CN113240022A - Wind power gear box fault detection method of multi-scale single-classification convolutional network - Google Patents
Wind power gear box fault detection method of multi-scale single-classification convolutional network Download PDFInfo
- Publication number
- CN113240022A CN113240022A CN202110547118.6A CN202110547118A CN113240022A CN 113240022 A CN113240022 A CN 113240022A CN 202110547118 A CN202110547118 A CN 202110547118A CN 113240022 A CN113240022 A CN 113240022A
- Authority
- CN
- China
- Prior art keywords
- data
- dimensional
- network
- wind power
- convolutional network
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 claims abstract description 17
- 239000013598 vector Substances 0.000 claims description 35
- 230000002159 abnormal effect Effects 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 17
- 238000012549 training Methods 0.000 claims description 15
- 238000012360 testing method Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 230000005856 abnormality Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 2
- 238000010801 machine learning Methods 0.000 abstract description 3
- 230000036541 health Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 238000010248 power generation Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000002123 temporal 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
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
Abstract
The invention provides a wind power gear box fault detection method based on a multi-scale single-classification convolutional network. Compared with the traditional machine learning method, the method can improve the accuracy and reliability of the fault detection of the gearbox.
Description
Technical Field
The invention belongs to the technical field of wind power generation state monitoring, and particularly relates to a wind power gear box fault detection method based on a multi-scale single-classification convolutional network.
Background
In recent years, wind energy has received much attention from all over the world as a clean renewable energy source which is inexhaustible and rapidly developed. The wind turbine generator is widely installed on land and offshore wind farms as an important power generation device for wind power generation. The gear box is one of the core components in the power transmission system of the wind turbine generator, is used as a pivot between a main shaft of the connecting unit and a generator, and plays roles of increasing the rotating speed and transmitting energy. However, due to the influence of the change of the operating environment, such as the irregular and uncertain wind speed, wind direction and the like, various types of faults are easy to occur in the operation process of the wind power gear box, and even the unit is stopped in the serious case. The economic benefit of the wind power plant and the healthy development of the wind power industry are seriously influenced by the faults and the unplanned shutdown, so that the method has important practical significance for timely and accurately detecting the faults of the wind power gearbox.
Most high-power wind turbines are provided with vibration monitoring systems, and the collected signals are mainly vibration signals. These signals contain a large amount of information on the state of health of the gearbox, which can be used to characterize the state of health of the interior of the wind turbine mechanically. Through effective analysis of the vibration signals of the wind power gearbox, the healthy operation conditions and the degradation degree of internal parts of the gearbox, such as gears and bearings, can be known.
At present, a wind power gear box fault detection method based on vibration signals mainly depends on manual feature design, and requirements on signal processing knowledge and expert diagnosis experience are high. On the other hand, the wind power gearbox operates in a healthy state most of the time, so that fault data is difficult to acquire, and the problems of incomplete fault mode, high marking cost and the like are faced. And less research is being done on fault detection methods that require only the first type of data.
Disclosure of Invention
The invention aims to solve the technical problem of providing a method which can effectively detect the fault of a wind power gear box only based on data modeling under the health state of the gear box, thereby overcoming the problem that the reliability of the detection performance can be ensured only when the balance condition of data and fault data samples under the health state is required to be met by most of the existing wind power gear box fault detection methods.
In order to solve the technical problem, the invention provides a wind power gear box fault detection method based on a multi-scale single-classification convolutional network, which comprises the following steps of:
step S1: acquiring one-dimensional vibration signal data of the wind power gear box in different running states, and preprocessing the data to acquire one-dimensional time sequence input vectors;
step S2: inputting one-dimensional time sequence input vectors into M convolutional network feature extraction modules respectively in a parallel mode, performing multi-scale time feature learning, and extracting multi-scale time features under different filter scales, wherein M is an integer and is greater than 1; splicing the acquired multi-scale time characteristics to obtain a one-dimensional characteristic vector;
s3, inputting the one-dimensional feature vector acquired in the S2 into a full connection layer, inputting the obtained output into a single classification objective function, and training through first class data to obtain a network model; the network model obtained by training comprises the following specific steps:
step S31, defining the wind power gear box fault detection task as a gear box abnormity detection task;
s32, inputting the one-dimensional feature vector obtained in the S2 into a full connection layer, and training through first class data to obtain a network model; the network model is obtained by training first-class data, and the computational expression of the single-classification objective function is as follows:
where n represents the number of one-dimensional time series input vectors, yiRepresenting the input vector of the ith one-dimensional time sequence after preprocessing, w representing all network layer parameters of the model, lambda representing a hyper-parameter, L representing the number of layers of a convolutional network, wlParameter representing the l-th layer of convolutional network, | | · | | non-calculationFDenotes the Frobenius norm, phi (y)i(ii) a w) multidimensional data representing a fully connected layer output of the network model, c represents an origin of the network model, the origin being obtained by directly inputting the first type of data into the network model;
step S4: inputting the test sample containing the first kind of data and the second kind of data into the network model obtained in step S3 to obtain an abnormal score value of the test sample, thereby obtaining a fault detection result.
