CN116976530B - Cable equipment state prediction method, device and storage medium - Google Patents
Cable equipment state prediction method, device and storage medium Download PDFInfo
- Publication number
- CN116976530B CN116976530B CN202311234823.6A CN202311234823A CN116976530B CN 116976530 B CN116976530 B CN 116976530B CN 202311234823 A CN202311234823 A CN 202311234823A CN 116976530 B CN116976530 B CN 116976530B
- Authority
- CN
- China
- Prior art keywords
- state
- data
- correlation
- quantities
- states
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012544 monitoring process Methods 0.000 claims abstract description 25
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000004364 calculation method Methods 0.000 claims description 13
- 230000009977 dual effect Effects 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 230000007613 environmental effect Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000009469 supplementation Effects 0.000 claims description 2
- 230000000875 corresponding effect Effects 0.000 description 11
- 230000006872 improvement Effects 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 239000010979 ruby Substances 0.000 description 1
- 229910001750 ruby Inorganic materials 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The application relates to a cable equipment state prediction method, a device and a storage medium, wherein the method comprises the following steps: acquiring state monitoring data of various cable equipment, classifying state quantities, and dividing the state quantities into various states; carrying out data preprocessing on the state quantity; calculating the correlation between the state quantities under each state, and determining the state quantity weight; respectively constructing a state prediction model based on an LSTM network for each state, taking time sequence state quantity and state quantity weight under the corresponding state as input, and outputting a state prediction result; calculating the correlation between states according to the state prediction result, and determining the state weight; and constructing a cable equipment state prediction model based on the LSTM network, and taking state sequence data and state weight obtained by prediction of each state prediction model as input to output a cable equipment state prediction result. Compared with the prior art, the method and the device have the advantages that the device states are predicted by calculating the correlation of a plurality of sequences and the coordination of a plurality of states, and the prediction is accurate.
Description
Technical Field
The present application relates to the field of cable device state prediction, and in particular, to a method, an apparatus, and a storage medium for predicting a cable device state based on data correlation and a dual LSTM network.
Background
Live detection, on-line monitoring and periodic testing will continuously generate new status data during the operation of the cable plant. Because the cable equipment is large in laying space and the number of state feature detection points is large, the realization difficulty of directly acquiring the actual running state condition of the cable equipment from massive original state data is high, so that the monitoring data needs to be identified, analyzed, extracted and generalized, the actual running state of the cable equipment is combined to be predicted, and the state of the cable equipment is evaluated, so that the safe and sustainable running of the cable equipment is ensured.
The common state monitoring data of the operation state of the cable equipment can reflect the operation state of the cable equipment, such as abnormal sound, partial discharge, heating defects, current and resistance. For example, CN112070322a discloses a method for predicting the running state of a high-voltage cable circuit based on a long-short-term memory network, which utilizes deep learning to realize multi-state-quantity joint prediction of a single node, improves the accuracy of cable state prediction, and ensures the running reliability of a high-voltage power transmission cable. However, the application does not consider the correlation between the state amounts. The time sequence rule of each state monitoring data reflects the change trend of the corresponding state, and the correlation relationship between different state monitoring data can reflect the mutual influence relationship between different states of the equipment. For example, after insulation breakage occurs in the intermediate connector of the cable device, the data related to the partial discharge state may have a significant mutation, and the temperature data and the sound data may also change. At present, the correlation between the state quantities has no clear regularity, and no cable equipment state prediction method considering the state quantity correlation exists in the prior art.
Disclosure of Invention
The application aims to provide a cable equipment state prediction method, a device and a storage medium based on data correlation and a double LSTM network, which are used for predicting a monitoring value by analyzing time sequence data of a plurality of correlated state quantities, predicting and evaluating equipment states by utilizing the cooperation of the plurality of state data sequences through analyzing the correlation or similarity among the plurality of state quantities, and improving the prediction precision of the whole running state of the cable equipment.
