CN114169237A - Power cable joint temperature abnormity early warning method combining EEMD-LSTM and isolated forest algorithm - Google Patents
Power cable joint temperature abnormity early warning method combining EEMD-LSTM and isolated forest algorithm Download PDFInfo
- Publication number
- CN114169237A CN114169237A CN202111473398.7A CN202111473398A CN114169237A CN 114169237 A CN114169237 A CN 114169237A CN 202111473398 A CN202111473398 A CN 202111473398A CN 114169237 A CN114169237 A CN 114169237A
- Authority
- CN
- China
- Prior art keywords
- temperature
- joint
- sequence
- lstm
- detection
- 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.)
- Granted
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000001514 detection method Methods 0.000 claims abstract description 55
- 238000012544 monitoring process Methods 0.000 claims abstract description 16
- 230000002159 abnormal effect Effects 0.000 claims abstract description 14
- 230000008859 change Effects 0.000 claims abstract description 5
- 238000000354 decomposition reaction Methods 0.000 claims description 24
- 238000012360 testing method Methods 0.000 claims description 13
- 230000009466 transformation Effects 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 230000005856 abnormality Effects 0.000 claims description 6
- 238000013145 classification model Methods 0.000 claims description 5
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 210000004027 cell Anatomy 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 229920003020 cross-linked polyethylene Polymers 0.000 description 2
- 239000004703 cross-linked polyethylene Substances 0.000 description 2
- 210000001503 joint Anatomy 0.000 description 2
- IYLGZMTXKJYONK-ACLXAEORSA-N (12s,15r)-15-hydroxy-11,16-dioxo-15,20-dihydrosenecionan-12-yl acetate Chemical compound O1C(=O)[C@](CC)(O)C[C@@H](C)[C@](C)(OC(C)=O)C(=O)OCC2=CCN3[C@H]2[C@H]1CC3 IYLGZMTXKJYONK-ACLXAEORSA-N 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- IYLGZMTXKJYONK-UHFFFAOYSA-N ruwenine Natural products O1C(=O)C(CC)(O)CC(C)C(C)(OC(C)=O)C(=O)OCC2=CCN3C2C1CC3 IYLGZMTXKJYONK-UHFFFAOYSA-N 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Geometry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention provides a power cable joint temperature abnormity early warning method combining EEMD-LSTM and isolated forest algorithm, aiming at a temperature hot spot power cable joint of a medium and low voltage distribution network, carrying out temperature abnormity early warning; establishing an EEMD-LSTM temperature prediction model according to joint historical temperature monitoring data; decomposing an original temperature data sequence into a plurality of subsequences with smaller magnitude by using the EEMD, and extracting joint temperature change trend information; predicting each subsequence through LSTM, reconstructing the prediction result of the subsequence, and outputting the predicted value of the temperature of the joint at the future time; carrying out temperature anomaly detection on the temperature index predicted by the EEMD-LSTM by adopting an isolated forest algorithm; establishing a plurality of sub-detection classifiers, and combining the joint surface temperature, the core temperature and the relative temperature difference in pairs to obtain three groups of classifiers; and finally, early warning of the joint temperature is realized. The method provided by the invention can be used for pre-judging the potential abnormal temperature of the joint in time and carrying out temperature early warning on the cable joint with overhigh temperature and too fast temperature rise.
Description
Technical Field
The invention relates to the field of power cable line fault detection, in particular to a power cable joint temperature abnormity early warning method combining EEMD-LSTM and an isolated forest algorithm.
Background
Along with the acceleration of industrialization and urbanization in China, the proportion of power cable lines in a power transmission and distribution network is continuously increased, and crosslinked polyethylene (XLPE) cables with the voltage level of 35kV and below are more and more widely applied.
The safe and stable operation of the power cable line is vital, and when the cable line breaks down, power failure can occur, and joint explosion can occur in serious conditions, so that fire disasters and casualties are caused. The cable intermediate joint is one of main accessories of a cable line, is used for connecting cables at two ends and is the weakest link of the cable line. Through accident statistical analysis, the occurrence of the middle joint fault is not as sudden as a line short-circuit fault, but is a continuously accumulated process. The faults that are typically present accumulate a significant amount of joule heating, causing the cable temperature to rise continuously. The system is overloaded for a long time, and when the applied load exceeds the current-carrying capacity of the cable, the temperature at the joint is also increased, so that potential safety hazards exist. The temperature is taken as a non-electrical quantity, can only reflect the insulation performance of the cable and the running load state of a system, is easy to measure and characterize, and is considered as an important cable condition detection means at home and abroad. The existing temperature detection technologies such as a digital temperature sensor and a fiber grating can accurately detect the temperature of the cable. The cable temperature on-line monitoring system is developed by combining a computer technology and an electronic technology, and the temperature of the cable joint and other indexes are monitored in real time. Meanwhile, the phenomena of connector temperature data transmission lag and insufficient data utilization exist, and the problem of overhigh maintenance cost caused by the complex cable line of the actual distribution network is solved.
Disclosure of Invention
The invention aims to solve the technical problems of wrong judgment and missed judgment of single threshold early warning according to relevant standards and detection experiences; based on the distribution network equipment state evaluation guide (Q/GDW 645-2011), the problems of excessive indexes and high index acquisition difficulty are solved. The LSTM neural network and the isolated forest algorithm are combined to solve the problems.
A power cable joint temperature abnormity early warning method combining EEMD-LSTM and isolated forest algorithm comprises the following steps:
step 3, establishing an LSTM network unit, optimizing LSTM network unit parameters, predicting each decomposed subsequence by using the LSTM network unit, reconstructing a subsequence prediction result, and outputting a joint temperature value at a future moment;
step 4, extracting a joint temperature index data set, performing characteristic transformation on the original data set by using haar wavelet transformation, and extracting temperature change trend information to obtain a temperature detection sample;
step 5, setting attention values and early warning values corresponding to all temperature indexes of the joint;
step 6, detecting temperature anomaly by adopting an isolated forest algorithm, dividing a detection sample into a training set and a testing set, carrying out anomaly marking on the training set, and establishing a detection classifier of joint surface temperature-core temperature, surface temperature-relative temperature difference and core temperature-relative temperature difference;
and 7, carrying out temperature anomaly detection on the test set by the established detection model according to the steps 3, 4, 5 and 6, and realizing joint temperature early warning.
Further, the integrated empirical mode decomposition described in step 2 is specifically expressed as:
the procedure for processing the linker temperature sequence using EEMD is shown below:
(1) setting zero mean Gaussian white noise with amplitude of k and decomposition repetition frequency of M, wherein the noise obeys normal distribution;
(2) adding the Gaussian white noise n (t) into a cable joint temperature initial sequence x (t) to perform an mth EMD decomposition;
1) searching all local minimum points and local maximum points in the sequence after noise is added, obtaining an upper envelope line sequence u (t) and a lower envelope line sequence v (t) by using 3 times of spline interpolation function fitting, and taking the mean value of the upper envelope line sequence u (t) and the lower envelope line sequence v (t) to obtain a mean value sequence m (t) (u (t)) + v (t))/2;
2) subtracting the mean sequence from the sequence to which noise is added to obtain an intermediate sequence h (t) ═ x (t) — m (t);
3) determining whether h (t) satisfies the eigenmode function (IMF), if yes, defining it as IMF1, otherwise, regarding it as a new x (t) and repeating the steps 1) -2). By parity of reasoning, the mth EMD decomposition is completed;
4) the residual component r (t) x (t) -h (t) is calculated. Repeating the steps 1) -3) on the residual component, and stopping when the decomposition termination condition is met;
(3) and performing ensemble averaging on the IMFs obtained by the M times of EMD decomposition to offset the amplitude influence caused by the added white noise.
Further, the establishment of the LSTM network element in step 3 is expressed as:
the sub-sequence decomposed in step 2 is predicted using the LSTM network element. The LSTM network element parameters are optimized. In the LSTM network structure parameters, the single-step prediction node of an output layer is 1, and the window length is 24; the number of the three-step prediction nodes is 3, and the window length is 72. The input window length is the length of the historical sequence used to make the joint temperature prediction. The use of a single LSTM layer can control the computational cost within a reasonable range,and at the same time, the network performance requirements are met, so the number of hidden layers is 1. During each training, after the lengths of an input layer and an output layer are determined, a particle swarm algorithm is used for searching a neuron dimension cell _ num of a hidden layer, a network learning rate factor c and a batch size Sbatch. And predicting each sub-sequence obtained by decomposition by using an LSTM network unit, reconstructing a sub-sequence prediction result, and outputting a joint temperature value at a future moment.
Further, in step 4, the original data set is subjected to feature transformation by haar wavelet transformation, temperature change trend information is extracted, and a temperature detection sample is obtained, wherein the specific process comprises the following steps:
temperature index in a monitoring sample, and joint wire core temperature TjcSurface temperature TbcInternal and external temperature difference delta TcThe data of (2) are extracted to obtain respective one-dimensional matrixes.
Tjc=[Tjc1 Tjc2 … Tjci … Tjcn]
Tbc=[Tbc1 Tbc2 … Tbci … Tbcn]
ΔTc=[Tc1 Tc2 … Tci … Tcn]
(1) Calculating the average value A of adjacent datajc、Abc、Ac。
(2) And calculating the difference value of the original data and the mean value. And calculating an average value, losing part of information and making the matrix size be half of the original size. For storing the missing part of the detail information, the calculated average value is subtracted from the previous data of the adjacent data pair to obtain the difference Djc、Dbc、Dc. When the temperature acquisition period is short, this difference can be regarded as the temperature rise rate. And storing the calculated mean value and difference value to obtain a new matrix after transformation, wherein the new matrix is used as a characteristic vector of the joint temperature and is expressed as:
Hjc=[Ajc1 Ajc2 … Ajci … Ajc(n+1)/2 Djc1 Djc2 … Djci … Djc(n+1)/2]
Hbc=[Abc1 Abc2 … Abci … Abc(n+1)/2 Dbc1 Dbc2 … Dbci … Dbc(n+1)/2]
Hc=[Ac1 Ac2 … Aci … Ac(n+1)/2 Dc1 Dc2 … Dci … Dc(n+1)/2]
combining the original data, the finally obtained detection vector group is as follows:
Xjc=[Tjc Hjc]T
Xbc=[Tbc Hbc]T
Xc=[ΔTc Hc]T
further, the setting of the attention value and the early warning value corresponding to each temperature index of the joint in the step 5 is specifically expressed as:
according to the operation experience, attention values and early warning values are set:
further, the step 6 of detecting the temperature abnormality by using the isolated forest algorithm specifically includes:
the isolated forest algorithm is set as a standard parameter: sample size NumsubIs 256; the tree depth, HeightLimit, is 8; number Num of isolated forest treestreeIs 100; the maximum number of features is 1. And after the setting of the index threshold is finished, selecting different classification models. And combining the transformed complete sample and the sample obtained by combining the temperature of the wire core of the joint, the surface temperature of the joint and the relative internal and external temperature difference in pairs, and carrying out anomaly detection test.
Further, step 7 the detection model to be established detects the temperature anomaly of the test set, and realizes the joint temperature early warning, including:
the detection performance is judged by an ROC (receiver Operating characteristics) curve. The curve has a False Positive Rate (FPR) as an abscissa and a detection rate (TPR) as an ordinate. Wherein, the calculation formulas of TPR and FPR are as follows:
TPR=TP/(TP+FN)
FPR=FP/(FP+TN)
tp (true positive), that is, the actual splice temperature is abnormal and the detection result is abnormal, fn (false negative), that is, the actual splice temperature is abnormal and the detection result is normal, fp (false positive), that is, the actual splice temperature is normal and the detection result is abnormal, tn (true negative), that is, the actual splice temperature is normal and the detection result is normal.
The area under the ROC curve is called AUC (area under the dark), and when different classification models are compared, the AUC of each model can be compared to serve as an index of the quality of the model.
And inputting the temperature test set to be detected into the selected detection classifier for detection.
The invention has the beneficial effects that:
and (3) fully utilizing a historical monitoring data source, and processing and analyzing the joint temperature sequence by using integrated empirical mode decomposition (EEMD) to obtain a plurality of subsequences with smaller orders of magnitude. And the LSTM network is adopted to predict each subsequence, so that the joint temperature trend can be predicted. And by using an isolated forest algorithm, multi-index detection can be considered at the same time. Compared with single index detection, the reliability is higher. The method provided by the invention can be used for pre-judging the potential abnormal temperature of the joint in time and carrying out temperature early warning on the cable joint with overhigh temperature and too fast temperature rise.
Drawings
FIG. 1 is a flow chart of a power cable joint temperature anomaly early warning method combining EEMD-LSTM and isolated forest algorithm;
FIG. 2 is joint initial temperature monitoring data;
FIG. 3 is a joint temperature trend prediction;
FIG. 4 is an isolated forest algorithm detection ROC curve;
fig. 5 is a detection of abnormal temperature at the junction of tag 13a 3.
Detailed Description
The following describes in detail a power cable joint temperature anomaly early warning method combining the EEMD-LSTM and the isolated forest algorithm according to the present invention with reference to the embodiments and the accompanying drawings. As shown in FIG. 1, the power cable joint temperature abnormity early warning method combining EEMD-LSTM and isolated forest algorithm of the invention comprises the following steps:
(1) monitoring temperature information of a corresponding label joint based on an online monitoring system aiming at a cable line in a 10kV distribution network cable trench in a certain area, wherein the temperature information comprises joint surface temperature, core temperature and internal and external relative temperature difference, and an initial temperature sequence is obtained;
(2) decomposing the initial temperature sequence by adopting integrated empirical mode decomposition (EEMD), decomposing a plurality of subsequences step by step, and processing a non-stable sequence;
the procedure for processing the linker temperature sequence using EEMD is shown below:
1) setting zero mean Gaussian white noise with amplitude of k and decomposition repetition frequency of M, wherein the noise obeys normal distribution;
2) adding the Gaussian white noise n (t) into a cable joint temperature initial sequence x (t) to perform an mth EMD decomposition;
searching all local minimum points and local maximum points in the sequence after noise is added, fitting by using a 3-time spline interpolation function to obtain an upper envelope line sequence u (t) and a lower envelope line sequence v (t), and taking the mean value of the upper envelope line sequence u (t) and the lower envelope line sequence v (t) to obtain a mean value sequence m (t) ((u) (t) + v (t))/2;
subtracting the mean sequence from the sequence with the noise added to obtain an intermediate sequence h (t) ═ x (t) — m (t);
thirdly, judging whether h (t) meets an eigenmode function (IMF), if so, defining the h (t) as IMF1, otherwise, regarding the h (t) as a new x (t), and repeating the steps (I) -II. By parity of reasoning, the mth EMD decomposition is completed;
calculating residual component r (t) x (t) -h (t). Repeating the steps I-III on the residual components, carrying out the same 'stripping', and stopping when the decomposition termination condition is met;
3) and performing ensemble averaging on the IMFs obtained by the M times of EMD decomposition to offset the amplitude influence caused by the added white noise.
(3) Establishing an LSTM network unit, optimizing LSTM network unit parameters, predicting each decomposed subsequence by using the LSTM network unit, reconstructing a subsequence prediction result, and outputting a joint temperature value at a future moment;
and (3) predicting the sub-sequence decomposed in the step (2) by using an LSTM network unit. The LSTM network element parameters are optimized. In the LSTM network structure parameters, the single-step prediction node of an output layer is 1, and the window length is 24; the number of the three-step prediction nodes is 3, and the window length is 72. The input window length is the length of the historical sequence used to make the joint temperature prediction. The use of a single LSTM layer can control the computational cost within a reasonable range while meeting network performance requirements, so the number of hidden layers is 1. During each training, after the lengths of an input layer and an output layer are determined, a particle swarm algorithm is used for searching a neuron dimension cell _ num of a hidden layer, a network learning rate factor c and a batch size Sbatch. And predicting each sub-sequence obtained by decomposition by using an LSTM network unit, reconstructing a sub-sequence prediction result, and outputting a joint temperature value at a future moment.
(4) Extracting a joint temperature index data set, performing characteristic transformation on the original data set by using haar wavelet transformation, and extracting temperature change trend information to obtain a temperature detection sample;
temperature index in a monitoring sample, and joint wire core temperature TjcSurface temperature TbcInternal and external temperature difference delta TcThe data of (2) are extracted to obtain respective one-dimensional matrixes.
Tjc=[Tjc1 Tjc2 … Tjci … Tjcn]
Tbc=[Tbc1 Tbc2 … Tbci … Tbcn]
ΔTc=[Tc1 Tc2 … Tci … Tcn]
1) Calculating the average value A of adjacent datajc、Abc、Ac。
2) And calculating the difference value of the original data and the mean value. And calculating an average value, losing part of information and making the matrix size be half of the original size. For storing the detail information of the missing part, the calculated average value is subtracted from the previous data of the adjacent data pair to obtain a difference valueDjc、Dbc、Dc. When the temperature acquisition period is short, this difference can be regarded as the temperature rise rate. And storing the calculated mean value and difference value to obtain a new matrix after transformation, wherein the new matrix is used as a characteristic vector of the joint temperature and is expressed as:
Hjc=[Ajc1 Ajc2 … Ajci … Ajc(n+1)/2 Djc1 Djc2 … Djci … Djc(n+1)/2]
Hbc=[Abc1 Abc2 … Abci … Abc(n+1)/2 Dbc1 Dbc2 … Dbci … Dbc(n+1)/2]
Hc=[Ac1 Ac2 … Aci … Ac(n+1)/2 Dc1 Dc2 … Dci … Dc(n+1)/2]
combining the original data, the finally obtained detection vector group is as follows:
Xjc=[Tjc Hjc]T
Xbc=[Tbc Hbc]T
Xc=[ΔTc Hc]T
(5) setting attention values and early warning values corresponding to all temperature indexes of the joint;
setting an attention value and an early warning value according to operation experience, and respectively giving an upper threshold value and a lower threshold value:
index (I) | Attention value | Early warning value |
Temperature of |
70℃ | 85℃ |
Temperature of |
45 |
50℃ |
|
20 |
30℃ |
(6) Detecting temperature abnormality by adopting an isolated forest algorithm, dividing a detection sample into a training set and a testing set, carrying out abnormality marking on the training set, and establishing a detection classifier of joint surface temperature-core temperature, surface temperature-relative temperature difference and core temperature-relative temperature difference;
the isolated forest algorithm is set as a standard parameter: the sample size is 256; the depth of the tree is 8; the number of the isolated forest trees is 100; the maximum number of features is 1. And after the setting of the index threshold is finished, selecting different classification models. And combining the transformed complete sample and the sample obtained by combining the temperature of the wire core of the joint, the surface temperature of the joint and the relative internal and external temperature difference in pairs, and carrying out anomaly detection test.
(7) And (4) according to the steps (3), (4), (5) and (6), carrying out temperature anomaly detection on the test set by the established detection model, and realizing joint temperature early warning.
Specific examples are given below:
and taking a data set of 30 days of temperature data of different joints in a cable trench of a certain 10kV distribution network from 7 month and 15 days to 8 month and 13 days in 2019 as a research sample. The monitoring interval time is 1h, and the period is 24 h. The 480 groups of monitoring data from 7-month 15-day zero to 8-month 3-day zero are used as initial training samples, the training samples are added in a subsequent rolling mode, and the data from 8-month 4-day zero to 8-month 13-day zero are tested.
FIG. 2 is a sequence of initial joint temperatures. FIG. 3 shows the joint temperature trend predicted by the LSTM network single-step and three-step predictions, which show that the EEMD-LSTM single-step and three-step predictions can predict the joint temperature trend. FIG. 4 is an ROC curve of different classifiers, the AUC value of each classifier is 0.83-0.85, two groups of indexes of core temperature and relative temperature difference are selected to be combined, and the anomaly detection effect is better; selecting two sets of index combinations of surface temperature and relative temperature difference to obtain the second best detection effect; the AUC average value detected by the complete sample is 0.981, and the detection effect is excellent. FIG. 5 shows that the abnormal points of the joint detection of the label 13A3 in the cable trench, which are far away from the normal temperature data, are distributed outside the abnormal value isoline, and the abnormal value of the label 13A3 can be detected in the range of-6.05044 to-7.67423.
The foregoing merely represents preferred embodiments of the invention, which are described in some detail and detail, and therefore should not be construed as limiting the scope of the invention. It should be noted that, for those skilled in the art, various changes, modifications and substitutions can be made without departing from the spirit of the present invention, and these are all within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (6)
1. A power cable joint temperature abnormity early warning method combining EEMD-LSTM and isolated forest algorithm is characterized by comprising the following steps:
step 1, monitoring temperature information of a label joint based on a corresponding online monitoring system aiming at a cable line in a 10kV distribution network cable trench in a certain area, wherein the temperature information comprises joint surface temperature, core temperature and internal and external relative temperature difference, and obtaining an initial temperature sequence;
step 2, decomposing the initial temperature sequence by adopting integrated empirical mode decomposition (EEMD), decomposing a plurality of subsequences step by step, and processing the unstable sequence of the initial temperature of the joint;
step 3, establishing an LSTM network unit, optimizing LSTM network unit parameters, predicting each decomposed subsequence by using the LSTM network unit, reconstructing a subsequence prediction result, and outputting a joint temperature value at a future moment;
step 4, extracting a temperature index data set from the joint monitoring real-time temperature data and the prediction data, performing characteristic transformation on the original data set by using haar wavelet transformation, and extracting temperature change trend information to obtain a temperature detection sample;
step 5, setting attention values and early warning values corresponding to all temperature indexes of the joint;
step 6, detecting temperature anomaly by adopting an isolated forest algorithm, dividing a detection sample into a training set and a testing set, carrying out anomaly marking on the training set, and establishing detection sub-classifiers of joint surface temperature-core temperature, surface temperature-relative temperature difference and core temperature-relative temperature difference;
and 7, carrying out temperature anomaly detection on the test set by the established detection model according to the steps 3, 4, 5 and 6, and realizing joint temperature early warning.
2. The power cable joint temperature anomaly early warning method combining EEMD-LSTM and isolated forest algorithm according to claim 1, wherein the integrated empirical mode decomposition of step 2 is specifically expressed as:
the procedure for processing the linker temperature sequence using EEMD is shown below:
(1) setting zero mean Gaussian white noise with amplitude of k and decomposition repetition frequency of M, wherein the noise obeys normal distribution;
(2) adding the Gaussian white noise n (t) into a cable joint temperature initial sequence x (t) to perform an mth EMD decomposition;
1) searching all local minimum points and local maximum points in the sequence after noise is added, obtaining an upper envelope line sequence u (t) and a lower envelope line sequence v (t) by using 3 times of spline interpolation function fitting, and taking the mean value of the upper envelope line sequence u (t) and the lower envelope line sequence v (t) to obtain a mean value sequence m (t) (u (t)) + v (t))/2;
2) subtracting the mean sequence from the sequence to which noise is added to obtain an intermediate sequence h (t) ═ x (t) — m (t);
3) judging whether h (t) meets the intrinsic mode function IMF, if yes, defining the intrinsic mode function IMF as IMF1, otherwise, regarding the intrinsic mode function as a new x (t), and repeating the steps 1) to 2); by parity of reasoning, the mth EMD decomposition is completed;
4) calculating a residual component r (t) ═ x (t) — h (t); repeating the steps 1) -3) on the residual component, and stopping when the decomposition termination condition is met;
(3) and performing ensemble averaging on the IMFs obtained by the M times of EMD decomposition to offset the amplitude influence caused by the added white noise.
3. The method for early warning of temperature abnormality of power cable joint in combination with EEMD-LSTM and isolated forest algorithm as claimed in claim 1, wherein the establishment of LSTM network unit in step 3 is expressed as:
predicting the sub-sequence decomposed in the step 2 by using an LSTM network unit; optimizing LSTM network element parameters; in the LSTM network structure parameters, the single-step prediction node of an output layer is 1, and the window length is 24; the number of the three-step prediction nodes is 3, and the window length is 72; inputting a window length as a historical sequence length for joint temperature prediction; the calculation cost is controlled within a reasonable range by using a single LSTM layer, the network performance requirement is met, and the number of hidden layers is 1; during each training, after the lengths of an input layer and an output layer are determined, a particle swarm algorithm is used for searching a neuron dimension cell _ num of a hidden layer, a network learning rate factor c and a batch size Sbatch(ii) a And predicting each sub-sequence obtained by decomposition by using an LSTM network unit, reconstructing a sub-sequence prediction result, and outputting a joint temperature value at a future moment.
4. The power cable joint temperature anomaly early warning method combining EEMD-LSTM and isolated forest algorithm as claimed in claim 1, wherein said step 4 of using haar wavelet transform to perform feature transformation on original data set, extracting temperature variation trend information to obtain temperature detection sample, the specific process is:
temperature index in a monitoring sample, and joint wire core temperature TjcSurface temperature TbcInternal and external temperature difference delta TcExtracting the data to obtain respective one-dimensional matrixes;
Tjc=[Tjc1 Tjc2…Tjci…Tjcn]
Tbc=[Tbc1 Tbc2…Tbci…Tbcn]
ΔTc=[Tc1 Tc2…Tci…Tcn]
(1) calculating the average value A of adjacent datajc、Abc、Ac;
(2) Calculating the difference value between the original data and the mean value; calculating an average value, losing part of information and making the matrix size be half of the original size; for storing the missing part of the detail information, the calculated average value is subtracted from the previous data of the adjacent data pair to obtain the difference Djc、Dbc、Dc(ii) a When the temperature acquisition period is short, the difference value can be regarded as the temperature rise rate; and storing the calculated mean value and difference value to obtain a new matrix after transformation, wherein the new matrix is used as a characteristic vector of the joint temperature and is expressed as:
Hjc=[Ajc1 Ajc2…Ajci…Ajc(n+1)/2 Djc1 Djc2…Djci…Djc(n+1)/2]
Hbc=[Abc1 Abc2…Abci…Abc(n+1)/2 Dbc1 Dbc2…Dbci…Dbc(n+1)/2]
Hc=[Ac1 Ac2…Aci…Ac(n+1)/2 Dc1 Dc2…Dci…Dc(n+1)/2]
combining the original data, the finally obtained detection vector group is as follows:
Xjc=[Tjc Hjc]T
Xbc=[Tbc Hbc]T
Xc=[ΔTc Hc]T
5. the power cable joint temperature anomaly early warning method combining EEMD-LSTM and the isolated forest algorithm according to claim 1, wherein the isolated forest algorithm is adopted to detect temperature anomaly in step 6, and the specific expression is as follows:
the isolated forest algorithm is set as a standard parameter: sample size NumsubIs 256; the tree depth, HeightLimit, is 8; number Num of isolated forest treestreeIs 100; the maximum number of features is 1; after setting the index threshold value, selecting different classification models; and combining the transformed complete sample and the sample obtained by combining the temperature of the wire core of the joint, the surface temperature of the joint and the relative internal and external temperature difference in pairs, and carrying out anomaly detection test.
6. The method for early warning of temperature abnormality of power cable joint in combination with EEMD-LSTM and isolated forest algorithm as claimed in claim 1, wherein said step 7 of detecting temperature abnormality of test set by established detection model to realize early warning of joint temperature includes:
judging the detection performance by an ROC curve, wherein the curve takes a false alarm rate FPR as an abscissa and a detection rate TPR as an ordinate; wherein, the calculation formulas of TPR and FPR are as follows:
TPR=TP/(TP+FN)
FPR=FP/(FP+TN)
TP represents true positive, namely the actual joint temperature is abnormal and the detection result is abnormal, FN represents false negative, namely the actual joint temperature is abnormal and the detection result is normal, FP represents false positive, namely the actual joint temperature is normal and the detection result is abnormal, TN represents true negative, namely the actual joint temperature is normal and the detection result is normal;
the area under the ROC curve is called AUC, and when different classification models are compared, the AUC of each model is compared to serve as an index of the quality of the model;
and inputting the temperature test set to be detected into the selected detection classifier for detection.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111473398.7A CN114169237B (en) | 2021-11-30 | 2021-11-30 | Power cable joint temperature abnormality early warning method combining EEMD-LSTM and isolated forest algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111473398.7A CN114169237B (en) | 2021-11-30 | 2021-11-30 | Power cable joint temperature abnormality early warning method combining EEMD-LSTM and isolated forest algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114169237A true CN114169237A (en) | 2022-03-11 |
CN114169237B CN114169237B (en) | 2024-05-03 |
Family
ID=80483206
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111473398.7A Active CN114169237B (en) | 2021-11-30 | 2021-11-30 | Power cable joint temperature abnormality early warning method combining EEMD-LSTM and isolated forest algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114169237B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115438102A (en) * | 2022-11-09 | 2022-12-06 | 成都新希望金融信息有限公司 | Space-time data anomaly identification method and device and electronic equipment |
CN115760484A (en) * | 2022-12-07 | 2023-03-07 | 湖北华中电力科技开发有限责任公司 | Method, device and system for improving potential danger identification capability of power distribution station area and storage medium |
CN116627953A (en) * | 2023-05-24 | 2023-08-22 | 首都师范大学 | Method for repairing loss of groundwater level monitoring data |
CN117346917A (en) * | 2023-10-13 | 2024-01-05 | 大唐同心新能源有限公司 | Cable joint temperature prediction system and prediction method |
CN117713505A (en) * | 2024-02-06 | 2024-03-15 | 北京嘉捷恒信能源技术股份有限公司 | High-side current detection method and system for switching power supply |
CN117874434A (en) * | 2024-03-11 | 2024-04-12 | 湖南工程学院 | Data cleaning method in intelligent environmental temperature and humidity detection process |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019233189A1 (en) * | 2018-06-04 | 2019-12-12 | 江南大学 | Method for detecting sensor network abnormal data |
CN111241466A (en) * | 2020-01-15 | 2020-06-05 | 上海海事大学 | Ship flow prediction method based on deep learning |
CN111738520A (en) * | 2020-06-24 | 2020-10-02 | 中国电子科技集团公司第二十八研究所 | System load prediction method fusing isolated forest and long-short term memory network |
CN112287602A (en) * | 2020-10-28 | 2021-01-29 | 北京国信会视科技有限公司 | Motor car axle temperature fault early warning method based on machine learning and isolated forest |
-
2021
- 2021-11-30 CN CN202111473398.7A patent/CN114169237B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019233189A1 (en) * | 2018-06-04 | 2019-12-12 | 江南大学 | Method for detecting sensor network abnormal data |
CN111241466A (en) * | 2020-01-15 | 2020-06-05 | 上海海事大学 | Ship flow prediction method based on deep learning |
CN111738520A (en) * | 2020-06-24 | 2020-10-02 | 中国电子科技集团公司第二十八研究所 | System load prediction method fusing isolated forest and long-short term memory network |
CN112287602A (en) * | 2020-10-28 | 2021-01-29 | 北京国信会视科技有限公司 | Motor car axle temperature fault early warning method based on machine learning and isolated forest |
Non-Patent Citations (1)
Title |
---|
杨涛;白晗;刘洋;王云龙;张楚曼;: "基于EEMD和改进Elman神经网络的电力电缆故障监测及预警算法研究", 实验室研究与探索, no. 10, 15 October 2018 (2018-10-15) * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115438102A (en) * | 2022-11-09 | 2022-12-06 | 成都新希望金融信息有限公司 | Space-time data anomaly identification method and device and electronic equipment |
CN115760484A (en) * | 2022-12-07 | 2023-03-07 | 湖北华中电力科技开发有限责任公司 | Method, device and system for improving potential danger identification capability of power distribution station area and storage medium |
CN116627953A (en) * | 2023-05-24 | 2023-08-22 | 首都师范大学 | Method for repairing loss of groundwater level monitoring data |
CN116627953B (en) * | 2023-05-24 | 2023-10-27 | 首都师范大学 | Method for repairing loss of groundwater level monitoring data |
CN117346917A (en) * | 2023-10-13 | 2024-01-05 | 大唐同心新能源有限公司 | Cable joint temperature prediction system and prediction method |
CN117713505A (en) * | 2024-02-06 | 2024-03-15 | 北京嘉捷恒信能源技术股份有限公司 | High-side current detection method and system for switching power supply |
CN117713505B (en) * | 2024-02-06 | 2024-04-05 | 北京嘉捷恒信能源技术股份有限公司 | High-side current detection method and system for switching power supply |
CN117874434A (en) * | 2024-03-11 | 2024-04-12 | 湖南工程学院 | Data cleaning method in intelligent environmental temperature and humidity detection process |
CN117874434B (en) * | 2024-03-11 | 2024-05-24 | 湖南工程学院 | Data cleaning method in intelligent environmental temperature and humidity detection process |
Also Published As
Publication number | Publication date |
---|---|
CN114169237B (en) | 2024-05-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN114169237B (en) | Power cable joint temperature abnormality early warning method combining EEMD-LSTM and isolated forest algorithm | |
CN113156917B (en) | Power grid equipment fault diagnosis method and system based on artificial intelligence | |
CN103886374B (en) | A kind of cable connector conductor temperature Forecasting Methodology based on RBF neural | |
CN112561736A (en) | Fault diagnosis system and method for relay protection device of intelligent substation | |
CN111650451A (en) | Converter transformer fault reason identification method and system | |
CN114254818A (en) | Low-voltage distribution network cascading failure early warning method based on risk assessment model | |
CN113780401A (en) | Composite insulator fault detection method and system based on principal component analysis method | |
CN107703407A (en) | power cable diagnostic method and device | |
CN117609841A (en) | Cable fault pre-judging analysis method based on neural network algorithm | |
CN116677570A (en) | Fault early warning method and system based on cabin temperature monitoring of offshore wind turbine | |
CN111157045B (en) | Artificial intelligence electric line abnormal risk degree monitoring method and system | |
CN115600754A (en) | Method, system, equipment and storage medium for predicting potential safety hazard of user power utilization | |
CN112632148A (en) | Intelligent early warning method for power distribution network components under extreme weather condition | |
CN115034606A (en) | Method for evaluating running state of low-voltage busbar insulating sheath of transformer | |
CN117394311B (en) | Power distribution network toughness assessment and emergency control method based on multi-source information fusion | |
CN113935423A (en) | Power battery fault early warning method and system for coupling fuzzy control rule with Elman neural network | |
CN117491813A (en) | Insulation abnormality detection method for power battery system of new energy automobile | |
CN114492636B (en) | Transformer winding state signal acquisition system | |
Luo et al. | Research on Intelligent Mining Algorithm for Distribution Network Transformer Fault Early Warning | |
CN115224684A (en) | Intelligent power distribution network risk state identification method and system based on immune hazard theory | |
Zhang et al. | Electric vehicle charging fault diagnosis based on bp neural network | |
CN114899946B (en) | Electric power system distributed fault diagnosis method based on average consistency | |
CN111199254A (en) | High-voltage electrical equipment real-time detection method based on RBF neural network and Bayesian network | |
CN114897035B (en) | Multi-source data feature fusion method for 10kV cable state evaluation | |
CN117783792B (en) | Valve side sleeve insulation state detection method and system based on multiparameter real-time monitoring |
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 |