In a preferred embodiment, the step S1 includes the following steps:
step S11, standardizing the acquired one-dimensional vibration time sequence data by adopting a z-score method, and calculating the expression as follows:
wherein y is one-dimensional time series data after normalization processing, x is original one-dimensional vibration time series data, and μ and σ are a mean value and a standard deviation of the original one-dimensional time series data, respectively;
step S12, the normalized one-dimensional time-series data is divided into a plurality of non-overlapping segments with the length of N, and N one-dimensional time-series input vectors are obtained.
In another preferred embodiment, in step S2, the size of the one-dimensional time-series input vector is 1 × N, where N is the number of sampling points, i.e., the length of each segment; and three convolutional network feature extraction modules are adopted to carry out multi-scale time feature learning, and the method specifically comprises the following steps:
step S21, respectively inputting the one-dimensional time sequence input vectors obtained in the step S1 to three convolutional network feature extraction modules with different filter scales in parallel, wherein the filter of each convolutional network feature extraction module slides along the direction of a time axis;
step S22, setting the number of layers of each convolutional network feature extraction module, wherein each module comprises three convolutional layers, three batch normalization layers, three regularization layers and a global average pooling layer, and the size of a filter of each global average pooling layer is the same;
and step S23, splicing the time features extracted under different filter scales to obtain a one-dimensional feature vector.
Preferably, the first type of data refers to data obtained under a healthy condition of the gearbox, and the second type of data refers to data under a faulty condition of the gearbox.
In another preferred embodiment, the step S4 specifically includes the following steps:
step S41, inputting the test sample into the network model obtained in the step S3 to obtain an abnormal point value; meanwhile, a group of abnormal score values are obtained from the first-class data in the step S3 through a network model, the maximum value of the group of abnormal score values is taken as a threshold value, and whether the wind power gear box is in a healthy state or not is detected by comparing the threshold value with the abnormal score value of the test sample; wherein the calculation expression of the abnormality score value is as follows:
s(y)=||φ(y;ω*)-c||2
wherein phi (y; w)*) Output, w, of the full-connectivity-layer network representing the model obtained by training*Representing all network layer parameters of the model resulting from the training.
Compared with the prior art, the invention has the technical progress that:
the invention provides a wind power gear box fault detection method based on a multi-scale single-classification convolutional network, which is characterized in that convolutional network feature extraction modules with different convolutional kernel structures are designed, and time features of gear box vibration signals are dug under multiple filter scales in a parallel mode.
Drawings
FIG. 1 is a flow chart of one embodiment of a wind turbine gearbox fault detection method based on a multi-scale single classification convolutional network of the present invention;
FIG. 2 is a flow diagram of multi-scale temporal feature learning according to an embodiment of the present invention;
FIG. 3 is a graph of average test results for condition one in accordance with an embodiment of the present invention;
FIG. 4 is a graph of the average results of condition two detection in one embodiment of the present invention.
Detailed Description
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
The embodiment of the invention adopts a gear box fault simulation experiment table which simulates a wind power gear box system, the collected vibration data is divided into bearing data and gear data, and the working conditions comprise two working conditions of 20Hz-0V and 30Hz-2V in rotating speed-load configuration. The embodiment of the invention uses a gear vibration signal data set of the planetary gearbox in the x direction, and each working condition comprises 1 normal state and 4 fault states.