The aim of the application can be achieved by the following technical scheme:
a cable equipment state prediction method based on data correlation and double LSTM network includes the following steps:
acquiring state monitoring data of various cable equipment, classifying state quantities, and dividing the state quantities into various states;
carrying out data preprocessing on various state quantities of the cable equipment state monitoring data;
calculating the correlation between the state quantities under each state, and determining the state quantity weight;
respectively constructing a state prediction model based on an LSTM network for each state, taking the preprocessed time sequence state quantity and state quantity weight in the corresponding state as input, and outputting a state prediction result;
calculating the correlation between states according to the state prediction result, and determining the state weight;
and constructing a cable equipment state prediction model based on the LSTM network, and taking state sequence data and state weight obtained by prediction of each state prediction model as input to output a cable equipment state prediction result.
The states include a partial discharge state, a mechanical state, a temperature state, and an environmental state.
The data preprocessing of the various state quantities of the cable equipment state monitoring data comprises the following steps:
carrying out missing value supplementation on the discontinuous data in the time sequence state quantity by adopting a piecewise linear interpolation method;
and carrying out normalization processing on each state quantity time sequence based on the standard deviation.
The method for calculating the correlation between the state quantities in each state comprises the following steps:
segmenting time series data of each state quantity, calculating characteristic mean value of each segment of data, and then obtaining state quantity f i And state quantity f j The correlation calculation formula is as follows:
，
wherein the time series of state quantities is divided into n 1 Segment corr (f) ik ,f jk ) Is the state quantity f i And state quantity f j And (3) measuring the similarity of the characteristic mean values in the same time period k, wherein the similarity measurement adopts cosine similarity.
The method for determining the state quantity weight comprises the following steps: calculating weights for different state quantities in each state, wherein the sum of the weights of the state quantities belonging to one state is 1; the initial weight value of each state quantity is recorded as 1, if the calculated correlation between the state quantity and other state quantities is larger than a preset threshold value, the state quantity and other state quantities are judged to have correlation, and the weight value of the state quantity is accumulated by 1;
the calculation formula of the state quantity weight is as follows:
，
wherein,t i is a state quantityiIs used to determine the final accumulated weight value of (c),i=（1,...,n) Represent the firstiThe number of state quantities is set to be equal to the number of state quantities,nthe number of state quantities.
The method for calculating the correlation between the states comprises the following steps: segmenting the predicted state time sequence data, and then obtaining a state F i And state F j The correlation calculation formula of (2) is as follows:
，
wherein the state time sequence is divided into n 2 Segment, count (F) ik ,F jk ) Is state F i And state F j The number of simultaneous occurrences within the same time period k.
The method for determining the state weight comprises the following steps: calculating weights for each state, wherein the sum of all state weights is 1; the initial weight value of each state is recorded as 1, if the calculated correlation between the state and other states is larger than a preset threshold value, the state and other states are judged to have correlation, and the weight value of the state is accumulated by 1;
the calculation formula of the state weight is as follows:
，
wherein,T i is a state ofiIs used to determine the final accumulated weight value of (c),i=（1,...,m) Represent the firstiThe status of the individual states is that,mis the number of states.
The basic unit of the LSTM network is a memory module, which comprises a memory unit and three gate structures for controlling the states of the memory unit: forget gate, input gate and output gate, forget gate decides to forget useless history information from the memory cell state, input gate decides the influence of current input data to the memory cell state, output gate decides output information;
set X t An input vector h representing time t t-1 Represents the output at time t-1, w f 、w i 、w c 、w o 、u f 、u i 、u c 、u o Representing a weight matrix, b f 、b i 、b c 、b o Representing the bias vector, the memory module performs the following processes of state update and information output:
1) Forget to forget the door forget the useless history information:
，
2) The input gate performs state update according to the input data and the history information:
，
，
，
3) The output gate outputs information of the current moment:
，
，
wherein σ is a logistic sigmoid function, f t 、i t And o t Respectively indicating the output states of the forgetting gate, the input gate and the output gate at the moment t, C t The memory cell state at time t is indicated.
