CN113723478A - Track circuit fault diagnosis method based on priori knowledge - Google Patents
Track circuit fault diagnosis method based on priori knowledge Download PDFInfo
- Publication number
- CN113723478A CN113723478A CN202110943713.1A CN202110943713A CN113723478A CN 113723478 A CN113723478 A CN 113723478A CN 202110943713 A CN202110943713 A CN 202110943713A CN 113723478 A CN113723478 A CN 113723478A
- Authority
- CN
- China
- Prior art keywords
- analog quantity
- quantity information
- track circuit
- fault diagnosis
- prior knowledge
- 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 34
- 238000003745 diagnosis Methods 0.000 title claims abstract description 22
- 230000008569 process Effects 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims description 17
- 238000010606 normalization Methods 0.000 claims description 15
- 238000012544 monitoring process Methods 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 12
- 230000003993 interaction Effects 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 5
- 230000002452 interceptive effect Effects 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000002194 synthesizing effect Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 claims description 2
- 238000004088 simulation Methods 0.000 claims 2
- 238000004458 analytical method Methods 0.000 abstract description 5
- 238000013024 troubleshooting Methods 0.000 abstract description 4
- 239000000047 product Substances 0.000 description 7
- 230000000875 corresponding effect Effects 0.000 description 5
- 238000012423 maintenance Methods 0.000 description 5
- 238000012549 training Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000007689 inspection Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000006227 byproduct Substances 0.000 description 1
- -1 carrier Substances 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000000306 component Substances 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/045—Combinations of 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Train Traffic Observation, Control, And Security (AREA)
Abstract
The invention discloses a fault diagnosis method of a track circuit based on prior knowledge, which is characterized in that on the basis of analog quantity analysis, characteristics with prior knowledge are generated, a model constructed by the characteristics can be used for more accurate fault diagnosis, a fault area can be judged by identifying codes, and the operation and learning cost is reduced; moreover, the method can be used for different interval track circuits, each interval section can construct an exclusive model according to the analog quantity of each interval section, and different types of track circuits can also adopt the model only by modifying the dimension of input data. Compared with the conventional manual troubleshooting, the fault locating process is carried out in hours on average, and is reduced to the second level, so that the fault troubleshooting time is greatly shortened.
Description
Technical Field
The invention relates to the technical field of track circuit fault diagnosis, in particular to a track circuit fault diagnosis method based on priori knowledge.
Background
At present, in the field of rail circuit fault diagnosis, a large number of excellent methods and models exist, and a design system can be used for reference. Such as a decision tree optimization C4.5 method, a rough set similarity calculation method, a Bayesian network, a fuzzy entropy theory and an asymmetric proximity principle and the like. The above is the application of the traditional machine learning in the field of the fault function analysis of the track circuit, and in recent years, the model is also deeply learned to judge the fault mode, such as a full-connection network, a deep belief network and the like.
The traditional track circuit fault intelligent diagnosis process is characterized in that an analog quantity alarm threshold value is defined in advance, and a patrol worker carries out manual investigation after a fault occurs. However, this process can lead to a large amount of inspection work delay and even serious missed inspection problems due to the misalignment of the inspection personnel or lack of personnel. Although some track circuit fault diagnosis methods based on deep learning exist at present, the low-order interaction information among analog quantities is not fully considered in the existing methods, and useful information is omitted from manual maintenance records of fault maintenance personnel, so that certain limitations exist.
Disclosure of Invention
The invention aims to provide a track circuit fault diagnosis method based on prior knowledge, which can find a fault occurring region in time by analyzing and diagnosing a track circuit in real time, can improve railway transportation efficiency, and has important significance and practical application value for ensuring the normal operation of the track circuit and reducing the labor intensity of operation and maintenance personnel.
The purpose of the invention is realized by the following technical scheme:
a track circuit fault diagnosis method based on prior knowledge comprises the following steps:
for the track circuit of any section, screening a plurality of analog quantity information meeting the set requirements from all analog quantity information measured by the monitoring point, and carrying out normalization processing;
taking the analog quantity information after normalization processing as the input of a previously constructed DeepMemory model; the Deep Memory model comprises a Deep part and a Memory part;
processing by a Deep part internal hidden layer to obtain hidden features corresponding to each analog quantity information; carrying out interactive processing on the analog quantity information after the normalization processing through a Memory part to generate data with prior knowledge; and (4) synthesizing the hidden features corresponding to the analog quantity information and the region codes in the data output track circuit with the prior knowledge to determine the fault region.
According to the technical scheme provided by the invention, on the basis of purely analog analysis, the characteristic with prior knowledge is generated, so that the constructed model can be used for more accurately diagnosing faults, and the fault area can be judged by identifying codes, so that the operation and learning cost is reduced; moreover, the method can be used for different interval track circuits, each interval section can construct an exclusive model according to the analog quantity of each interval section, and different types of track circuits can also adopt the model only by modifying the dimension of input data. Compared with the conventional manual troubleshooting, the fault locating process is carried out in hours on average, and is reduced to the second level, so that the fault troubleshooting time is greatly shortened.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a flowchart of a track circuit fault diagnosis method based on a priori knowledge according to an embodiment of the present invention;
fig. 2 is a schematic diagram of acquiring required data at a monitoring point provided in a track circuit according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a deep memory model according to an embodiment of the present invention;
FIG. 4 is a schematic illustration of the Deep portion provided by an embodiment of the present invention;
fig. 5 is a schematic diagram of a Memory portion according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
The terms that may be used herein are first described as follows:
the term "and/or" means that either or both can be achieved, for example, X and/or Y means that both cases include "X" or "Y" as well as three cases including "X and Y".
The terms "comprising," "including," "containing," "having," or other similar terms of meaning should be construed as non-exclusive inclusions. For example: including a feature (e.g., material, component, ingredient, carrier, formulation, material, dimension, part, component, mechanism, device, process, procedure, method, reaction condition, processing condition, parameter, algorithm, signal, data, product, or article of manufacture), is to be construed as including not only the particular feature explicitly listed but also other features not explicitly listed as such which are known in the art.
The track circuit fault diagnosis method based on prior knowledge provided by the invention is described in detail below. Details which are not described in detail in the embodiments of the invention belong to the prior art which is known to the person skilled in the art. Those not specifically mentioned in the examples of the present invention were carried out according to the conventional conditions in the art or conditions suggested by the manufacturer.
As shown in fig. 1, a method for diagnosing a fault of a track circuit based on a priori knowledge includes the following steps:
step 1, for the track circuit of any section, screening a plurality of analog quantity information meeting the set requirements from all analog quantity information measured by monitoring points, and carrying out normalization processing.
In the embodiment of the invention, the analog quantity monitored by the track circuit is collected and analyzed. For the track circuit of any interval, corresponding monitoring points can be arranged according to the actual environment, N analog quantity information which can be measured through all the monitoring points is obtained, and M analog quantities are screened out according to the change degree of the analog quantity information, wherein N and M are positive integers, and N is greater than M.
As shown in fig. 2, a schematic diagram of the track circuit layout monitoring points in a certain section is given, because there are many track circuit devices, the centralized monitoring system has five monitoring points in the track circuit, and three monitoring points are set in the track in front of the train operation, and through the analysis of the eight monitoring points, 27 analog quantities including voltage, current, frequency shift, carrier frequency and other parameters can be measured in the track circuit. Finally, 10 voltage and current analog quantities with large changes are selected, and the names and units of the analog quantities are shown in table 1.
TABLE 1 analog quantity information screened
Considering that the unit is different between different types of analog quantity information, the value is different in size, the value is different from zero to thousands of values, and therefore normalization processing is needed before the analog quantity information is input into the model. In the embodiment of the invention, a Min-Max normalization mode is selected to normalize the track circuit analog quantity to be between 0 and 1, and the formula is as follows:
wherein x isi、
Each representing a value of the analog quantity before and after normalization, xmin、xmaxEach representing a minimum value and a maximum value of the analog quantity of the corresponding category.
Step 2, taking the analog quantity information after normalization processing as the input of a previously constructed DeepMemory model; the Deep Memory model comprises a Deep part and a Memory part.
In the embodiment of the present invention, a deep memory model is constructed, as shown in fig. 3, which mainly includes: deep part and Memory part.
In the embodiment of the invention, the Deep part belongs to a neural network and can be understood as mining data information in a high-order space, and the Memory part generates regular characteristics of prior knowledge by adopting a characteristic pairwise product mode, has Memory capacity and can be understood as mining data information in a low-order space.
And 3, calculating and processing the normalized analog quantity information through the Deep part and the Memory part to obtain a final fault area code.
As shown in fig. 3, the normalized analog quantity information needs to be processed by two parts, respectively, on one hand, the analog quantity information is processed by forward propagation of a hidden layer inside the Deep module to obtain the hidden characteristics of each analog quantity information, and on the other hand, the analog quantity information is processed by interaction through a Memory module to generate data with prior knowledge; the hidden characteristics of each analog quantity information and the data with the prior knowledge are uniformly input to an output layer for fault classification, and fault area codes in the track circuit can be obtained.
Fig. 4 shows the principle of the Deep part, in which a plurality of hidden layers are arranged, the normalized analog quantity information obtained in the step 1 is input, and the hidden features of the analog quantity information are obtained through the hidden layers in the Deep part.
The input layer and the three hidden layers are four layers in total, and a neural network forward propagation algorithm is adopted between the data of the previous layer and the data of the next layer. And obtaining an intermediate vector by matrix operation of output data of the network layer of the previous layer, obtaining input of the network layer of the next layer by calculation of the intermediate vector through a loss function, wherein the loss function adopted by the first four layers is a Relu function.
If only the Deep part is adopted, it can be considered that 10 analog quantities need to be comprehensively considered for each fault diagnosis, and in the actual fault diagnosis process of the artificial track circuit, there is a case that only a few analog quantities need to be considered to approximately obtain a fault area, for example, if the supply voltage and the supply current drop by the same amplitude at the same time, the fault is judged to occur in the area 1 with greater confidence, so that a diagnosis rule with prior knowledge can be summarized, and thus the following guess is given: whether similar features with prior knowledge can be generated through interaction of pairwise products of the features, and therefore verification is carried out by adding a Memory part on the basis of the Deep part.
Fig. 5 shows the principle of the Memory part, in which the normalized analog quantity information obtained in the step 1 is also input, and feature interaction is performed on the input data in the Memory part, for example, pairwise product interaction is performed on the normalized analog quantity information, and data with prior knowledge is output. And finally, the Deep Memory model combines the hidden features obtained by the Deep part and the data with the prior knowledge obtained by the Memory part to output a certain fault area code in the track circuit area.
The principle of synthesizing the hidden features corresponding to each analog quantity information and the fault area codes in the data output track circuit with the prior knowledge can be understood as follows: and splicing the right side of the last layer (namely, the third hidden layer) of the middle layer of the Deep part, and splicing the output of the Memory part. The two parts are then entered into the output layer together, and the number of nodes of the output layer is the number of regions plus 1 (normal category), for example, when the track circuit of a section is divided into 9 regions as mentioned later, the number of nodes of the output layer is 10, and the loss function is selected as a multi-class cross entropy function. Illustratively, the resulting fault code is as follows, 1000000000 for data normal, 0100000000 for region 1 fault, and so on 0000000001 for region 9 fault.
In the training stage, a Relu function is adopted by an input layer and a hidden layer of the Deep part, a multi-classification cross entropy function is adopted from the hidden layer to an output layer, a multi-classification cross entropy function is also adopted after the Memory part is spliced, historical data can be used as a training set in the training stage, training is carried out in a conventional mode, the specific process can refer to conventional technologies, and details are omitted.
Taking the screening of 10 analog quantities mentioned in the foregoing step 1 as an example, the Memory part inputs 10 normalized analog quantity information, and product-by-product interactions result in 45 data in total, as shown in table 2.
| Interactive features | 1@2 | 1@3 | ... | 8@10 | 9@10 |
| Characteristic value | 0.8875 | 0.0658 | ... | 0.0293 | 0.9118 |
Table 2 product interaction procedure
In table 2, @ represents product interaction, and 1@2 represents product interaction between the normalized analog quantity 1 and the normalized analog quantity 2, so that a new characteristic value is generated, and the value range of the interactive characteristic is between 0 and 1 because the normalized data is between 0 and 1; the specific values shown in table 2 are examples and are not intended to be limiting.
In the embodiment of the invention, according to the standard of 'technical conditions for temporary operation of a ZPW-2000 interval track circuit outdoor monitoring and diagnosis system', a track circuit of one section is divided into 9 areas, different codes correspond to different areas, and a fault area can be positioned through the codes output by a deep memory model:
1. an indoor transmitting end direction switching circuit; 2. a sending end simulates a network; 3. an outdoor transmitting end channel; 4) a transmitting end tuning area; 5) a main rail line; 6) a receiving end tuning area; 7) an outdoor receiving end channel; 8) the receiving end simulates a network; 9) indoor receiving end direction switching circuit.
In the embodiment of the invention, the final output result of the model is a specific fault area code, and the output can be constructed in a visual interface mode, such as numbers 1-9 and the like representing fault areas 1-9. The fault area can be judged by identifying the code, and the operation and learning cost is reduced.
Compared with the prior art, the scheme of the embodiment of the invention mainly has the following beneficial effects:
1) the stable model obtained through training can replace workers to carry out centralized and correlated analysis on the track circuit data, accurately and timely identify the fault occurring region, and can guide maintenance personnel to overhaul.
2) By analyzing and diagnosing the condition of the track circuit in real time, the fault symptom of the equipment is found in advance, and scientific maintenance is guided.
3) The faults of the track circuit are classified into 9 large categories by coarse granularity, so that the output dimension of the model can be accurate, and the generalization capability of the model is improved; and moreover, the fault area can be judged by identifying the code, and the operation and learning cost is reduced.
4) Aiming at different interval track circuits, each interval can construct an exclusive model according to the analog quantity of each interval, different types of track circuits can adopt the model, only the dimension of input data needs to be modified, and the application range of the scheme is widened.
Through the above description of the embodiments, it is clear to those skilled in the art that the above embodiments can be implemented by software, and can also be implemented by software plus a necessary general hardware platform. With this understanding, the technical solutions of the embodiments can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods according to the embodiments of the present invention.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (6)
1. A track circuit fault diagnosis method based on prior knowledge is characterized by comprising the following steps:
for the track circuit of any section, screening a plurality of analog quantity information meeting the set requirements from all analog quantity information measured by the monitoring point, and carrying out normalization processing;
taking the analog quantity information after normalization processing as the input of a previously constructed DeepMemory model; the Deep Memory model comprises a Deep part and a Memory part;
processing by a Deep part internal hidden layer to obtain hidden features corresponding to each analog quantity information; carrying out interactive processing on the analog quantity information after the normalization processing through a Memory part to generate data with prior knowledge; and (4) synthesizing the hidden features corresponding to the analog quantity information and the region codes in the data output track circuit with the prior knowledge to determine the fault region.
2. The track circuit fault diagnosis method based on the prior knowledge as claimed in claim 1, wherein the step of screening out a plurality of analog quantity information meeting the setting requirement from all analog quantity information measured from the monitoring points comprises:
n analog quantity information which can be measured through all monitoring points is screened out M analog quantities according to the change degree of the analog quantity information, wherein N and M are positive integers, and N > M.
3. The method of claim 1, wherein the normalization process is performed by a Min-Max normalization.
4. The track circuit fault diagnosis method based on the prior knowledge of claim 1, wherein the Deep part comprises: the input layer and the three hidden layers are four layers in total, a neural network forward propagation algorithm is adopted between the previous layer and the next layer of data, the output data of the network layer of the previous layer is subjected to matrix operation to obtain an intermediate vector, the intermediate vector is subjected to loss function calculation to obtain the input of the next layer of network layer, and the loss functions adopted by the previous four layers are Relu functions.
5. The track circuit fault diagnosis method based on the priori knowledge of claim 1, wherein the Memory part interactively processes the analog quantity information after the normalization processing, and comprises:
and the Memory part carries out product interaction on the analog quantity information after the normalization processing.
6. The track circuit fault diagnosis method based on the priori knowledge of claim 1, wherein the track circuit is divided into the following 9 regions, corresponding region codes are arranged in different regions, and corresponding fault regions are determined according to the output region codes:
the system comprises an indoor sending end direction switching circuit area, a sending end simulation network, an outdoor sending end channel, a sending end tuning area, a main rail line, a receiving end tuning area, an outdoor receiving end channel, a receiving end simulation network and an indoor receiving end direction switching circuit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110943713.1A CN113723478A (en) | 2021-08-17 | 2021-08-17 | Track circuit fault diagnosis method based on priori knowledge |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110943713.1A CN113723478A (en) | 2021-08-17 | 2021-08-17 | Track circuit fault diagnosis method based on priori knowledge |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113723478A true CN113723478A (en) | 2021-11-30 |
Family
ID=78676092
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110943713.1A Pending CN113723478A (en) | 2021-08-17 | 2021-08-17 | Track circuit fault diagnosis method based on priori knowledge |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113723478A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111626416A (en) * | 2020-04-24 | 2020-09-04 | 黑龙江瑞兴科技股份有限公司 | Automatic rail circuit fault diagnosis method based on deep convolutional neural network |
| CN111813084A (en) * | 2020-07-10 | 2020-10-23 | 重庆大学 | Mechanical equipment fault diagnosis method based on deep learning |
| CN112561035A (en) * | 2020-12-08 | 2021-03-26 | 上海海事大学 | Fault diagnosis method based on CNN and LSTM depth feature fusion |
| CN113109666A (en) * | 2021-04-09 | 2021-07-13 | 河南省博海大数据科技有限公司 | Track circuit fault diagnosis method based on deep convolutional neural network |
| CN113139335A (en) * | 2021-04-09 | 2021-07-20 | 郑州宥新算法智能科技有限公司 | BP neural network-based track circuit fault intelligent diagnosis method |
-
2021
- 2021-08-17 CN CN202110943713.1A patent/CN113723478A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111626416A (en) * | 2020-04-24 | 2020-09-04 | 黑龙江瑞兴科技股份有限公司 | Automatic rail circuit fault diagnosis method based on deep convolutional neural network |
| CN111813084A (en) * | 2020-07-10 | 2020-10-23 | 重庆大学 | Mechanical equipment fault diagnosis method based on deep learning |
| CN112561035A (en) * | 2020-12-08 | 2021-03-26 | 上海海事大学 | Fault diagnosis method based on CNN and LSTM depth feature fusion |
| CN113109666A (en) * | 2021-04-09 | 2021-07-13 | 河南省博海大数据科技有限公司 | Track circuit fault diagnosis method based on deep convolutional neural network |
| CN113139335A (en) * | 2021-04-09 | 2021-07-20 | 郑州宥新算法智能科技有限公司 | BP neural network-based track circuit fault intelligent diagnosis method |
Non-Patent Citations (1)
| Title |
|---|
| 高敬鹏: "深度学习 卷积神经网络技术与实践", 北京：机械工业出版社, pages: 39 - 42 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | A fuzzy TOPSIS and rough set based approach for mechanism analysis of product infant failure | |
| US20210003640A1 (en) | Fault locating method and system based on multi-layer evaluation model | |
| CN106780121B (en) | Power consumption abnormity identification method based on power consumption load mode analysis | |
| CN110703057B (en) | Power equipment partial discharge diagnosis method based on data enhancement and neural network | |
| CN109597836A (en) | A kind of communication equipment alarm association rule digging method based on weighting matrix | |
| CN106708738B (en) | Software test defect prediction method and system | |
| CN105488740A (en) | Method and system for evaluating risk grade of operation mode of urban power distribution network | |
| CN111340063A (en) | Coal mill data anomaly detection method | |
| US20230083078A1 (en) | Method for intelligent fault detection and location of power distribution network | |
| CN116388402B (en) | Abnormality alarm analysis method applied to power transformation equipment | |
| CN105975797A (en) | Product early-fault root cause recognition method based on fuzzy data processing | |
| CN116737510B (en) | Data analysis-based intelligent keyboard monitoring method and system | |
| CN112488371A (en) | Pipeline intelligent early warning method and system based on big data | |
| CN105471647A (en) | Power communication network fault positioning method | |
| Eltyshev et al. | Intelligent decision support in the electrical equipment diagnostics | |
| CN115130578A (en) | Incremental rough clustering-based online evaluation method for state of power distribution equipment | |
| Dong | Combining unsupervised and supervised learning for asset class failure prediction in power systems | |
| Lyu et al. | A data-driven approach for identifying possible manufacturing processes and production parameters that cause product defects: A thin-film filter company case study | |
| CN113740666B (en) | Method for positioning root fault of storm alarm in power system of data center | |
| CN114548493A (en) | Method and system for predicting current overload of electric energy meter | |
| CN110532122B (en) | Fault analysis method and system, electronic equipment and storage medium | |
| CN111553581A (en) | Equipment maintainability evaluation model based on entropy value | |
| CN113723478A (en) | Track circuit fault diagnosis method based on priori knowledge | |
| Sachan et al. | Multiple correspondence analysis to study failures in a diverse population of a cable | |
| CN112732773B (en) | Method and system for checking uniqueness of relay protection defect data |
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 |