The embodiment of the invention adopts a flow of a wind power gear box fault detection method based on a multi-scale single-classification convolutional network as shown in figure 1, and the method comprises the following steps:
step S1: the method comprises the following steps of collecting original vibration signals of a wind power gear box state monitoring system, obtaining one-dimensional vibration signal data in different running states, and preprocessing the data to obtain one-dimensional time sequence input vectors, wherein the specific method comprises the following steps:
step S11, standardizing the acquired one-dimensional time series data by adopting a z-score method, and calculating the expression as follows:
wherein y is one-dimensional time series data after normalization processing, x is original one-dimensional time series data, and μ and σ are mean and standard deviation of the one-dimensional time series data, respectively;
step S12, dividing the standardized one-dimensional time sequence data into a plurality of non-overlapping segments with the length of N to obtain N one-dimensional time sequence input vectors;
step S2: preprocessing an original vibration signal, as shown in fig. 2, inputting one-dimensional time sequence input vectors of first-class data into M convolutional network feature extraction modules respectively in a parallel manner, performing multi-scale time feature learning, and extracting multi-scale time features under different filter scales, wherein M is an integer and is greater than 1; splicing the acquired multi-scale time characteristics to obtain a one-dimensional characteristic vector, which comprises the following specific steps:
step S21, respectively inputting the one-dimensional time sequence input vectors obtained in the step S1 to three convolutional network feature extraction modules with different filter scales in parallel, wherein the filter of each convolutional network feature extraction module slides along the direction of a time axis;
step S22, setting the number of layers of each convolutional network feature extraction module, wherein each module comprises three convolutional layers, three batch normalization layers, three regularization layers and a global average pooling layer, and the size of a filter of each global average pooling layer is the same;
s23, splicing the time features extracted under different filter scales to obtain a one-dimensional feature vector;
step S3: inputting the preprocessed one-dimensional time sequence vectors into a plurality of convolution networks in parallel for multi-scale time feature learning, namely inputting the one-dimensional feature vectors obtained in the step S2 into a full connection layer, inputting the obtained output vectors into a single classification objective function, and training based on the first type of data to obtain a network model, wherein the method specifically comprises the following steps:
step S31, defining the wind power gear box fault detection task as a gear box abnormity detection task;
s32, inputting the one-dimensional feature vector obtained in the S2 into a full connection layer, and training through first class data to obtain a network model; the first type of data refers to data under the condition that the gearbox is in a healthy state, and the second type of data refers to data under the condition that the gearbox is in a fault state; the network model is obtained by training first-class data, and the computational expression of the single-classification objective function is as follows:
where n represents the number of one-dimensional time series input vectors, yiRepresenting the input vector of the ith one-dimensional time sequence after preprocessing, w representing all network layer parameters of the model, lambda representing a hyper-parameter, L representing the number of layers of a convolutional network, wlParameter representing the l-th layer of convolutional network, | | · | | non-calculationFDenotes the Frobenius norm, phi (y)i(ii) a w) represents multidimensional data output by a full connection layer of the network model, c represents an origin of the network model, and the origin is obtained by directly inputting the first type of data into the network model;
step S4: inputting the obtained multi-scale time characteristics into a single classifier to generate an abnormal score value, namely inputting the test sample into the network model obtained in the step S3 to obtain the abnormal score value, and generating a final abnormal detection result, wherein the method specifically comprises the following steps: inputting the test sample into the network model obtained in the step S3 to obtain an abnormal score value; meanwhile, a group of abnormal score values are obtained from the first-class data in the step S3 through a network model, the maximum value of the group of abnormal score values is taken as a threshold value, and whether the wind power gear box is in a healthy state or not is detected by comparing the threshold value with the abnormal score value of the test sample; wherein the calculation expression of the abnormality score value is as follows:
s(y)=||φ(y;ω*)-c||2
wherein phi (y; w)*) Output, w, of the fully-connected layer network representing the trained network model*All network layer parameters of the trained network model are represented.
In one embodiment of the invention, two operation conditions of the wind power gearbox system are considered together, wherein the first operation condition is a condition that the rotating speed-load is set to be 20Hz-0V in the first embodiment, and the second operation condition is a condition that the rotating speed-load is set to be 30Hz-2V in the second embodiment, each operation condition comprises a normal state and four fault states, and the four fault states are a gear tooth surface peeling state, a gear root crack state, a gear tooth breaking state and a gear surface abrasion state respectively. In order to effectively detect the wind power gear fault, the average result of ten times of repeated operation is used as the final diagnosis result in the experiment.
Fig. 3 shows a graph of the average detection result of the first operating condition, and fig. 4 shows a graph of the average detection result of the second operating condition. Compared with the traditional Machine learning method, the traditional Machine learning method comprises a Local Outlier Factor (LOF) algorithm, a single-Class Support Vector Machine (One-Class-SVM), an isolated Forest (IForest), and the Area under the curve (AUC) value is obviously improved, so that the enhanced fault detection performance is obtained.