A cable plant state prediction apparatus based on data correlation and dual LSTM network, comprising a memory, a processor, and a program stored in the memory, the processor implementing the method as described above when executing the program.
A storage medium having stored thereon a program which when executed performs a method as described above.
Compared with the prior art, the application has the following beneficial effects:
1. the application can improve the accuracy of the prediction result. In practical application, the state monitoring data often has the characteristics of non-smoothness, instability, noise value, missing value and the like, so that the accuracy of equipment state judgment can be affected.
2. The application has higher real-time performance. The application introduces real-time monitoring data and non-real-time detection data, and can timely discover the poor state of the cable equipment by adjusting the prediction step length of the LSTM, thereby avoiding potential equipment damage and unstable power system.
3. The application scene is wider. The application is suitable for different power equipment state application scenes, can effectively improve the state evaluation accuracy performance of the power equipment, and ensures the safe operation of the power equipment.
Drawings
FIG. 1 is a flow chart of the method of the present application;
FIG. 2 is a flow chart of calculating state quantity weights according to the present application;
FIG. 3 is a schematic diagram of the data processing process of the model of the present application.
Detailed Description
The application will now be described in detail with reference to the drawings and specific examples. The present embodiment is implemented on the premise of the technical scheme of the present application, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present application is not limited to the following examples.
Example 1
The present embodiment provides a cable equipment status prediction method based on data correlation and dual LSTM networks, as shown in fig. 1, including the following steps:
s1, acquiring state monitoring data of various cable equipment, classifying state quantities, and dividing the state quantities into various states;
each state of the cable plant will typically have a plurality of state monitoring data, all of which may be acquired, classified. In this embodiment, the cable apparatus states include a partial discharge state, a mechanical performance state, a temperature state, an environmental state, and the like, each of which reflects the operation condition of the cable apparatus from one side.
For example, the partial discharge state includes ultrahigh frequency monitoring data, ultrasonic monitoring data, high frequency monitoring data, transient ground voltage monitoring data, infrared monitoring data, temperature monitoring data, gas variation experimental data, etc.
S2, preprocessing various state quantities of the cable equipment state monitoring data.
In this embodiment, step S2 includes the following steps:
s21, supplementing the missing value of the discontinuous data in the time sequence state quantity by adopting a piecewise linear interpolation method.
The state data of the cable equipment has linear continuous data and discontinuous data, and the analysis of the time sequence related data is preferably consistent in data length, so that the discontinuous data is supplemented with missing values by adopting a piecewise linear interpolation method. The piecewise interpolation formula is as follows:
，
wherein [ thex k+1 -x k ]Is the interval in which interpolation is required,y k is thatx k The value of the point is set to be,y k+1 is thatx k+1 The value of the dot.
S22, carrying out normalization processing on each state quantity time sequence based on standard deviation.
Each state quantity data represents a different physical quantity, and therefore normalization processing is required for the data. The present embodiment uses normalization of the standard deviation to the original time series for each state quantity:
，
wherein,x mean is the mean value of the state quantity data sequence,x std is the variance of the time series of state quantities。
In other embodiments, different normalization methods may be employed without affecting the achievement of the objectives of the present application.
S3, calculating the correlation between the state quantities under each state, and determining the state quantity weight.
The state quantities in the same state have correlation, for example, noise data and the like can appear when the amplitude of the partial discharge data becomes large. Therefore, it is necessary to calculate the correlation between the state amounts as the state amount weights.
In this embodiment, the time series data of each state quantity is segmented, and the feature mean value of each segment of data is calculated, so that the state quantity f i And state quantity f j The correlation calculation formula is as follows:
，
wherein the time series of state quantities is divided into n 1 Segment corr (f) ik ,f jk ) Is the state quantity f i And state quantity f j Measurement of feature mean similarity over the same time period k. In this embodiment, the similarity measure is cosine similarity.
The characteristic amount of the discharge data may include a discharge amplitude, the number of discharge pulses, whether or not it is discharge, whether or not it is noise, discharge temperature, etc., a characteristic average value being set according to the characteristic amount, and a calculation method thereof.
In one embodiment, a method for determining the state quantity weight is shown in fig. 2, and specifically includes: calculating weights for different state quantities in each state, wherein the sum of the weights of the state quantities belonging to one state is 1; the initial weight value of each state quantity is recorded as 1, if the calculated correlation c between the state quantity and other state quantities ij If the state quantity is larger than the preset threshold value, the correlation between the state quantity and the preset threshold value is judged, and the weight value of the state quantity is accumulated by 1.
Then, the calculation formula of the state quantity weight is:
，
wherein,t i is a state quantityiIs used to determine the final accumulated weight value of (c),i=（1,...,n) Represent the firstiThe number of state quantities is set to be equal to the number of state quantities,nthe number of state quantities.
S4, respectively constructing a state prediction model based on the LSTM network for each state, taking the preprocessed time sequence state quantity and state quantity weight in the corresponding state as input, and outputting a state prediction result.
In this embodiment, the basic unit of the LSTM network is a memory block (memory block), which includes a memory cell (memory cell) and three gate structures for controlling the states of the memory cell (cell state): forget gate (forget gate), input gate (input gate) and output gate (output gate), the forget gate determining to forget useless history information from the memory cell state, the input gate determining the effect of the current input data on the memory cell state, the output gate determining the output information;
set X t An input vector h representing time t t-1 Represents the output at time t-1, w f 、w i 、w c 、w o 、u f 、u i 、u c 、u o Representing a weight matrix, b f 、b i 、b c 、b o Representing the bias vector, the memory module performs the following processes of state update and information output:
1) Forget to forget the door forget the useless history information:
，
2) The input gate performs state update according to the input data and the history information:
，
，
，
3) The output gate outputs information of the current moment:
，
，
wherein σ is a logistic sigmoid function, f t 、i t And o t Respectively indicating the output states of the forgetting gate, the input gate and the output gate at the moment t, C t The memory cell state at time t is indicated.
The state prediction model outputs a prediction score of a corresponding time segment of each state, the prediction score is a numerical value of a [0,1] interval, and the larger the value is, the more serious the state is. The output of the state prediction model may be classified into a plurality of levels according to the prediction score and a preset interval division threshold. In this embodiment, the prediction results are classified into four classes of normal, general, serious, and crisis.
For example, the input of the partial discharge state prediction model is the partial discharge state F P Time-series data P of a plurality of state quantities of (2) 1 ,..P i ,…P n And the corresponding state quantity weight, n is the number of partial discharge state quantities, wherein, P is assumed 1 Is divided into n 1 Segments, in which the time series data is，aThe number of time series. The output of the partial discharge state prediction model is，/>Belongs to one of { normal, general, severe, crisis } and represents the predicted outcome of segment i. Namely, partial discharge state predictionThe prediction result of the model output is expressed as [ severe, general, crisis, … … ]]. The prediction results of other state prediction models are similar.
S5, calculating the correlation between the states according to the state prediction result, and determining the state weight.
In addition to the correlation between state quantities in the same state, there is also a correlation between different states, for example, after partial discharge occurs, some cases may generate heat, i.e., there is some correlation between the partial discharge state and the temperature state.
The method for calculating the correlation between the states in this embodiment is as follows: segmenting the predicted state time sequence data, and then obtaining a state F i And state F j The correlation calculation formula of (2) is as follows:
，
wherein the state time sequence is divided into n 2 Segment, count (F) ik ,F jk ) Is state F i And state F j The number of simultaneous occurrences within the same time period k.
In one embodiment, the method for determining the state weight is as follows: calculating weights for each state, wherein the sum of all state weights is 1; the initial weight value of each state is recorded as 1, if the calculated correlation C between the state and other states ij If the weight value is larger than the preset threshold value, the correlation between the two is judged, and the weight value of the state is accumulated by 1.
The calculation formula of the state weight is as follows:
，
wherein,T i is a state ofiIs used to determine the final accumulated weight value of (c),i=（1,...,m) Represent the firstiThe status of the individual states is that,mis the number of states.
S6, constructing a cable equipment state prediction model based on the LSTM network, taking state sequence data and state weights obtained by prediction of each state prediction model as input, and outputting a cable equipment state prediction result.
The LSTM network structure on which the cable plant state prediction model is based may refer to the state prediction model. It should be understood that the LSTM network structure may be in other forms, or may be combined with other neural network structures such as an attention mechanism, which are conventional means for those skilled in the art, so as to avoid obscuring the purpose of the present application, and this embodiment will not be described herein.
The inputs to the state prediction model of the cable plant are time-series data of a plurality of states of the plant and corresponding state weights thereof, e.g. partial discharge state time-series dataTemperature state time series data->Equal and corresponding weights W P ，W T Etc. The output of the cable plant state prediction model is the plant state time series prediction result F 1 ,F 2 ,…]Can be expressed as [ general, severe, crisis … … ]]。
In steps S3-S6, the model-to-data processing procedure is as shown in FIG. 3, and the input data of the state prediction model is a preprocessed state quantity time sequence and a corresponding state quantity weight, and in one embodiment, the input data includes a state quantity { f } associated with a partial discharge state i1 ,f i2 …f in Mechanical state-related state quantity { f } (i+1)1 ,f (i+1)2 …f (i+1)n … …, temperature state dependent state quantity { f m1 ,f m2 …f mn And each state prediction model outputs state prediction results of different states respectively, and the prediction results are also expressed in a time sequence form. And then, the state prediction results of the different states and the corresponding state weights are input into a cable equipment state prediction model, and finally the state prediction results of the cable equipment, which take the correlation among various cable equipment state monitoring data into consideration, are output.
Example 2
The implementation procedure of this embodiment is shown in embodiment 1, and differs from embodiment 1 in that the similarity measurement in step S3 uses euclidean distance.
Example 3
The implementation steps of this embodiment are seen in embodiment 1, and are different from embodiment 1 in that the output of the state prediction model and the cable plant state prediction model are classified into three classes according to actual needs: normal, mild and severe can also be classified into more grades, which are determined according to the accuracy of the prediction result, the actual application scene requirement and the like, and the different grade classifications do not affect the protection scope of the application and the realization of the application.
Example 4
The present embodiment provides a cable equipment status prediction apparatus based on data correlation and dual LSTM network, including a memory, a processor, and a program stored in the memory, where the processor implements the method described in embodiment 1 when executing the program.
At the hardware level, the cable device state prediction apparatus based on the data correlation and the dual LSTM network includes a processor, an internal bus, a network interface, a memory and a nonvolatile memory, and may of course include hardware required by other services. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs to implement the method of data acquisition described above with respect to fig. 1. Of course, other implementations, such as logic devices or combinations of hardware and software, are not excluded from the present application, that is, the execution subject of the following processing flows is not limited to each logic unit, but may be hardware or logic devices.
Improvements to one technology can clearly distinguish between improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) and software (improvements to the process flow). However, with the development of technology, many improvements of the current method flows can be regarded as direct improvements of hardware circuit structures. Designers almost always obtain corresponding hardware circuit structures by programming improved method flows into hardware circuits. Therefore, an improvement of a method flow cannot be said to be realized by a hardware entity module. For example, a programmable logic device (Programmable Logic Device, PLD) (e.g., field programmable gate array (Field Programmable Gate Array, FPGA)) is an integrated circuit whose logic function is determined by the programming of the device by a user. A designer programs to "integrate" a digital system onto a PLD without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Moreover, nowadays, instead of manually manufacturing integrated circuit chips, such programming is mostly implemented by using "logic compiler" software, which is similar to the software compiler used in program development and writing, and the original code before the compiling is also written in a specific programming language, which is called hardware description language (Hardware Description Language, HDL), but not just one of the hdds, but a plurality of kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), lava, lola, myHDL, PALASM, RHDL (Ruby Hardware Description Language), etc., VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog are currently most commonly used. It will also be apparent to those skilled in the art that a hardware circuit implementing the logic method flow can be readily obtained by merely slightly programming the method flow into an integrated circuit using several of the hardware description languages described above.
Example 5
The present embodiment provides a storage medium having stored thereon a program which, when executed, implements the method described in embodiment 1.
Storage media, including both permanent and non-permanent, removable and non-removable media, may be implemented in any method or technology for storage of information. 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 describes in detail preferred embodiments of the present application. It should be understood that numerous modifications and variations can be made in accordance with the concepts of the application by one of ordinary skill in the art without undue burden. Therefore, all technical solutions which can be obtained by logic analysis, reasoning or limited experiments based on the prior art by a person skilled in the art according to the inventive concept shall be within the scope of protection defined by the claims.
Claims (10)
1. The cable equipment state prediction method based on the data correlation and the double LSTM network is characterized by comprising the following steps of:
acquiring state monitoring data of various cable equipment, classifying state quantities, and dividing the state quantities into various states;
carrying out data preprocessing on various state quantities of the cable equipment state monitoring data;
calculating the correlation between the state quantities under each state, and determining the state quantity weight;
respectively constructing a state prediction model based on an LSTM network for each state, taking the preprocessed time sequence state quantity and state quantity weight in the corresponding state as input, and outputting a state prediction result;
calculating the correlation between states according to the state prediction result, and determining the state weight;
and constructing a cable equipment state prediction model based on the LSTM network, and taking state sequence data and state weight obtained by prediction of each state prediction model as input to output a cable equipment state prediction result.
2. The method for predicting the state of a cable device based on a data correlation and dual LSTM network as claimed in claim 1, wherein said states include a partial discharge state, a mechanical state, a temperature state, and an environmental state.
3. The method for predicting the state of a cable device based on data correlation and dual LSTM network as claimed in claim 1, wherein said data preprocessing of a plurality of state quantities of the cable device state monitoring data comprises the steps of:
carrying out missing value supplementation on the discontinuous data in the time sequence state quantity by adopting a piecewise linear interpolation method;
and carrying out normalization processing on each state quantity time sequence based on the standard deviation.
4. The method for predicting states of cable plant based on data correlation and dual LSTM network as claimed in claim 1, wherein the method for calculating the correlation between state quantities in each state is as follows:
segmenting time series data of each state quantity, calculating characteristic mean value of each segment of data, and then obtaining state quantity f i And state quantity f j The correlation calculation formula is as follows:
，
wherein the time series of state quantities is divided into n 1 Segment corr (f) ik ,f jk ) Is the state quantity f i And state quantity f j And (3) measuring the similarity of the characteristic mean values in the same time period k, wherein the similarity measurement adopts cosine similarity.
5. The method for predicting the state of a cable device based on a data correlation and dual LSTM network according to claim 1, wherein the method for determining the state quantity weight is as follows: calculating weights for different state quantities in each state, wherein the sum of the weights of the state quantities belonging to one state is 1; the initial weight value of each state quantity is recorded as 1, if the calculated correlation between the state quantity and other state quantities is larger than a preset threshold value, the state quantity and other state quantities are judged to have correlation, and the weight value of the state quantity is accumulated by 1;
the calculation formula of the state quantity weight is as follows:
，
wherein,t i is a state quantityiIs used to determine the final accumulated weight value of (c),i=（1,...,n) Represent the firstiThe number of state quantities is set to be equal to the number of state quantities,nthe number of state quantities.
6. The method for predicting the state of a cable device based on data correlation and dual LSTM network according to claim 1, wherein the method for calculating the correlation between states is as follows: segmenting the predicted state time sequence data, and then obtaining a state F i And state F j The correlation calculation formula of (2) is as follows:
，
wherein the state time sequence is divided into n 2 Segment, count (F) ik ,F jk ) Is state F i And state F j The number of simultaneous occurrences within the same time period k.
7. The method for predicting the state of a cable device based on data correlation and dual LSTM networks according to claim 1, wherein the method for determining the state weight is as follows: calculating weights for each state, wherein the sum of all state weights is 1; the initial weight value of each state is recorded as 1, if the calculated correlation between the state and other states is larger than a preset threshold value, the state and other states are judged to have correlation, and the weight value of the state is accumulated by 1;
the calculation formula of the state weight is as follows:
，
wherein,T i is a state ofiIs used to determine the final accumulated weight value of (c),i=（1,...,m) Represent the firstiThe status of the individual states is that,mis the number of states.
8. The method for predicting the state of a cable plant based on a data correlation and dual LSTM network according to claim 1, wherein the basic unit of the LSTM network is a memory module, and the memory module comprises a memory unit and three gate structures for controlling states of the memory unit: forget gate, input gate and output gate, forget gate decides to forget useless history information from the memory cell state, input gate decides the influence of current input data to the memory cell state, output gate decides output information;
set X t An input vector h representing time t t-1 Represents the output at time t-1, w f 、w i 、w c 、w o 、u f 、u i 、u c 、u o Representing a weight matrix, b f 、b i 、b c 、b o Representing the bias vector, the memory module performs the following processes of state update and information output:
1) Forget to forget the door forget the useless history information:
，
2) The input gate performs state update according to the input data and the history information:
，
，
，
3) The output gate outputs information of the current moment:
，
，
wherein σ is a logistic sigmoid function, f t 、i t And o t Respectively indicating the output states of the forgetting gate, the input gate and the output gate at the moment t, C t The memory cell state at time t is indicated.
9. A cable plant status prediction apparatus based on data correlation and dual LSTM networks, comprising a memory, a processor, and a program stored in said memory, wherein said processor implements the method of any of claims 1-8 when executing said program.
10. A storage medium having a program stored thereon, wherein the program, when executed, implements the method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311234823.6A CN116976530B (en) | 2023-09-25 | 2023-09-25 | Cable equipment state prediction method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311234823.6A CN116976530B (en) | 2023-09-25 | 2023-09-25 | Cable equipment state prediction method, device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116976530A CN116976530A (en) | 2023-10-31 |
| CN116976530B true CN116976530B (en) | 2023-12-08 |
Family
ID=88477124
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311234823.6A Active CN116976530B (en) | 2023-09-25 | 2023-09-25 | Cable equipment state prediction method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116976530B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070322A (en) * | 2020-09-28 | 2020-12-11 | 国网河北省电力有限公司雄安新区供电公司 | High-voltage cable line running state prediction method based on long-short term memory network |
| CN112257967A (en) * | 2020-12-21 | 2021-01-22 | 中国电力科学研究院有限公司 | Method for predicting state quantity of cable, method and device for determining operating state of cable |
| CN112989601A (en) * | 2021-03-10 | 2021-06-18 | 西南石油大学 | Submarine cable state evaluation method based on subjective and objective combination weighting |
| CN113468764A (en) * | 2021-07-22 | 2021-10-01 | 西南石油大学 | Submarine cable state prediction method based on time sequence |
| AU2021106675A4 (en) * | 2021-03-23 | 2021-11-11 | Guangdong AnHance Electric Power Technology Co., Ltd | Cable state monitoring method and storage medium |
| CN114818907A (en) * | 2022-04-21 | 2022-07-29 | 广东电网有限责任公司 | State monitoring method, device and equipment of power transmission line and storage medium |
| CN115689114A (en) * | 2022-10-27 | 2023-02-03 | 西南石油大学 | Submarine cable running state prediction method based on combined neural network |
-
2023
- 2023-09-25 CN CN202311234823.6A patent/CN116976530B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070322A (en) * | 2020-09-28 | 2020-12-11 | 国网河北省电力有限公司雄安新区供电公司 | High-voltage cable line running state prediction method based on long-short term memory network |
| CN112257967A (en) * | 2020-12-21 | 2021-01-22 | 中国电力科学研究院有限公司 | Method for predicting state quantity of cable, method and device for determining operating state of cable |
| CN112989601A (en) * | 2021-03-10 | 2021-06-18 | 西南石油大学 | Submarine cable state evaluation method based on subjective and objective combination weighting |
| AU2021106675A4 (en) * | 2021-03-23 | 2021-11-11 | Guangdong AnHance Electric Power Technology Co., Ltd | Cable state monitoring method and storage medium |
| CN113468764A (en) * | 2021-07-22 | 2021-10-01 | 西南石油大学 | Submarine cable state prediction method based on time sequence |
| CN114818907A (en) * | 2022-04-21 | 2022-07-29 | 广东电网有限责任公司 | State monitoring method, device and equipment of power transmission line and storage medium |
| CN115689114A (en) * | 2022-10-27 | 2023-02-03 | 西南石油大学 | Submarine cable running state prediction method based on combined neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116976530A (en) | 2023-10-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Rafiq et al. | Regularized LSTM based deep learning model: first step towards real-time non-intrusive load monitoring | |
| US9679248B2 (en) | Automated creation of semantically-enriched diagnosis models using time series data of temperatures collected by a network of sensors | |
| CN111091233A (en) | Wind power plant short-term wind power prediction modeling method based on wavelet analysis and multi-model AdaBoost depth network | |
| CN110245047B (en) | Time sequence abnormality detection method, device and equipment | |
| CN113093985B (en) | Sensor data link abnormity detection method and device and computer equipment | |
| CN111680712A (en) | Transformer oil temperature prediction method, device and system based on similar moments in the day | |
| Lei et al. | Prediction of PM2. 5 concentration considering temporal and spatial features: A case study of Fushun, Liaoning Province | |
| CN116976530B (en) | Cable equipment state prediction method, device and storage medium | |
| Xu et al. | An improved LOF outlier detection algorithm | |
| CN116665798A (en) | Air pollution trend early warning method and related device | |
| CN114757097B (en) | Line fault diagnosis method and device | |
| CN115796361A (en) | Wind speed interval prediction method and device for ground stage of overhead line engineering | |
| CN115567371A (en) | Abnormity detection method, device, equipment and readable storage medium | |
| Wang et al. | A conditional random field based feature learning framework for battery capacity prediction | |
| Ye et al. | A Novel Self-Supervised Learning-Based Anomalous Node Detection Method Based on an Autoencoder for Wireless Sensor Networks | |
| CN111027680A (en) | Monitoring quantity uncertainty prediction method and system based on variational self-encoder | |
| Marei et al. | CNC Machining Centre Thermal Error Prediction from Temperature Sensors Using Autoencoder-Augmented Long Short Term Memory Networks | |
| CN115982622B (en) | Nuclear reactor coolant system operation transient state rapid identification method, device and system | |
| CN116433050B (en) | Abnormality alarm method and system applied to agricultural big data management system | |
| CN116736174B (en) | Method, apparatus, computer device and storage medium for predicting remaining life of battery | |
| CN116956197B (en) | Deep learning-based energy facility fault prediction method and device and electronic equipment | |
| CN113806172B (en) | Method and device for processing equipment state parameters, electronic equipment and storage medium | |
| CN115408449B (en) | User behavior processing method, device and equipment | |
| CN117493766A (en) | Regional photovoltaic typical daily curve similarity anomaly identification method, device and medium | |
| JP3309554B2 (en) | Chaos time series short-term prediction device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |