomaly detection on the data to be detected according to the adjusted convolutional neural network feature extraction model and the anomaly classification model, and performs anomaly score calculation according to the anomaly detection result to obtain an anomaly score calculation result;
and step S24, comparing the abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment in normal operation, predicting whether the primary side transformer equipment can generate faults, and sending the prediction result to a centralized data processing center.
3. The method for monitoring the anomaly of the transformer based on the edge calculation according to claim 1, wherein the training process of the convolutional neural network feature extraction model in the step S1 is specifically as follows:
decomposing the training data into one-dimensional time operation data and two-dimensional day operation data;
inputting the one-dimensional time operation data and the two-dimensional day operation data into a convolutional neural network feature extraction model, sequentially carrying out local perception, parameter sharing and convolutional calculation on the input data by the convolutional neural network feature extraction model, extracting data features of normal sample data, and obtaining training data;
wherein the activation function and the optimization function in the convolutional neural network feature extraction model are respectivelyAnd->Algorithm, use ++in multi-layer convolution structure of convolutional neural network feature extraction model>。
4. The method for monitoring the anomaly of the transformer based on the edge calculation of claim 3, wherein the anomaly classification model adopts an isolated forest algorithm, and the specific process of obtaining the anomaly score distribution parameter estimation by using the anomaly classification model is as follows:
sampling from training data to obtain a plurality of sample points, respectively forming a plurality of sub-samples by the plurality of sample points, respectively constructing an isolated tree for each sub-sample, testing each isolated tree in a forest, and recording the path length of each isolated tree;
calculating an anomaly score by utilizing an anomaly score calculation formula according to the path length of each tree to obtain an anomaly score of each sample point;
and summarizing the abnormal score of each sample point to obtain the corresponding primary side transformer equipment abnormal score distribution parameter estimation in normal operation.
5. The method for monitoring the anomaly of the transformer based on the edge calculation as claimed in claim 4, wherein the specific process of testing each isolated tree is as follows:
randomly selecting psi data from the training data as sub-samples, and putting the sub-samples into a root node of an isolated tree;
randomly designating a dimension, and randomly generating a cutting point p in the current node data range, wherein the cutting point p is generated between the maximum value and the minimum value of the designated dimension in the current node data;
dividing the data space of the current node by using the cutting point p, and continuously repeating until the height of only one data in the subsamples or the isolated tree in the root node corresponding to the current node reaches log2 (ψ), and recording the path length of the isolated tree at the moment.
6. The method for monitoring the anomaly of the transformer based on the edge calculation according to claim 5, wherein the anomaly score is calculated, and the anomaly score of each sample point in the training data is calculated according to an anomaly score calculation formula, wherein the calculation formula is as follows:
;
where x represents one sample point in the subsamples, h (x) represents the path length that the sample point x passes to reach the leaf node on the ith isolated tree, and E (h (x)) represents the average path length of x on Nt trees in the forest, expressed as:
;
the average path length required when unsuccessful searching occurs in a binary search tree is represented by the following calculation formula:
when the abnormality score s (x, psi) of x is close to 1, the sample point is a discrete point, and the corresponding primary side transformer equipment operation state is abnormal;
when the abnormality score s (x, ψ) of x is close to 0, the sample point is a normal point, and the corresponding primary side transformer device operation state is normal.
7. An edge calculation based transformer anomaly monitoring system for performing an edge calculation based transformer anomaly monitoring method of any one of claims 1 to 6, comprising a primary side transformer device, a sensor, a collector, an edge data center, and a centralized data processing center, wherein,
the primary side transformer device is used for providing normal sample data and real-time operation data of normal operation;
the sensor is used for collecting real-time operation data and normal sample data of the primary side transformer equipment and summarizing and transmitting the real-time operation data and the normal sample data to the collector;
the collector is used for uploading real-time operation data and normal sample data of the primary side transformer equipment to the edge data center;
the edge data center is used for uploading normal sample data to the centralized data processing center, downloading KNN classification model parameters and convolutional neural network characteristics under corresponding categories from the centralized data processing center, extracting model parameters and abnormal classification model parameters, classifying real-time operation data according to the downloaded models, calculating abnormal scores, and giving an alarm to the centralized data processing center according to calculation results;
the centralized data processing center is used for regularly receiving normal sample data in normal operation from the edge data center, clustering all the normal sample data, respectively training a convolutional neural network feature extraction model and an abnormal classification model according to the clustered data to obtain convolutional neural network feature extraction model parameters and abnormal classification model parameters under different categories, calculating abnormal scores of the normal sample data according to the abnormal classification model, obtaining primary transformer equipment abnormal score distribution parameter estimation in normal operation, and issuing each obtained model parameter and primary transformer equipment abnormal score distribution parameter estimation to the edge data center according to the categories, training a KNN classification model according to the normal sample data, obtaining parameters of the KNN classification model and issuing the parameters to the edge data center.
8. The system for monitoring anomaly of a transformer based on edge calculation of claim 7,
the edge data center includes an edge classification data center and an edge processing data center,
the edge classification data center is used for uploading normal sample data to the centralized data processing center, classifying the real-time operation data according to the trained KNN classification model to obtain data to be detected, and distributing the data to be detected to the edge processing data center under the corresponding classification;
the edge processing data center comprises a model training module, a data processing module and an early warning module,
the model training module is used for downloading the convolutional neural network characteristic extraction model parameters and the abnormal classification model parameters of the edge processing data center under the corresponding classification from the centralized data processing center, and adjusting the corresponding convolutional neural network characteristic extraction model and the abnormal classification model according to the model parameters;
the data processing module is used for carrying out data feature extraction and anomaly detection on the data to be detected distributed from the edge classification data center according to the adjusted convolutional neural network feature extraction model and the anomaly classification model, and carrying out anomaly score calculation according to the anomaly detection result to obtain an anomaly score calculation result;
the early warning module is used for comparing the obtained abnormal score calculation result with the abnormal score distribution parameter estimation of the primary side transformer equipment, predicting whether the corresponding primary side transformer equipment fails or not, and sending the prediction result to the centralized data processing center.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110648925.7A CN113391239B (en) | 2021-06-10 | 2021-06-10 | Mutual inductor anomaly monitoring method and system based on edge calculation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110648925.7A CN113391239B (en) | 2021-06-10 | 2021-06-10 | Mutual inductor anomaly monitoring method and system based on edge calculation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113391239A CN113391239A (en) | 2021-09-14 |
CN113391239B true CN113391239B (en) | 2024-01-23 |
Family
ID=77620272
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110648925.7A Active CN113391239B (en) | 2021-06-10 | 2021-06-10 | Mutual inductor anomaly monitoring method and system based on edge calculation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113391239B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116256690A (en) * | 2023-02-15 | 2023-06-13 | 广州市德珑电子器件有限公司 | Current transformer safety detection method and system |
CN116699318B (en) * | 2023-07-27 | 2023-10-20 | 成都汉度科技有限公司 | Electric leakage detection method and system based on edge calculation |
CN117148775B (en) * | 2023-10-31 | 2024-01-23 | 中国电建集团山东电力管道工程有限公司 | Remote monitoring method, system, equipment and medium for pipeline production process |
CN117216673B (en) * | 2023-11-08 | 2024-03-12 | 国网江西省电力有限公司电力科学研究院 | Current transformer monitoring evaluation overhauls platform |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020133639A1 (en) * | 2018-12-29 | 2020-07-02 | 东北大学 | Intelligent analysis system for magnetic flux leakage detection data in pipeline |
CN111522804A (en) * | 2020-04-15 | 2020-08-11 | 国网江苏省电力有限公司 | Cleaning method for abnormal data of transformer equipment state monitoring |
-
2021
- 2021-06-10 CN CN202110648925.7A patent/CN113391239B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020133639A1 (en) * | 2018-12-29 | 2020-07-02 | 东北大学 | Intelligent analysis system for magnetic flux leakage detection data in pipeline |
CN111522804A (en) * | 2020-04-15 | 2020-08-11 | 国网江苏省电力有限公司 | Cleaning method for abnormal data of transformer equipment state monitoring |
Non-Patent Citations (3)
Title |
---|
云边智能: 电力系统运行控制的边缘计算方法 及其应用现状与展望;白昱阳等;《自动化学报》;46(第3期);第397页至第410页 * |
基于孤立森林算法的用电数据异常检测研究;余翔等;《信息技术》;第88页至第92页 * |
基于改进快速密度峰值聚类算法的 电力大数据异常值检测;刘凤魁等;《电力信息与通信技术》;15(第6期);第36页至第40页 * |
Also Published As
Publication number | Publication date |
---|---|
CN113391239A (en) | 2021-09-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113391239B (en) | Mutual inductor anomaly monitoring method and system based on edge calculation | |
CN113255795B (en) | Equipment state monitoring method based on multi-index cluster analysis | |
CN113156917B (en) | Power grid equipment fault diagnosis method and system based on artificial intelligence | |
WO2011043108A1 (en) | Equipment status monitoring method, monitoring system, and monitoring program | |
CN108537154A (en) | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning | |
CN110837866A (en) | XGboost-based electric power secondary equipment defect degree evaluation method | |
CN104732276A (en) | On-line diagnosing method for faults of metering production facility | |
CN111754020A (en) | Intelligent power plant equipment inspection system and method based on data analysis | |
CN114856941A (en) | Offshore wind power plant and unit fault diagnosis operation and maintenance system and diagnosis operation and maintenance method thereof | |
CN114723285A (en) | Power grid equipment safety evaluation prediction method | |
CN115511367B (en) | Intelligent quality management system of production line | |
CN111767657B (en) | Nuclear power system fault diagnosis method and system | |
CN115409131A (en) | Production line abnormity detection method based on SPC process control system | |
CN115060312A (en) | Building material safety monitoring system based on artificial intelligence | |
CN113071966A (en) | Elevator fault prediction method, device, equipment and storage medium | |
CN114487705A (en) | Power grid equipment fault positioning detection method | |
CN117277566B (en) | Power grid data analysis power dispatching system and method based on big data | |
CN112734977A (en) | Equipment risk early warning system and algorithm based on Internet of things | |
CN117151684A (en) | Wind power fan data analysis early warning method, system, device and readable storage medium | |
CN111459697A (en) | Excitation system fault monitoring method based on deep learning network | |
CN117113135A (en) | Carbon emission anomaly monitoring and analyzing system capable of sorting and classifying anomaly data | |
CN115600695B (en) | Fault diagnosis method for metering equipment | |
CN116244581A (en) | Wind power blade fault early warning method and system based on artificial intelligent platform | |
CN114936600A (en) | Document abnormity monitoring method, device, equipment and storage medium | |
CN115713027A (en) | Transformer state evaluation method, device and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |