CN115796057A - Cable joint temperature prediction method and system based on BAS-LSTM - Google Patents
Cable joint temperature prediction method and system based on BAS-LSTM Download PDFInfo
- Publication number
- CN115796057A CN115796057A CN202310064346.7A CN202310064346A CN115796057A CN 115796057 A CN115796057 A CN 115796057A CN 202310064346 A CN202310064346 A CN 202310064346A CN 115796057 A CN115796057 A CN 115796057A
- Authority
- CN
- China
- Prior art keywords
- lstm
- temperature prediction
- bas
- long
- term memory
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000012549 training Methods 0.000 claims abstract description 56
- 230000007787 long-term memory Effects 0.000 claims abstract description 51
- 230000006403 short-term memory Effects 0.000 claims abstract description 51
- 210000001503 joint Anatomy 0.000 claims abstract description 32
- 241001481710 Cerambycidae Species 0.000 claims abstract description 23
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 230000009467 reduction Effects 0.000 claims description 31
- 238000010845 search algorithm Methods 0.000 claims description 18
- 238000012360 testing method Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 17
- 230000015654 memory Effects 0.000 claims description 15
- 238000005259 measurement Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 abstract description 5
- 238000013528 artificial neural network Methods 0.000 description 5
- 239000013598 vector Substances 0.000 description 5
- 230000008859 change Effects 0.000 description 3
- 238000009413 insulation Methods 0.000 description 3
- 238000000513 principal component analysis Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 150000001875 compounds Chemical class 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000223679 Beauveria Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 239000012774 insulation material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a cable joint temperature prediction method and a system based on BAS-LSTM, which comprises the steps of collecting state parameters related to temperature prediction of a cable middle joint and a terminal joint; preprocessing the state parameters and inputting the state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints; the cable joint temperature prediction model takes state parameters as sample data to train the long and short term memory network model, meanwhile, the best super parameter value of the long and short term memory network model is obtained through optimization of a longicorn searching algorithm, and the best long and short term memory network training model is obtained based on the best super parameter value. The invention considers the internal temperature conditions of the intermediate joint and the terminal joint, optimizes the hyperparameter in the LSTM by combining the BAS, and can realize high-precision temperature prediction.
Description
Technical Field
The invention belongs to the technical field of cable joint temperature prediction, and particularly relates to a cable joint temperature prediction method and system based on BAS-LSTM.
Background
Currently, as power cables can greatly save resources, more and more cable lines are laid, so that cable accessories (including terminals and intermediate connectors) are also increased. In the whole cable line system, the cable terminal, the intermediate joint and the cable are all key factors influencing the safe operation of the system. According to the field operation statistical data, more than 90% of cable operation faults occur at the cable joint position, so that electric field concentration and local temperature rise are caused; when the temperature is too high, the cable insulation material is subjected to electric-thermal insulation breakdown, so that the electric energy supply of industries such as industrial production, people life, traffic, information and the like is influenced, the safety and the reliability of the operation of a power grid are reduced, and huge economic loss of the power grid is caused.
The cable joint temperature is used as an important index, real-time temperature data can be acquired through the acquisition device, and the cable joint temperature has no autonomous prediction capability. Therefore, prediction based on joint temperature data collected in actual operation becomes a hotspot of research in recent years, the insulation level of the cable joint can be estimated in advance through cable joint temperature prediction, faults of the cable can be diagnosed in time, and a foundation is provided for controlling cost and ensuring system safety. Therefore, the research on the temperature prediction method of the cable middle-terminal joint is of great significance.
The prior art provides a temperature prediction method based on a BP neural network and a Kalman algorithm, but most of the prior art directly inputs cable surface temperature data into a prediction model for prediction, and due to the fact that the cable surface temperature and the actual joint internal temperature are different, the prediction precision of the conventional method is not ideal.
Disclosure of Invention
In view of the above, the present invention is directed to solve the problem that the prediction accuracy is not accurate enough when the joint prediction is performed by using the cable surface temperature data in the conventional cable joint temperature prediction method.
