CN117034180B - Power communication equipment data anomaly detection method, system and storage medium - Google Patents
Power communication equipment data anomaly detection method, system and storage medium Download PDFInfo
- Publication number
- CN117034180B CN117034180B CN202311300263.XA CN202311300263A CN117034180B CN 117034180 B CN117034180 B CN 117034180B CN 202311300263 A CN202311300263 A CN 202311300263A CN 117034180 B CN117034180 B CN 117034180B
- Authority
- CN
- China
- Prior art keywords
- data
- time
- power communication
- model
- communication equipment
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 66
- 238000004891 communication Methods 0.000 title claims abstract description 57
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 32
- 238000012549 training Methods 0.000 claims abstract description 19
- 238000013528 artificial neural network Methods 0.000 claims abstract description 18
- 230000005856 abnormality Effects 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 230000000306 recurrent effect Effects 0.000 claims abstract description 8
- 230000002159 abnormal effect Effects 0.000 claims description 20
- 230000006870 function Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000013480 data collection Methods 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000012300 Sequence Analysis Methods 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 238000012217 deletion Methods 0.000 claims description 3
- 230000037430 deletion Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims description 3
- 238000004590 computer program Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 125000004122 cyclic group Chemical group 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The application relates to a method, a system and a storage medium for detecting data abnormality of power communication equipment, wherein the method comprises the following steps of 1) collecting historical data of various power communication equipment, wherein the historical data mainly comprises network performance data, equipment performance data, temperature, current and voltage and timestamp information; 2) Preprocessing various data collected in the step 1); 3) Converting the processed data into a time sequence format, and constructing an abnormality detection data set; 4) Inputting the anomaly detection data set obtained in the step 3) into an RNN (RNN recurrent neural network) for training, and generating a data anomaly detection model; 5) And (3) extracting real-time data of the power communication equipment, processing the data, inputting the processed data into the abnormality detection model generated in the step (4), and judging the state of the power communication equipment according to the model output result. By using the time sequence and combining the neural network algorithm, the anomaly detection can be carried out on the data of various power communication devices at the same time.
Description
Technical Field
The present disclosure relates to the field of power system security, and in particular, to a method, a system, and a storage medium for detecting data anomalies in power communication devices.
Background
The power communication equipment is a core component in the power system and is used for realizing information transmission and remote control between the power equipment. These devices generate a large amount of data, including information on voltage, current, frequency, temperature, etc., which is critical to the detection, operation and safety of the power system. In order to ensure the normal operation of the power system, abnormal conditions in the data of the power communication equipment must be found and processed in time. Therefore, the data abnormality detection of the power communication equipment has important significance in a modern power system, and is helpful for ensuring the normal operation of the power equipment and improving the stability and reliability of the power system.
Because of the increasing variety and number of power communication devices, it is a difficult problem how to design a unified system to detect anomalies of multiple data at the same time. The data anomaly detection of the power communication equipment mainly has the following problems, namely the data generated by the power communication equipment is high-dimensional nowadays, and the traditional anomaly detection method is low in efficiency when processing the data. Secondly, in practical application, the labels of the abnormal data are usually missing, and the working condition of the power system can also change at any time, so the model needs to be robust. Thirdly, abnormal data is usually in a few classes, while normal data occupies most of data, so that the problem of data unbalance is caused.
Disclosure of Invention
An object of the embodiments of the present application is to provide a method, a system, and a storage medium for detecting data anomalies of a power communication device, which solve the problem of low model robustness caused by complex and diversified data by combining a time sequence with a neural network algorithm, and can detect anomalies of data of multiple power communication devices at the same time by constructing a deep learning model.
In order to achieve the above purpose, the present application provides the following technical solutions:
in a first aspect, an embodiment of the present application provides a method for detecting an abnormality of data in an electric power communication device, including the steps of:
1) Collecting historical data of various power communication devices, wherein the historical data mainly comprises network performance data, device performance data, temperature, current voltage and timestamp information;
2) Preprocessing various data collected in the step 1);
3) Converting the processed data into a time sequence format, and constructing an abnormality detection data set;
4) Inputting the anomaly detection data set obtained in the step 3) into an RNN (RNN recurrent neural network) for training, and generating a data anomaly detection model;
5) And (3) extracting real-time data of the power communication equipment, processing the data, inputting the processed data into the abnormality detection model generated in the step (4), and judging the state of the power communication equipment according to the model output result.
The network performance data in the step 1) comprises broadband utilization rate, packet loss rate, delay, throughput and the like; the equipment state data comprises the running state of equipment and the memory utilization rate; each data is secured with time stamp information to facilitate subsequent data processing.
In the step 2), various data collected in the step 1) need to be preprocessed, and the following processing modes are mainly performed:
processing the missing values: detecting missing values in the data, including deleting rows containing missing values, filling the missing values with a mean or median value, estimating the missing values using interpolation,
outlier processing: detecting and processing abnormal values, including identifying abnormal values by statistical methods or based on domain knowledge, and performing deletion or replacement processing;
timestamp processing: firstly, ensuring that the time stamps are ordered according to a time sequence so as to perform time sequence analysis, if the time stamps are not ordered, ordering the time stamps, calculating intervals among the time stamps, and ensuring that the time intervals among the data points are uniform;
and carrying out normalization processing on the data of different features, and shrinking the data to the same scale range.
In the step 3), the processed data is converted into a time sequence format, and an abnormal detection data set is constructed, and the time window sample sequence is created mainly by using a sliding window technology when the data is converted into the time sequence format.
In step 4), the anomaly detection data set obtained in step 3) is input into an RNN recurrent neural network for training, and a data anomaly detection model is generated, wherein the training based on the RNN recurrent neural network is obtained through the following process training:
forward propagation
Input data: feature sequences within a time windowWhere t represents a time step, initializing the hidden state h (0) to a zero vector or a random vector when time step t=0,
calculating a hidden state, wherein h (t) is the hidden state at time step t,is the hidden state at time step t-1, X (t) is the input at time step t, W and U are weight matrices, b is the bias term, and tanh is the hyperbolic tangent activation function:
，
calculating an output o (t), where o (t) is the model output at time step t,is the output layerIs used for the weight matrix of the (c),is the bias term for the output layer:
，
loss function: calculating the difference between the model output and the true value using the mean square error, where Loss (t) is the Loss at time step t, o (t) is the output of the model, y (t) is the true value,the method is quadratic:
，
counter-propagation
The gradient of the loss function with respect to the model parameters is calculated using the chain law, in order to update the weights and biases of the model,
calculating the gradient of the output layer weight and bias:
，
calculating gradient of hidden state:
，
calculating the parameter gradient of the RNN layer in the model:
，
parameter update, whereinIs learning rate, controls the step length of parameter updating:
，
、/>is a weight matrix after parameter update, +.>Is an offset term after the parameter is updated,is the weight matrix of the output layer after parameter update, < >>Is the bias term of the output layer after parameter updating, and the iteration is repeated until a certain iteration number is reached or the loss converges to a satisfactory level.
In a second aspect, an embodiment of the present application provides a system for detecting data anomalies in a power communication device, including a data collection module, a data preprocessing module, a data set construction module, a detection model generation module, and a result output module,
the data collection module is used for collecting historical data of various power communication devices;
the data preprocessing module is used for preprocessing various data collected by the data collecting module;
the data set construction module converts the processed data into a time sequence format and constructs an abnormality detection data set;
the detection model generation module inputs the abnormal detection data set obtained by the data set construction module into the RNN circulating neural network for training to generate a data abnormal detection model;
the result output module extracts real-time data of the power communication equipment, processes the data, inputs the processed data into the abnormal detection model generated by the detection model generation module, and judges the state of the power communication equipment according to the model output result.
In a third aspect, embodiments of the present application provide a computer-readable storage medium storing program code that, when executed by a processor, implements the steps of the power communication device data anomaly detection method described above.
Compared with the prior art, the beneficial effects of this application are: the method and the device collect historical data of various power communication equipment, preprocess all the data, convert the processed data into a time sequence format and construct an abnormality detection data set. Inputting the data set into the RNN cyclic neural network for training, and generating a data abnormality detection model. Finally, real-time data of the power communication equipment are extracted and input into an anomaly detection model after being processed, and the state of the power communication equipment is judged according to the output result of the model. Compared with the existing data anomaly detection method of the power communication equipment, the method has the advantages that the problem of low model robustness caused by complex and diversified data is solved by combining the time sequence with the neural network algorithm, and anomaly detection can be carried out on data of various power communication equipment at the same time by constructing a deep learning model.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a graph of an RNN-based recurrent neural network model algorithm in the present application;
FIG. 2 is a flow chart of the method of the present application;
fig. 3 is a system block diagram of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.
The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The terms "first," "second," and the like, are used merely to distinguish one entity or action from another entity or action, and are not to be construed as indicating or implying any actual such relationship or order between such entities or actions.
As shown in fig. 1-2, the present application provides a method for detecting data anomalies of a power communication device based on deep learning, which includes the following steps:
step 1: historical data of various power communication devices is collected, and the historical data mainly comprises network performance data, device performance data, temperature, current voltage and timestamp information. Wherein the network performance data comprises broadband utilization, packet loss, delay, throughput and the like; the equipment state data comprises the running state of equipment and the memory utilization rate; wherein each data is secured with time stamp information to facilitate subsequent data processing. The extraction of various data is mainly extracted from power communication equipment, and specifically comprises the following steps:
1) The data is obtained from the monitoring system of the power communication equipment, and the equipment has built-in monitoring functions.
2) Sensor devices are used to measure environmental parameters such as temperature, current, voltage, etc.
3) Network traffic data is captured using a network capture traffic tool, wireshark.
Step 2: preprocessing various data collected in the step 1, and mainly performing the following processing modes:
1) Processing the missing values: the missing values in the data are detected by a method comprising deleting the row containing the missing values, filling the missing values with the mean or median value, and estimating the missing values using an interpolation method.
2) Outlier processing: outliers are detected and processed using methods including statistical methods and domain-based knowledge to identify outliers and using appropriate processing means such as deletion, substitution.
3) Timestamp processing: first, the time stamps are ensured to be ordered in time sequence for time sequence analysis, and if the time stamps are not ordered, the ordering is performed, and the intervals between the time stamps are calculated, so that the time intervals between the data points are ensured to be uniform.
4) And carrying out normalization processing on the data of different features, and shrinking the data to the same scale range so as to avoid overlarge influence of certain features on the model.
Step 3: the processed data is converted into a time series format, and a different detection data set is constructed. Converting data to a time series format creates a sequence of time window samples, mainly using a sliding window technique. Defining a time set T to include time nodesDefine F as the feature set,> representing the individual features and defining Representing the respective characteristic values. Initially a time series of data is represented as follows:
TABLE 1 initial time series data
Converting the time sequence data into a time window, and storing the processed data into a data format suitable for model training, wherein the converted time window data is as follows:
TABLE 2 converted time window data
Step 4: inputting the anomaly detection data set obtained in the step 3 into an RNN (RNN-network-based) cyclic neural network for training, and generating a data anomaly detection model, wherein the training based on the RNN cyclic neural network is obtained through training by the following process:
1) Forward propagation
a. Input data: feature sequences within a time windowWhere t represents a time step. At time step t=0, the hidden state h (0) is initialized to a zero vector.
b. Calculating a hidden state, wherein h (t) is the hidden state at time step t, X (t) is the input at time step t, W and U are weight matrices, b is a bias term, and tanh is a hyperbolic tangent activation function:
，
c. calculating an output o (t), where o (t) is the model output at time step t,is a matrix of weights for the output layer,is the bias term for the output layer:
，
d. loss function: calculating the difference between the model output and the true value using the mean square error, where Loss (t) is the Loss at time step t, o (t) is the output of the model, and y (t) is the true value:
，
2) Counter-propagation
Gradients of the loss function relative to the model parameters are calculated using the chain law to update the weights and biases of the model.