The method provided by the embodiment of the invention is used for carrying out feature extraction and classification on the vibration data of the wind power gearbox so as to realize gearbox fault detection, and the core of the method is to obtain better fault detection performance. The method comprises the steps of obtaining one-dimensional vibration signal data of the gear box, preprocessing the one-dimensional vibration signal data to obtain one-dimensional time sequence input vectors, designing a convolution network time characteristic extraction module with different convolution kernel structures, learning multi-scale time characteristics of the vibration data of the gear box, splicing the vibration data, inputting the vibration data into a full-connection layer network and a single-classification target function, obtaining different abnormal score values by utilizing the characteristic that the multi-scale time characteristics of the first type of vibration data and the second type of vibration data are different, and comparing the abnormal score values with a threshold value to further detect whether the wind power gear box is in a healthy state. This result further illustrates that the present invention is worth applying to the fault detection of an actual wind power gearbox.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements made to the technical solution of the present invention by those skilled in the art without departing from the spirit of the present invention shall fall within the protection scope defined by the claims of the present invention.
Claims (5)
1. A wind power gear box fault detection method based on a multi-scale single-classification convolutional network is characterized by comprising the following steps:
step S1: acquiring one-dimensional vibration signal data of the wind power gear box in different running states, and preprocessing the data to acquire one-dimensional time sequence input vectors;
step S2: inputting one-dimensional time sequence input vectors into M convolutional network feature extraction modules respectively in a parallel mode, performing multi-scale time feature learning, and extracting multi-scale time features under different filter scales, wherein M is an integer and is greater than 1; splicing the acquired multi-scale time characteristics to obtain a one-dimensional characteristic vector;
s3, inputting the one-dimensional feature vector acquired in the S2 into a full connection layer, inputting the obtained output into a single classification objective function, and training through first class data to obtain a network model; the network model obtained by training comprises the following specific steps:
step S31, defining the wind power gear box fault detection task as a gear box abnormity detection task;
s32, inputting the one-dimensional feature vector obtained in the S2 into a full connection layer, and training through first class data to obtain a network model; the network model is obtained by training first-class data, and the computational expression of the single-classification objective function is as follows:
wherein n represents a one-dimensional timeNumber of sequential input vectors, yiRepresenting the input vector of the ith one-dimensional time sequence after preprocessing, w representing all network layer parameters of the model, lambda representing a hyper-parameter, L representing the number of layers of a convolutional network, wlParameter representing the l-th layer of convolutional network, | | · | | non-calculationFDenotes the Frobenius norm, phi (y)i(ii) a w) multidimensional data representing a fully connected layer output of the network model, c represents an origin of the network model, the origin being obtained by directly inputting the first type of data into the network model;
step S4: inputting the test sample containing the first kind of data and the second kind of data into the network model obtained in step S3 to obtain an abnormal score value of the test sample, thereby obtaining a fault detection result.
2. The wind power gearbox fault detection method based on the multi-scale single classification convolutional network as claimed in claim 1, wherein the step S1 comprises the following steps:
step S11, standardizing the acquired one-dimensional vibration time sequence data by adopting a z-score method, and calculating the expression as follows:
wherein y is one-dimensional time series data after normalization processing, x is original one-dimensional vibration time series data, and μ and σ are a mean value and a standard deviation of the original one-dimensional time series data, respectively;
step S12, the normalized one-dimensional time-series data is divided into a plurality of non-overlapping segments with the length of N, and N one-dimensional time-series input vectors are obtained.
3. The wind power gearbox fault detection method based on the multi-scale single-classification convolutional network as claimed in claim 1, wherein in step S2, the size of the one-dimensional time series input vector is 1 × N, where N is the number of sampling points, i.e. the length of each segment; and three convolutional network feature extraction modules are adopted to carry out multi-scale time feature learning, and the method specifically comprises the following steps:
step S21, respectively inputting the one-dimensional time sequence input vectors obtained in the step S1 to three convolutional network feature extraction modules with different filter scales in parallel, wherein the filter of each convolutional network feature extraction module slides along the direction of a time axis;
step S22, setting the number of layers of each convolutional network feature extraction module, wherein each module comprises three convolutional layers, three batch normalization layers, three regularization layers and a global average pooling layer, and the size of a filter of each global average pooling layer is the same;
and step S23, splicing the time features extracted under different filter scales to obtain a one-dimensional feature vector.
4. The wind power gearbox fault detection method based on the multi-scale single classification convolutional network as claimed in claim 1, wherein the first class of data refers to data under the healthy state of the gearbox, and the second class of data refers to data under the fault state of the gearbox.
5. The wind power gearbox fault detection method based on the multi-scale single-classification convolutional network as claimed in claim 1, wherein the step S4 specifically comprises the following steps:
step S41, inputting the test sample into the network model obtained in the step S3 to obtain an abnormal point value; meanwhile, a group of abnormal score values are obtained from the first-class data in the step S3 through a network model, the maximum value of the group of abnormal score values is taken as a threshold value, and whether the wind power gear box is in a healthy state or not is detected by comparing the threshold value with the abnormal score value of the test sample; wherein the calculation expression of the abnormality score value is as follows:
s(y)=||φ(y;ω*)-c||2
wherein phi (y; w)*) Output, w, of the full-connectivity-layer network representing the model obtained by training*Representing all network layer parameters of the model resulting from the training.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110547118.6A CN113240022A (en) | 2021-05-19 | 2021-05-19 | Wind power gear box fault detection method of multi-scale single-classification convolutional network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110547118.6A CN113240022A (en) | 2021-05-19 | 2021-05-19 | Wind power gear box fault detection method of multi-scale single-classification convolutional network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113240022A true CN113240022A (en) | 2021-08-10 |
Family
ID=77137674
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110547118.6A Pending CN113240022A (en) | 2021-05-19 | 2021-05-19 | Wind power gear box fault detection method of multi-scale single-classification convolutional network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113240022A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116718894A (en) * | 2023-06-19 | 2023-09-08 | 上饶市广强电子科技有限公司 | Circuit stability test method and system for corn lamp |
CN117515131A (en) * | 2024-01-04 | 2024-02-06 | 之江实验室 | Method, device, storage medium and equipment for monitoring abrasion of planetary reducer |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108154504A (en) * | 2017-12-25 | 2018-06-12 | 浙江工业大学 | A kind of detection method of the Surface Defects in Steel Plate based on convolutional neural networks |
CN109598283A (en) * | 2018-10-22 | 2019-04-09 | 中南大学 | A kind of aluminium electroloysis degree of superheat recognition methods based on semi-supervised extreme learning machine |
CN110940523A (en) * | 2019-11-29 | 2020-03-31 | 盐城工学院 | Unsupervised domain adaptive fault diagnosis method |
CN110991424A (en) * | 2019-12-25 | 2020-04-10 | 安徽工业大学 | Fault diagnosis method based on minimum entropy deconvolution and stacking sparse self-encoder |
CN111400547A (en) * | 2020-03-05 | 2020-07-10 | 西北工业大学 | Human-computer cooperation video anomaly detection method |
CN111426950A (en) * | 2020-03-19 | 2020-07-17 | 燕山大学 | Wind driven generator fault diagnosis method of multi-scale space-time convolution depth belief network |
CN111444747A (en) * | 2019-01-17 | 2020-07-24 | 复旦大学 | Epileptic state identification method based on transfer learning and cavity convolution |
US20200251127A1 (en) * | 2019-02-06 | 2020-08-06 | Hitachi, Ltd. | Abnormal sound detection device and abnormal sound detection method |
CN111898095A (en) * | 2020-07-10 | 2020-11-06 | 佛山科学技术学院 | Deep migration learning intelligent fault diagnosis method and device, storage medium and equipment |
CN112254964A (en) * | 2020-09-03 | 2021-01-22 | 太原理工大学 | Rolling bearing fault diagnosis method based on rapid multi-scale convolution neural network |
-
2021
- 2021-05-19 CN CN202110547118.6A patent/CN113240022A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108154504A (en) * | 2017-12-25 | 2018-06-12 | 浙江工业大学 | A kind of detection method of the Surface Defects in Steel Plate based on convolutional neural networks |
CN109598283A (en) * | 2018-10-22 | 2019-04-09 | 中南大学 | A kind of aluminium electroloysis degree of superheat recognition methods based on semi-supervised extreme learning machine |
CN111444747A (en) * | 2019-01-17 | 2020-07-24 | 复旦大学 | Epileptic state identification method based on transfer learning and cavity convolution |
US20200251127A1 (en) * | 2019-02-06 | 2020-08-06 | Hitachi, Ltd. | Abnormal sound detection device and abnormal sound detection method |
CN110940523A (en) * | 2019-11-29 | 2020-03-31 | 盐城工学院 | Unsupervised domain adaptive fault diagnosis method |
CN110991424A (en) * | 2019-12-25 | 2020-04-10 | 安徽工业大学 | Fault diagnosis method based on minimum entropy deconvolution and stacking sparse self-encoder |
CN111400547A (en) * | 2020-03-05 | 2020-07-10 | 西北工业大学 | Human-computer cooperation video anomaly detection method |
CN111426950A (en) * | 2020-03-19 | 2020-07-17 | 燕山大学 | Wind driven generator fault diagnosis method of multi-scale space-time convolution depth belief network |
CN111898095A (en) * | 2020-07-10 | 2020-11-06 | 佛山科学技术学院 | Deep migration learning intelligent fault diagnosis method and device, storage medium and equipment |
CN112254964A (en) * | 2020-09-03 | 2021-01-22 | 太原理工大学 | Rolling bearing fault diagnosis method based on rapid multi-scale convolution neural network |
Non-Patent Citations (2)
Title |
---|
LEI ZHANG;DONG LI;QUANSHENG GUO: "《Deep Learning From Spatio-Temporal Data Using Orthogonal Regularizaion Residual CNN for Air Prediction》", 《IEEE ACCESS》 * |
余增: "《深度学习中特征表示方法研究与应用》", 《中国博士学位论文全文数据库 (信息科技辑)》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116718894A (en) * | 2023-06-19 | 2023-09-08 | 上饶市广强电子科技有限公司 | Circuit stability test method and system for corn lamp |
CN116718894B (en) * | 2023-06-19 | 2024-03-29 | 上饶市广强电子科技有限公司 | Circuit stability test method and system for corn lamp |
CN117515131A (en) * | 2024-01-04 | 2024-02-06 | 之江实验室 | Method, device, storage medium and equipment for monitoring abrasion of planetary reducer |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Saufi et al. | Gearbox fault diagnosis using a deep learning model with limited data sample | |
Saufi et al. | Challenges and opportunities of deep learning models for machinery fault detection and diagnosis: A review | |
Han et al. | An adaptive spatiotemporal feature learning approach for fault diagnosis in complex systems | |
Jiao et al. | Deep coupled dense convolutional network with complementary data for intelligent fault diagnosis | |
CN111426950B (en) | Wind driven generator fault diagnosis method of multi-scale space-time convolution depth belief network | |
CN113240022A (en) | Wind power gear box fault detection method of multi-scale single-classification convolutional network | |
CN113076834B (en) | Rotating machine fault information processing method, processing system, processing terminal, and medium | |
Sun et al. | A robust approach of multi-sensor fusion for fault diagnosis using convolution neural network | |
CN112163472A (en) | Rolling bearing diagnosis method based on multi-view feature fusion | |
CN111198098A (en) | Wind power generator bearing fault prediction method based on artificial neural network | |
Marugán et al. | Multivariable analysis for advanced analytics of wind turbine management | |
Zare et al. | Convolutional neural networks for wind turbine gearbox health monitoring. | |
CN111855202A (en) | Gear box fault diagnosis method and system | |
Baltazar et al. | A review on neurocomputing based wind turbines fault diagnosis and prognosis | |
Mohammadi et al. | DeepWind: an accurate wind turbine condition monitoring framework via deep learning on embedded platforms | |
Zhu et al. | Fault diagnosis of wind turbine gearbox under limited labeled data through temporal predictive and similarity contrast learning embedded with self-attention mechanism | |
CN112734001A (en) | Wind power transmission chain intelligent fault diagnosis method based on order spectrum migration | |
CN113673442A (en) | Variable working condition fault detection method based on semi-supervised single classification network | |
Shaw et al. | Prediction of remaining useful life of wind turbine shaft bearings using machine learning | |
Long et al. | Wind turbine anomaly identification based on improved deep belief network with SCADA data | |
Cao et al. | A novel method for detection of wind turbine blade imbalance based on multi-variable spectrum imaging and convolutional neural network | |
CN115165364A (en) | Wind turbine generator bearing fault diagnosis model construction method based on transfer learning | |
Liu et al. | Acoustic emission analysis for wind turbine blade bearing fault detection using sparse augmented Lagrangian algorithm | |
CN114897292A (en) | Fault diagnosis method and system based on multi-scale residual convolution variational network | |
Saha et al. | Investigating Rotor Conditions on Wind Turbines Using Integrating Tree Classifiers |
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: 20210810 |
|
RJ01 | Rejection of invention patent application after publication |