In order to solve the technical problems, the invention provides the following technical scheme:
in a first aspect, the invention provides a cable joint temperature prediction method based on BAS-LSTM, comprising the following steps:
acquiring state parameters related to temperature prediction of a cable intermediate joint and a terminal joint;
preprocessing the state parameters and inputting the state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints;
the BAS-LSTM-based cable joint temperature prediction model takes state parameters as sample data, trains the long and short term memory network model, optimizes and obtains the optimal super parameter value of the long and short term memory network model by utilizing a longicorn whisker search algorithm, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and serves as the BAS-LSTM-based cable joint temperature prediction model.
Further, the state parameters specifically include:
ambient temperature, humidity, core to sheath current ratio and historical temperature, wherein, ambient temperature includes the highest, the lowest and average ambient temperature of day, and historical temperature includes the surface of cable intermediate head and terminal joint and inside measurement temperature.
Further, the state parameters are preprocessed, specifically:
and performing dimensionality reduction on the state parameters by using a WPCA characteristic dimensionality reduction method, and obtaining characteristic data subjected to WPCA dimensionality reduction when the contribution degree of a certain dimension characteristic is greater than a contribution degree threshold value.
Further, the state parameters are preprocessed and then input into a cable joint temperature prediction model based on BAS-LSTM, so as to obtain temperature prediction results of cable middle and terminal joints, and the method specifically comprises the following steps:
performing WPCA characteristic dimension reduction processing by taking the state parameters as sample data;
normalizing the feature data subjected to dimension reduction and dividing the data;
dividing the processed characteristic data into a training set and a testing set, and training the long-term and short-term memory network model;
continuously updating and iteratively calculating the optimal longicorn position by using the root mean square error in the training set as a fitness value and using a longicorn stigma search algorithm;
assigning the optimal position to a long-term and short-term memory network model;
training the long-short term memory network model after assignment to obtain an optimal long-short term memory network training model;
testing the prediction precision of the optimal long-term and short-term memory network training model by using the test set;
and performing inverse normalization processing on the output quantity of the optimal long and short term memory network training model to obtain temperature prediction results of cable middle and terminal joints.
In a second aspect, the present invention provides a BAS-LSTM based cable joint temperature prediction system, comprising:
the acquisition unit is used for acquiring state parameters related to temperature prediction of the cable intermediate joint and the terminal joint;
the prediction unit is used for preprocessing the state parameters and inputting the preprocessed state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of the cable middle and terminal joints;
the BAS-LSTM-based cable joint temperature prediction model takes state parameters as sample data, trains the long and short term memory network model, optimizes and obtains the optimal super parameter value of the long and short term memory network model by utilizing a longicorn whisker search algorithm, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and serves as the BAS-LSTM-based cable joint temperature prediction model.
Further, in the acquisition unit, the state parameters specifically include:
ambient temperature, humidity, core to sheath current ratio and historical temperature, wherein, ambient temperature includes the highest, the lowest and average ambient temperature of day, and historical temperature includes the surface of cable intermediate head and terminal joint and inside measurement temperature.
Further, in the prediction unit, the state parameter is preprocessed, specifically:
and performing dimensionality reduction on the state parameters by using a WPCA characteristic dimensionality reduction method, and obtaining characteristic data subjected to WPCA dimensionality reduction when the contribution degree of a certain dimension characteristic is greater than a contribution degree threshold value.
Further, in the prediction unit, the state parameters are preprocessed and then input into a cable joint temperature prediction model based on BAS-LSTM, so as to obtain temperature prediction results of cable middle and terminal joints, specifically including:
performing WPCA characteristic dimension reduction processing by taking the state parameters as sample data;
normalizing and dividing the feature data subjected to dimension reduction;
dividing the processed characteristic data into a training set and a testing set, and training the long-term and short-term memory network model;
continuously updating and iteratively calculating the optimal longicorn position by using the root mean square error in the training set as a fitness value and using a longicorn stigma search algorithm;
assigning the optimal position to a long-term and short-term memory network model;
training the long and short term memory network model after assignment to obtain an optimal long and short term memory network training model;
testing the prediction precision of the optimal long-term and short-term memory network training model by using the test set;
and performing inverse normalization processing on the output quantity of the optimal long-short term memory network training model to obtain temperature prediction results of the cable middle and terminal joints.