a. Calculating the gradient of the output layer weight and bias:
，
b. calculating gradient of hidden state:
，
c. calculating the parameter gradient of the RNN layer in the model:
，
d. parameter update, whereinIs learning rate, controls the step length of parameter updating:
，
、/>is a weight matrix after parameter update, +.>Is an offset term after the parameter is updated,is the weight matrix of the output layer after parameter update, < >>Is the bias term of the output layer after parameter updating,
3) And repeating the iteration until a certain iteration number or loss converges to a satisfactory level, setting a threshold value threshold according to the iteration training result, wherein the condition that the threshold value is lower indicates that the data is normal, and the condition that the data is higher indicates that the data is abnormal and needs to be further analyzed.
Step 5: and (4) extracting real-time data of the power communication equipment, processing the data, inputting the processed data into the abnormality detection model generated in the step (4), and judging the state of the power communication equipment according to the model output result. The output value is mainly compared with threshold, and the state of the power communication equipment corresponding to the abnormal data is further analyzed.
As shown in fig. 3, the embodiment of the application provides a data anomaly detection system of power communication equipment, which comprises a data collection module 1, a data preprocessing module 2, a data set construction module 3, a detection model generation module 4 and a result output module 5,
the data collection module 1 collects historical data of various power communication devices;
the data preprocessing module 2 preprocesses various data collected by the data collecting module;
the data set construction module 3 converts the processed data into a time sequence format and constructs an abnormality detection data set;
the detection model generation module 4 inputs the abnormal detection data set obtained by the data set construction module into the RNN circulating neural network for training to generate a data abnormal detection model;
the result output module 5 extracts real-time data of the power communication equipment, processes the data, inputs the processed data into the abnormal detection model generated by the detection model generation module, and judges the state of the power communication equipment according to the model output result.
According to the deep learning-based power communication equipment data anomaly detection method, historical data of various power communication equipment are collected, all data are preprocessed, the processed data are converted into a time sequence format, and an anomaly detection data set is constructed. Inputting the data set into the RNN cyclic neural network for training, and generating a data abnormality detection model. Finally, real-time data of the power communication equipment are extracted and input into an anomaly detection model after being processed, and the state of the power communication equipment is judged according to the output result of the model.
Compared with the existing data anomaly detection method of the power communication equipment, the method has the advantages that the problem of low model robustness caused by complex and diversified data is solved by combining the time sequence with the neural network algorithm, and anomaly detection can be carried out on data of various power communication equipment at the same time by constructing a deep learning model.
The embodiment of the application provides a computer readable storage medium storing program code which, when executed by a processor, implements the steps of the power communication device data anomaly detection method described above.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
The foregoing is merely exemplary embodiments of the present application and is not intended to limit the scope of the present application, and various modifications and variations may be suggested to one skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
Claims (3)
1. A method for detecting anomalies in data of an electrical communication device, comprising the steps of:
1) Collecting historical data of various power communication devices, including network performance data, device performance data, temperature, current voltage and timestamp information;
2) Preprocessing various data collected in the step 1);
3) Converting the processed data into a time sequence format, and constructing an abnormality detection data set;
4) Inputting the anomaly detection data set obtained in the step 3) into an RNN (RNN recurrent neural network) for training, and generating a data anomaly detection model;
5) Extracting real-time data of the power communication equipment, processing the data, inputting the processed data into the abnormality detection model generated in the step 4), and judging the state of the power communication equipment according to the model output result;
the network performance data in step 1) comprises broadband utilization, packet loss, delay and throughput; the equipment state data comprises the running state of equipment and the memory utilization rate; each data is ensured to have time stamp information so as to facilitate subsequent data processing;
in the step 2), various data collected in the step 1) need to be preprocessed, and the following processing modes are performed:
processing the missing values: detecting missing values in the data, including deleting rows containing missing values, filling the missing values with a mean or median value, estimating the missing values using interpolation,
outlier processing: detecting and processing abnormal values, including identifying abnormal values by statistical methods or based on domain knowledge, and performing deletion or replacement processing;
timestamp processing: firstly, ensuring that the time stamps are ordered according to a time sequence so as to perform time sequence analysis, if the time stamps are not ordered, ordering the time stamps, calculating intervals among the time stamps, and ensuring that the time intervals among the data points are uniform;
carrying out normalization processing on the data of different characteristics, and shrinking the data to the same scale range;
in the step 3), the processed data is converted into a time sequence format, an abnormal detection data set is constructed, the data is converted into the time sequence format, and a time window sample sequence is created by using a sliding window technology;
in step 4), the anomaly detection data set obtained in step 3) is input into an RNN recurrent neural network for training, and a data anomaly detection model is generated, wherein the training based on the RNN recurrent neural network is obtained through the following process training:
input data: feature sequences within a time windowWhere t represents a time step, initializing the hidden state h (0) to a zero vector or a random vector when time step t=0,
calculating a hidden state, wherein h (t) is the hidden state at time step t,is the hidden state at time step t-1, X (t) is the input at time step t, W and U are weight matrices, b is the bias term, and tanh is the hyperbolic tangent activation function:
，
calculating an output o (t), where o (t) is the model output at time step t,is the weight matrix of the output layer, +.>Is the bias term for the output layer:
，
loss function: calculating the difference between the model output and the true value using the mean square error, where Loss (t) is the Loss at time step t, o (t) is the output of the model, y (t) is the true value,the method is quadratic:
，
the gradient of the loss function with respect to the model parameters is calculated using the chain law, in order to update the weights and biases of the model,
calculating the gradient of the output layer weight and bias:
，
calculating gradient of hidden state:
，
calculating the parameter gradient of the RNN layer in the model:
，
parameter update, whereinIs learning rate, controls the step length of parameter updating:
，
、/>is a weight matrix after parameter update, +.>Is an offset item after parameter update, +.>Is the weight matrix of the output layer after parameter update, < >>Is the bias term of the output layer after parameter updating, and the iteration is repeated until a certain iteration number is reached or the loss converges to a satisfactory level.
2. A data anomaly detection system of power communication equipment for implementing the method as claimed in claim 1, comprising a data collection module, a data preprocessing module, a data set construction module, a detection model generation module, and a result output module,
the data collection module is used for collecting historical data of various power communication devices;
the data preprocessing module is used for preprocessing various data collected by the data collecting module;
the data set construction module converts the processed data into a time sequence format and constructs an abnormality detection data set;
the detection model generation module inputs the abnormal detection data set obtained by the data set construction module into the RNN circulating neural network for training to generate a data abnormal detection model;
the result output module extracts real-time data of the power communication equipment, processes the data, inputs the processed data into the abnormal detection model generated by the detection model generation module, and judges the state of the power communication equipment according to the model output result.