In summary, the invention provides a cable joint temperature prediction method and system based on BAS-LSTM, comprising collecting state parameters related to temperature prediction of a cable middle joint and a terminal joint; preprocessing the state parameters and inputting the state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints; the BAS-LSTM-based cable joint temperature prediction model takes state parameters as sample data, trains the long and short term memory network model, optimizes and obtains the optimal super parameter value of the long and short term memory network model by utilizing a longicorn whisker search algorithm, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and serves as the BAS-LSTM-based cable joint temperature prediction model. The invention considers the internal temperature conditions of the intermediate joint and the terminal joint, optimizes the hyper-parameters in the LSTM by combining the BAS, and can realize high-precision temperature prediction.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a BAS-LSTM-based cable joint temperature prediction method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an LSTM network according to an embodiment of the present invention.
Detailed Description
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
For temperature prediction of cable joints, the prior art provides a power cable temperature early warning system based on data analysis and temperature prediction, comprising: the on-line monitoring equipment is used for acquiring and storing the ambient temperature, the ambient humidity, the sheath or core current and the cable joint temperature; the temperature prediction device is connected with the online monitoring device and used for predicting the temperature of the cable joint on the day according to the acquired environmental temperature, environmental humidity and sheath or core current on the day to be predicted on the basis of the temperature prediction model; and the early warning equipment is respectively connected with the online monitoring equipment and the temperature prediction equipment and is used for early warning based on a predicted value and an actual measurement value of the temperature of the cable joint on the day to be predicted. Compared with the prior art, the method has the advantages that the historical data is fully utilized to obtain the predicted temperature, the predicted temperature and the measured temperature are combined to carry out early warning, the data change trend is analyzed through data mining, the temperature of the cable is predicted, the temperature auxiliary criterion is researched, the early warning system is established, the insulation level of the cable joint is predicted in advance, and effective early warning and forecasting are realized.
The method for predicting the temperature of the cable joint wire based on the GA optimized serial BP network further comprises the following steps: selecting and normalizing related data; establishing a training model; and predicting by using the trained model. The innovation of the invention is that two BP neural networks are established in series in the modeling process, and the reflection factor data and the prediction result are trained and learned respectively, wherein the output vector of the network 1, namely the prediction result of the reflection factor data, is used as the input vector of the network 2, and the weight and the threshold of the two networks are optimized by using a genetic algorithm, so that the group of target values can be predicted more accurately under the condition of no reflection factor data, namely, the prediction model uses the network 1 to train the reflection factor data to obtain the temperature value at the corresponding moment, then the network 2 trains the prediction results of the reflection factor data at three continuous moments to obtain the temperature value at the fourth moment, and the whole process of solving the temperature value at the fourth moment does not need the reflection factor data at the moment. The model has strong learning capacity of the BP neural network, combines excellent global search capacity of a genetic algorithm, and simultaneously, the serial fusion of the two BP neural networks ensures that the prediction performance of the model is more excellent.
The cable joint surface temperature prediction method based on the Kalman algorithm is also provided, and is an algorithm for predicting the temperature of the cable joint surface at the next moment and early warning the working state of the cable joint. Firstly, measuring by a temperature sensor to obtain surface temperature data of a cable joint, and respectively adopting a three-section type fixed threshold detection algorithm and a temperature change rate detection algorithm to carry out operation processing on the surface temperature data; and then, the surface temperature of the cable joint at the next moment is estimated through a Kalman algorithm, so that the time difference of the transmission of the cable core temperature at the cable joint to the surface of the cable joint is made up, and the time is won for the rush repair of the cable fault.
However, for cable joint temperature prediction, hot spot temperature data is usually obtained on the cable surface, and the internal temperature conditions of the intermediate joint and the terminal joint are ignored, so that the prediction result of the temperature prediction model is different from the actual internal temperature of the joint, which may cause potential safety hazard. In addition, most of the multidimensional historical data are directly input into a prediction model for training, effective analysis on the data is not performed, and the prediction model is not improved, so that the accuracy of temperature prediction is not ideal.
Accordingly, the present invention provides a BAS-LSTM based cable joint temperature prediction method and system. In one embodiment of the invention, the high-precision cable middle-terminal joint temperature prediction is realized by collecting various parameters related to the cable middle-terminal joint temperature prediction, improving an LSTM prediction model by adopting a Tianniu whisker search algorithm and inputting various parameters into the BAS-LSTM-based temperature prediction model.
An embodiment of the BAS-LSTM based cable joint temperature prediction method of the present invention is described in detail below.
Referring to fig. 1, the present embodiment provides a method for predicting a temperature of a cable joint based on BAS-LSTM, including the following steps:
s100: acquiring state parameters related to temperature prediction of a cable intermediate joint and a terminal joint;
s200: preprocessing the state parameters and inputting the state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints;
the BAS-LSTM-based cable joint temperature prediction model takes state parameters as sample data, trains the long and short term memory network model, optimizes and obtains the optimal super parameter value of the long and short term memory network model by utilizing a longicorn whisker search algorithm, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and serves as the BAS-LSTM-based cable joint temperature prediction model.
The embodiment provides a cable joint temperature prediction method based on BAS-LSTM, which comprises the steps of collecting state parameters related to temperature prediction of a cable middle joint and a terminal joint; preprocessing the state parameters and inputting the state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints; the BAS-LSTM-based cable joint temperature prediction model takes state parameters as sample data, trains the long and short term memory network model, optimizes and obtains the optimal super parameter value of the long and short term memory network model by utilizing a longicorn whisker search algorithm, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and serves as the BAS-LSTM-based cable joint temperature prediction model. The invention considers the internal temperature conditions of the intermediate joint and the terminal joint, optimizes the hyper-parameters in the LSTM by combining the BAS, and can realize high-precision temperature prediction.
In an alternative embodiment, the collected state parameters related to the temperature prediction of the cable middle-terminal joint comprise ambient temperature, humidity, core/sheath current ratio and historical temperature, the ambient temperature can be divided into three types of highest ambient temperature, lowest ambient temperature and average ambient temperature on the day, the historical temperature comprises surface and internal measured temperature of the cable middle-terminal joint, and the measured time is 24 hours at each integral point, so that input state parameter vectors of a prediction model are jointly formed, and the total number of the input state parameter vectors is 53.
In an alternative embodiment, the WPCA characteristic dimension reduction is performed on the collected state parameters related to the temperature prediction of the cable middle-terminal joint.
WPCA introduces a weight coefficient to each dimension of characteristics on the basis of Principal Component Analysis (PCA) so as to distinguish the importance of different characteristics, and the calculation formula is as follows:
in the formula, wi (i =1,2, \ 8230;, k) represents weighting coefficients of k features, and fj (j =1,2, \ 8230;, n) represents categories of n samples; obtaining wi between the sample and the label by a least square method fitting method, wherein the optimal solution solving formula is as follows:
then, the weighting coefficient wi and the original input data are weighted to obtain:
finally at Y n×k Based on the above, PCA dimensionality reduction is performed (this part is a common method in the field, and therefore will not be described in detail), and the contribution degree of the t-dimensional feature is obtained. And when the contribution degree of the t-dimensional feature is greater than the contribution degree threshold value, obtaining the feature data after WPCA dimension reduction.
In an alternative embodiment, the BAS-LSTM based cable joint temperature prediction model includes both the LSTM algorithm for temperature prediction and the BAS algorithm for optimizing the hyper-parameters of the LSTM model. Wherein, the long-short term memory neural network (LSTM) makes the weight change continuously mainly through input, output and forgetting gate, the structure is shown in FIG. 2,S t-1 andh t-1 respectively representing the cell state and output at time t-1,x t andh t respectively representtThe input of the time of day and the state output of the cell,S t the output of the final LSTM.
Since some hyper-parameters in the LSTM have great influence on the model performance, the prediction accuracy of the model is directly influenced if the initial values of the hyper-parameters cannot be accurately obtained. The method takes the root mean square error in the LSTM model training process as a fitness value, and obtains the optimal super parameter value by utilizing the optimization of a celestial cow whisker search algorithm (BAS).
The steps of temperature prediction based on the cable joint temperature prediction model of BAS-LSTM are shown in fig. 1, and include:
1) Performing WPCA characteristic dimension reduction processing by taking the state parameters as sample data;
2) Normalizing and dividing the feature data subjected to dimension reduction;
3) Dividing the processed characteristic data into a training set and a testing set, and training the long-term and short-term memory network model;
4) Using the root mean square error in the training set as a fitness value, and continuously updating and iteratively calculating the optimal position of the longicorn by using a longicorn stigma search algorithm;
5) Assigning the optimal position to a long-term and short-term memory network model;
6) Training the long and short term memory network model after assignment to obtain an optimal long and short term memory network training model;
7) Testing the prediction precision of the optimal long-term and short-term memory network training model by using the test set;
8) And performing inverse normalization processing on the output quantity of the optimal long-short term memory network training model to obtain temperature prediction results of the cable middle and terminal joints.
The method comprises the following steps of obtaining the optimal hyper-parameter value by utilizing a celestial cow whisker search algorithm (BAS) optimization:
1) According to the following
Standardizing the random motion direction of each longicorn;
2) Establishing coordinates of the left side and the right side of the generated longicorn stigma:
in the formula (I), the compound is shown in the specification,x pt for optimizingtThe position of the second right hair is the position of the second right hair,x qt for optimizingtThe position of the next left hair is the position of the left hair,x t for optimizingtThe position of the second-to-last centroid,d 0 the distance between two whiskers. Setting a fitness objective functionf(x)And (3) updating the space coordinates after the food is sensed corresponding to the perceptibility of the two whiskers:
in the formula (I), the compound is shown in the specification,s t is as followstThe step size of the sub-optimization,sgis a symbolic function.
3) Changing the distance and step length of each optimization, continuously iterating, updating and calculating the optimal longicorn position, and when the iteration times reach the maximum iteration timest max And (5) stopping, and assigning the optimal longicorn position to the LSTM model.
When the prediction method provided by the embodiment is applied to a certain example, three indexes, namely a Mean Absolute Percentage Error (MAPE), a Root Mean Square Error (RMSE) and a Mean Absolute Error (MAE), are adopted to evaluate a model prediction result, and a cable joint temperature prediction result predicted only by using LSTM is shown in table 1.
TABLE 1 comparison of cable joint temperature prediction error results
Therefore, the prediction method provided by the embodiment has higher accuracy. Considering that most of the cable joint temperature data are usually obtained on the surface of a cable at present, the prediction method provided by the embodiment takes the internal temperatures of the intermediate joint and the terminal joint and other parameters for prediction as input parameters, and extracts effective characteristic parameters by using a WPCA dimension reduction method to construct a cable intermediate-terminal joint temperature input state parameter vector; in addition, a new prediction model of BAS-LSTM middle-terminal joint temperature is established by utilizing a Beauveria tiannikov search algorithm (BAS) to adaptively optimize hyper-parameters in the LSTM, so that multi-dimensional and high-precision temperature prediction is realized, a basis is provided for diagnosing cables, the operation efficiency and reliability of the cables are improved, and the electric energy supply efficiency and the overall economy of a power system are improved.
While the foregoing is a detailed description of one embodiment of the BAS-LSTM based cable joint temperature prediction method of the present invention, another embodiment of the BAS-LSTM based cable joint temperature prediction system of the present invention will be described in detail below.
The embodiment provides a BAS-LSTM-based cable joint temperature prediction system which comprises an acquisition unit and a prediction unit.
In this embodiment, the acquisition unit is configured to acquire state parameters related to the prediction of the temperature of the cable intermediate joint and the terminal joint.
In the acquisition unit, the state parameters specifically include:
ambient temperature, humidity, core to sheath current ratio and historical temperature, wherein, ambient temperature includes the highest, the lowest and average ambient temperature of day, and historical temperature includes the surface of cable intermediate head and terminal joint and inside measurement temperature.
In this embodiment, the prediction unit is configured to input the preprocessed state parameters into a cable joint temperature prediction model based on BAS-LSTM, and obtain temperature prediction results of the cable middle and terminal joints;
the BAS-LSTM-based cable joint temperature prediction model takes state parameters as sample data, trains the long and short term memory network model, optimizes and obtains the optimal super parameter value of the long and short term memory network model by utilizing a longicorn whisker search algorithm, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and serves as the BAS-LSTM-based cable joint temperature prediction model.
Further, in the prediction unit, the state parameter is preprocessed, specifically:
and performing dimensionality reduction on the state parameters by using a WPCA characteristic dimensionality reduction method, and obtaining characteristic data subjected to WPCA dimensionality reduction when the contribution degree of a certain dimension characteristic is greater than a contribution degree threshold value.
Preprocessing the state parameters and inputting the state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints, wherein the method specifically comprises the following steps:
performing WPCA characteristic dimension reduction processing by taking the state parameters as sample data;
normalizing the feature data subjected to dimension reduction and dividing the data;
dividing the processed characteristic data into a training set and a test set, and training the long-short term memory network model;
continuously updating and iteratively calculating the optimal longicorn position by using the root mean square error in the training set as a fitness value and using a longicorn stigma search algorithm;
assigning the optimal position to a long-term and short-term memory network model;
training the long and short term memory network model after assignment to obtain an optimal long and short term memory network training model;
testing the prediction precision of the optimal long-term and short-term memory network training model by using the test set;
and performing inverse normalization processing on the output quantity of the optimal long-short term memory network training model to obtain temperature prediction results of the cable middle and terminal joints.
It should be noted that the prediction system provided in this embodiment is used to implement the prediction method provided in the foregoing embodiment, and the specific settings of each unit are based on complete implementation of the method, which is not described herein again.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. The cable joint temperature prediction method based on BAS-LSTM is characterized by comprising the following steps:
acquiring state parameters related to temperature prediction of a cable intermediate joint and a terminal joint;
preprocessing the state parameters and inputting the preprocessed state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of cable middle and terminal joints;
the BAS-LSTM-based cable joint temperature prediction model takes the state parameters as sample data, trains a long and short term memory network model, optimizes by utilizing a long and short term whisker search algorithm to obtain the optimal super parameter value of the long and short term memory network model, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and takes the optimal long and short term memory network training model as the BAS-LSTM-based cable joint temperature prediction model.
2. The BAS-LSTM-based cable joint temperature prediction method of claim 1, wherein the state parameters specifically include:
ambient temperature, humidity, sinle silk and sheath current ratio and historical temperature, wherein, ambient temperature is including the highest, minimum and average ambient temperature of this day, historical temperature includes cable intermediate head and terminal joint's surface and inside measurement temperature.
3. The BAS-LSTM-based cable joint temperature prediction method of claim 1, wherein the state parameters are pre-processed, specifically:
and performing dimension reduction processing on the state parameters by adopting a WPCA (wi-Fi protected setup) feature dimension reduction method, and obtaining feature data subjected to the WPCA dimension reduction processing when the contribution degree of a certain dimension feature is greater than a contribution degree threshold value.
4. The BAS-LSTM-based cable joint temperature prediction method of claim 1, wherein the preprocessing of the state parameters is input into a BAS-LSTM-based cable joint temperature prediction model to obtain temperature prediction results for cable intermediate and terminal joints, specifically comprising:
performing WPCA characteristic dimension reduction processing by taking the state parameter as sample data;
normalizing the feature data subjected to dimension reduction and dividing the data;
dividing the processed characteristic data into a training set and a testing set, and training the long-short term memory network model;
continuously updating and iteratively calculating the optimal longicorn position by using the root mean square error in the training set as a fitness value and using a longicorn stigma search algorithm;
assigning the optimal position to the long-term and short-term memory network model;
training the long-short term memory network model after assignment to obtain the optimal long-short term memory network training model;
testing the prediction precision of the optimal long-short term memory network training model by using the test set;
and performing reverse normalization processing on the output quantity of the optimal long and short term memory network training model to obtain temperature prediction results of the cable middle and terminal joints.
5. BAS-LSTM based cable joint temperature prediction system, comprising:
the acquisition unit is used for acquiring state parameters related to temperature prediction of the cable intermediate joint and the terminal joint;
the prediction unit is used for preprocessing the state parameters and inputting the preprocessed state parameters into a cable joint temperature prediction model based on BAS-LSTM to obtain temperature prediction results of the cable middle and terminal joints;
the BAS-LSTM-based cable joint temperature prediction model takes the state parameters as sample data, trains a long and short term memory network model, optimizes by utilizing a long and short term whisker search algorithm to obtain the optimal super parameter value of the long and short term memory network model, obtains the optimal long and short term memory network training model based on the optimal super parameter value, and takes the optimal long and short term memory network training model as the BAS-LSTM-based cable joint temperature prediction model.
6. The BAS-LSTM based cable joint temperature prediction system of claim 5, wherein in the acquisition unit, the state parameters specifically include:
ambient temperature, humidity, sinle silk and sheath current ratio and historical temperature, wherein, ambient temperature is including the highest, minimum and average ambient temperature of this day, historical temperature includes cable intermediate head and terminal joint's surface and inside measurement temperature.
7. The BAS-LSTM-based cable joint temperature prediction system of claim 5, wherein said state parameters are preprocessed in said prediction unit, specifically:
and performing dimension reduction processing on the state parameters by adopting a WPCA (wi-Fi protected setup) feature dimension reduction method, and obtaining feature data subjected to the WPCA dimension reduction processing when the contribution degree of a certain dimension feature is greater than a contribution degree threshold value.
8. The BAS-LSTM-based cable joint temperature prediction system of claim 5, wherein the prediction unit is further configured to preprocess the state parameters and input the preprocessed state parameters into a BAS-LSTM-based cable joint temperature prediction model to obtain the temperature prediction results of the cable middle and end joints, and specifically comprises:
performing WPCA characteristic dimension reduction processing by taking the state parameter as sample data;
normalizing and dividing the feature data subjected to dimension reduction;
dividing the processed characteristic data into a training set and a test set, and training the long-short term memory network model;
continuously updating and iteratively calculating the optimal longicorn position by using the root mean square error in the training set as a fitness value and using a longicorn stigma search algorithm;
assigning the optimal position to the long-term and short-term memory network model;
training the long-short term memory network model after assignment to obtain the optimal long-short term memory network training model;
testing the prediction precision of the optimal long-short term memory network training model by using the test set;
and performing reverse normalization processing on the output quantity of the optimal long and short term memory network training model to obtain temperature prediction results of the cable middle and terminal joints.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310064346.7A CN115796057A (en) | 2023-02-06 | 2023-02-06 | Cable joint temperature prediction method and system based on BAS-LSTM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310064346.7A CN115796057A (en) | 2023-02-06 | 2023-02-06 | Cable joint temperature prediction method and system based on BAS-LSTM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115796057A true CN115796057A (en) | 2023-03-14 |
Family
ID=85429883
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310064346.7A Pending CN115796057A (en) | 2023-02-06 | 2023-02-06 | Cable joint temperature prediction method and system based on BAS-LSTM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115796057A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
| CN107103184A (en) * | 2017-03-28 | 2017-08-29 | 国网上海市电力公司 | A kind of high-voltage cable joint temperature predicting method |
| CN112215412A (en) * | 2020-09-27 | 2021-01-12 | 中国农业大学 | Dissolved oxygen prediction method and device |
| CN115112171A (en) * | 2022-06-20 | 2022-09-27 | 国网山东省电力公司曲阜市供电公司 | Cable intermediate joint monitoring method and system |
| CN115577643A (en) * | 2022-11-23 | 2023-01-06 | 广东电网有限责任公司中山供电局 | Temperature prediction method and device for cable terminal |
| CN115586468A (en) * | 2022-09-27 | 2023-01-10 | 福州亿力电力工程有限公司配电工程分公司 | Cable intermediate head comprehensive fault on-line monitoring and early warning system |
-
2023
- 2023-02-06 CN CN202310064346.7A patent/CN115796057A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
| CN107103184A (en) * | 2017-03-28 | 2017-08-29 | 国网上海市电力公司 | A kind of high-voltage cable joint temperature predicting method |
| CN112215412A (en) * | 2020-09-27 | 2021-01-12 | 中国农业大学 | Dissolved oxygen prediction method and device |
| CN115112171A (en) * | 2022-06-20 | 2022-09-27 | 国网山东省电力公司曲阜市供电公司 | Cable intermediate joint monitoring method and system |
| CN115586468A (en) * | 2022-09-27 | 2023-01-10 | 福州亿力电力工程有限公司配电工程分公司 | Cable intermediate head comprehensive fault on-line monitoring and early warning system |
| CN115577643A (en) * | 2022-11-23 | 2023-01-06 | 广东电网有限责任公司中山供电局 | Temperature prediction method and device for cable terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108022001B (en) | Short-term load probability density prediction method based on PCA (principal component analysis) and quantile regression forest | |
| CN110765700A (en) | Ultrahigh voltage transmission line loss prediction method based on quantum ant colony optimization RBF network | |
| CN102663412B (en) | Power equipment current-carrying fault trend prediction method based on least squares support vector machine | |
| CN112765873B (en) | LSTM algorithm-based power transformation equipment heating prediction method | |
| CN110685868A (en) | Wind turbine generator fault detection method and device based on improved gradient elevator | |
| CN111461922B (en) | Real-time prediction method for hot spot temperature of transformer based on extreme learning machine | |
| CN111401657B (en) | Transformer hot spot temperature time sequence prediction method based on support vector regression algorithm | |
| CN107103184A (en) | A kind of high-voltage cable joint temperature predicting method | |
| CN113484774B (en) | Lithium battery pack capacity estimation method based on temperature calibration and neural network | |
| CN110837915A (en) | Low-voltage load point prediction and probability prediction method for power system based on hybrid integrated deep learning | |
| Saatloo et al. | Hierarchical extreme learning machine enabled dynamic line rating forecasting | |
| CN114994537A (en) | EKF-LSTM-based lithium battery state of charge joint prediction measurement method | |
| CN112541634B (en) | Method and device for predicting top-layer oil temperature and discriminating false alarm and storage medium | |
| CN114357670A (en) | Power distribution network power consumption data abnormity early warning method based on BLS and self-encoder | |
| CN117575247A (en) | Province-station scheduling management method and system | |
| CN117113243B (en) | Photovoltaic equipment abnormality detection method | |
| CN115796057A (en) | Cable joint temperature prediction method and system based on BAS-LSTM | |
| CN112580254A (en) | Multi-dimensional analysis and early warning method for converter transformer based on multi-level decision fusion | |
| Sahoo et al. | Health index analysis of XLPE cable insulation using machine learning technique | |
| El-Hawary | Artificial neural networks and possible applications to desalination | |
| CN116738355A (en) | Electric power multi-mode feature level fusion method considering data difference | |
| CN115993537A (en) | Lithium battery capacity prediction method based on correlation analysis and WOA-LSTM | |
| CN115017828A (en) | Power cable fault identification method and system based on bidirectional long-short-time memory network | |
| CN107742886B (en) | Prediction method for load peak simultaneous coefficient of thermoelectric combined system | |
| Zhang et al. | Prediction of Power Outages in Distribution Network with Grey Theory |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230314 |
|
| RJ01 | Rejection of invention patent application after publication |