3. A computer-readable storage medium storing program code which, when executed by a processor, implements the steps of the power communication device data anomaly detection method of claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311300263.XA CN117034180B (en) | 2023-10-10 | 2023-10-10 | Power communication equipment data anomaly detection method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311300263.XA CN117034180B (en) | 2023-10-10 | 2023-10-10 | Power communication equipment data anomaly detection method, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117034180A CN117034180A (en) | 2023-11-10 |
| CN117034180B true CN117034180B (en) | 2023-12-22 |
Family
ID=88602756
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311300263.XA Active CN117034180B (en) | 2023-10-10 | 2023-10-10 | Power communication equipment data anomaly detection method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117034180B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117574302A (en) * | 2024-01-12 | 2024-02-20 | 威海海洋职业学院 | Ocean fishing boat power system abnormality monitoring method and monitoring system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107977710A (en) * | 2017-12-21 | 2018-05-01 | 南方电网科学研究院有限责任公司 | Multiplexing electric abnormality data detection method and device |
| CN111833583A (en) * | 2020-07-14 | 2020-10-27 | 南方电网科学研究院有限责任公司 | Training method, device, equipment and medium for power data anomaly detection model |
| CN114358152A (en) * | 2021-12-21 | 2022-04-15 | 国网江苏省电力有限公司苏州供电分公司 | Intelligent power data anomaly detection method and system |
| CN115688035A (en) * | 2022-10-19 | 2023-02-03 | 江苏电力信息技术有限公司 | Time sequence power data anomaly detection method based on self-supervision learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109710636B (en) * | 2018-11-13 | 2022-10-21 | 广东工业大学 | Unsupervised industrial system anomaly detection method based on deep transfer learning |
-
2023
- 2023-10-10 CN CN202311300263.XA patent/CN117034180B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107977710A (en) * | 2017-12-21 | 2018-05-01 | 南方电网科学研究院有限责任公司 | Multiplexing electric abnormality data detection method and device |
| CN111833583A (en) * | 2020-07-14 | 2020-10-27 | 南方电网科学研究院有限责任公司 | Training method, device, equipment and medium for power data anomaly detection model |
| CN114358152A (en) * | 2021-12-21 | 2022-04-15 | 国网江苏省电力有限公司苏州供电分公司 | Intelligent power data anomaly detection method and system |
| CN115688035A (en) * | 2022-10-19 | 2023-02-03 | 江苏电力信息技术有限公司 | Time sequence power data anomaly detection method based on self-supervision learning |
Non-Patent Citations (2)
| Title |
|---|
| " Research of Substation Equipment Abnormity Identification Based on Image Processing";Xu Changfu etc;《2017 International Conference on Smart Grid and Electrical Automation (ICSGEA)》;全文 * |
| 基于混合卷积神经网络和循环神经网络的入侵检测模型;方圆;李明;王萍;江兴何;张信明;;计算机应用(10);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117034180A (en) | 2023-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7010641B2 (en) | Abnormality diagnosis method and abnormality diagnosis device | |
| CN108429254B (en) | non-invasive electric load identification method | |
| CN117034180B (en) | Power communication equipment data anomaly detection method, system and storage medium | |
| US20090234607A1 (en) | Evaluating anomaly for one-class classifiers in machine condition monitoring | |
| CN104317681A (en) | Behavioral abnormality automatic detection method and behavioral abnormality automatic detection system aiming at computer system | |
| CN111767957B (en) | Log abnormality detection method and device, storage medium and electronic equipment | |
| CN111314329A (en) | Traffic intrusion detection system and method | |
| Tavakoli et al. | Event tracking for real-time unaware sensitivity analysis (EventTracker) | |
| CN114595210A (en) | Multi-dimensional data anomaly detection method and device and electronic equipment | |
| CN114118576A (en) | Regional geological disaster trend prediction method and system | |
| CN116559667A (en) | Model training method and device, battery detection method and device, equipment and medium | |
| CN114726581A (en) | Abnormity detection method and device, electronic equipment and storage medium | |
| CN115018012A (en) | Internet of things time sequence anomaly detection method and system under high-dimensional characteristic | |
| CN113093985B (en) | Sensor data link abnormity detection method and device and computer equipment | |
| CN115438897A (en) | Industrial process product quality prediction method based on BLSTM neural network | |
| CN111191720A (en) | Service scene identification method and device and electronic equipment | |
| CN117313015A (en) | Time sequence abnormality detection method and system based on time sequence and multiple variables | |
| CN113194068A (en) | Malicious encrypted flow detection method and device based on convolutional neural network | |
| CN115495274B (en) | Exception handling method based on time sequence data, network equipment and readable storage medium | |
| CN117131452A (en) | Abnormality detection method and system based on normalized flow and Bayesian network | |
| JP7173284B2 (en) | Event monitoring device, method and program | |
| CN108761250B (en) | Industrial control equipment voltage and current-based intrusion detection method | |
| EP4075219A1 (en) | System and method of monitoring an industrial environment | |
| US20230297095A1 (en) | Monitoring device and method for detecting anomalies | |
| Gatta et al. | Capability of the Bayesian Forecasting Method to Predict Field Time Series |
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 |