CN114358078A - Deep learning detection method for motor vibration - Google Patents
Deep learning detection method for motor vibration Download PDFInfo
- Publication number
- CN114358078A CN114358078A CN202111678049.9A CN202111678049A CN114358078A CN 114358078 A CN114358078 A CN 114358078A CN 202111678049 A CN202111678049 A CN 202111678049A CN 114358078 A CN114358078 A CN 114358078A
- Authority
- CN
- China
- Prior art keywords
- vibration
- motor
- top plate
- neural network
- network model
- 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
Images
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to a motor vibration deep learning detection method, which comprises the following steps: the motor is provided with a contact type vibration tool used for limiting the motor so that the motor is tightly attached to the vibration sensor; building and training a convolutional neural network model; the vibration sensor collects motor vibration signals in real time; preprocessing the collected motor vibration signals and extracting two-dimensional vibration characteristics; inputting the two-dimensional vibration characteristics into the convolutional neural network model, and calculating to obtain a two-classification probability value; and comparing the two classification probability values with a preset threshold value to determine the motor state, wherein the motor state comprises normal vibration and abnormal vibration. According to the invention, the accuracy of the acquired vibration signal is ensured by arranging the contact type vibration tool, and the accuracy of judging abnormal sound of the motor is effectively improved by extracting two-dimensional vibration characteristics in the vibration signal, adopting a deep learning method and utilizing a three-branch convolutional neural network model to carry out two-class judgment.
Description
Technical Field
The invention relates to the technical field of neural networks, in particular to a deep learning detection method for motor vibration.
Background
The judgment of abnormal motor sound is a problem which is very important for motor manufacturers, and is related to the improvement of product technology and the improvement of product quality. The traditional motor abnormal sound analyzing and judging method mainly comprises a sound signal collecting analysis method and a vibration signal collecting analysis method, and compared with sound signals, the vibration signals can avoid the influence of environmental noise and are more generally adopted in practice.
However, in practice, many motors or surfaces are uneven or have small volumes, and the vibration sensors cannot be directly attached to the motors or surfaces, so that the vibration signals acquired by the vibration sensors are inaccurate. Meanwhile, because the frequency band of the vibration signal is limited, the frequency range of a general vibration sensor is within 10000Hz, and the analysis accuracy is lower when the frequency spectrum analysis is carried out on the vibration signal.
Disclosure of Invention
In order to overcome the technical problems, the invention provides a motor vibration deep learning detection method, which is used for improving the acquisition accuracy and the analysis accuracy of a vibration sensor, so that the accuracy of judging abnormal sound of a motor is improved.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a motor vibration deep learning detection method, which comprises the following steps:
arranging a contact type vibration tool on the motor, wherein the contact type vibration tool is used for limiting the motor so that the motor is tightly attached to the vibration sensor;
building a convolutional neural network model, and training the convolutional neural network model by using a pre-built motor detection training set;
the vibration sensor collects motor vibration signals in real time;
preprocessing the collected motor vibration signals and extracting two-dimensional vibration characteristics;
inputting the two-dimensional vibration characteristics into the convolutional neural network model, and calculating to obtain a two-classification probability value;
and comparing the two classification probability values with a preset threshold value to determine the motor state, wherein the motor state comprises normal vibration and abnormal vibration.
Preferably, the contact vibration tool includes:
the device comprises a base, an upper supporting top plate and a lower supporting top plate, wherein the upper supporting top plate and the lower supporting top plate are arranged on the base; a limiting hole is formed in the upper supporting top plate; the upper supporting top plate is provided with a fixing structure for fixing the motor so that the motor is tightly clamped on the limiting hole; and a spring is arranged at the position of the lower supporting top plate corresponding to the limiting hole and used for providing supporting force for the vibration sensor so that the vibration sensor is tightly attached to the motor through the limiting hole.
Preferably, the fixing structure comprises at least two quick clamps; one end of the rapid clamp is provided with a fixing part and is fixed on the upper supporting top plate; the other end is arranged as a clamping part which is clamped on the motor.
Preferably, the vibration sensor is sleeved in the spring, and the spring is limited by the lower support top plate and can only move along the vertical direction.
Preferably, the upper supporting top plate and the lower supporting top plate are connected with the base through supporting rib plates; the upper supporting top plate and the lower supporting top plate are provided with corresponding connecting holes on the supporting rib plate; and the bolt is connected with the upper supporting top plate and the supporting rib plate through the connecting hole, is connected with the upper supporting top plate and the supporting rib plate, and is used for adjusting the height of the upper supporting top plate and the lower supporting top plate relative to the base.
Preferably, the convolutional neural network model includes: the system comprises at least one convolution layer block, at least one pooling layer and at least one full-connection layer which are connected in sequence; the convolutional layer block comprises a 3 × 3 convolutional layer branch, a 1 × 1 convolutional layer branch and a residual error branch; the three branches are respectively connected with a batch of standard layers and then linearly added, and the two classification probability values are output after the ReLU activation function operation.
Preferably, the batch specification layer is used for forcibly pulling back the distribution of any neuron of each layer of the neural network to a standard normal distribution with a mean value of 0 and a variance of 1.
Preferably, the training of the convolutional neural network model comprises the following steps:
s401, initializing the weight value of the convolutional neural network model;
s402, extracting two-dimensional vibration characteristics from the motor detection training set, inputting the two-dimensional vibration characteristics into the initialized convolutional neural network model, and performing iterative training on the convolutional neural network model; wherein the weight value is updated once per iteration;
s403, extracting two-dimensional vibration characteristics from a pre-established motor detection test set, inputting the two-dimensional vibration characteristics into the convolutional neural network model obtained through the training in the step S402, and outputting a test result;
s404, judging whether the detection precision of the convolutional neural network model obtained by the training in the step S402 on the motor state meets the preset requirement or not according to the test result; if the result is consistent, the training is finished; if not, returning to step S402 to continue training until the convolutional neural network model meeting the preset requirements is obtained by training.
Preferably, the motor detection training set includes a pre-collected motor vibration signal, the collected motor being pre-flagged as "normal" or "fault".
Preferably, the extracting the two-dimensional vibration feature comprises the following steps;
s501, collecting vibration signals of the motor;
s502, framing the vibration signal according to a preset frame time length to obtain N frames of vibration fragments; calculating a Mel filter bank according to the vibration signals of each frame, and obtaining logarithmic Mel feature vectors after logarithm;
and S503, combining the logarithmic Mel feature vectors of all the frames to generate a two-dimensional vibration feature matrix.
The invention has the beneficial effects that:
in the vibration signal acquisition process, the vibration sensor is fixed on the motor with small volume or uneven surface by adopting the contact type vibration tool, so that the accuracy of the acquired vibration signal is ensured, and meanwhile, the abnormal sound of the motor is predicted by extracting the two-dimensional vibration characteristics in the vibration signal and the three-branch convolution neural network model, so that the accuracy of judging the abnormal sound of the motor is greatly improved.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic diagram of a convolutional neural network detection flow of a motor vibration detection method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a contact type vibration tool in a motor vibration detection method according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a convolutional neural network of a method for detecting vibration of a motor according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a convolution block structure of a motor vibration detection method according to an embodiment of the present invention.
Reference numerals
1 base
2 upper supporting top plate
3 lower supporting top plate
4 electric machine
5 fixing structure
6 vibration sensor
7 spring
8 support rib plate
9 bolt
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more clearly and completely apparent, the technical solutions in the embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
In some of the flows described in the present specification and claims and in the above figures, a number of operations are included which occur in a particular order, but it should be clearly understood that these operations may be performed out of order or in parallel with the order in which they occur, the order of the operations being, for example, S401, S402, etc., merely to distinguish between various operations, and the order of the operations itself does not represent any order of performance. Additionally, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel.
The method for detecting deep learning of motor vibration according to the present invention will be described in detail with reference to the accompanying drawings and specific embodiments.
The invention provides a motor vibration deep learning detection method, which comprises the following steps:
arranging a contact type vibration tool on the motor, wherein the contact type vibration tool is used for limiting the motor so that the motor is tightly attached to the vibration sensor; the contact type vibration tool can fix the vibration sensor on a small motor or a motor with an irregular shape, and can still ensure the firmness of the vibration sensor even under the condition that a carrier provided with the motor moves at a high speed, so that the vibration sensor is not easy to shake, misplace or fall.
Specifically, taking one of the possible embodiments as an example for illustration, the present invention can be applied to abnormal sound detection of the motor of the automobile rearview mirror, because the motor of the automobile rearview mirror has a small volume and surface unevenness, and the vibration sensor cannot be directly fixed on the motor of the automobile rearview mirror. Therefore, when the abnormal sound of the automobile rearview mirror motor is tested, the contact type vibration tool provided with the vibration sensor needs to be fixed on the automobile rearview mirror motor at first, so that the automobile rearview mirror motor is tightly attached to the vibration sensor, the defect that the vibration sensor cannot be fixed independently is overcome, the accuracy of collecting vibration signals of the automobile rearview mirror motor is improved, and meanwhile, the vibration signals of the motor can still be firmly set in the driving process of an automobile to be collected.
Preferably, the contact type vibration tool is connected with a processor, and the acquired vibration signal or vibration is sent to the processor. The connection mode of the contact type vibration tool and the processor includes, but is not limited to, communication connection, signal line connection and the like, and the connection mode is claimed in the invention as long as the connection mode can transmit the vibration signal to the processor. It should be noted that the vibration signal mentioned in the present invention includes, but is not limited to, vibration acceleration, vibration displacement, amplitude, vibration velocity, and the like.
Fig. 1 shows a schematic detection flow diagram of a convolutional neural network of a motor vibration detection method according to an embodiment of the present invention.
With reference to fig. 1, the present invention builds a convolutional neural network model, and trains the convolutional neural network model using a pre-established motor detection training set;
the vibration sensor collects motor vibration signals in real time; the vibration signal is less susceptible to environmental noise interference and has higher reliability than the sound signal.
Preprocessing the collected motor vibration signals and extracting two-dimensional vibration characteristics;
inputting the two-dimensional vibration characteristics into the convolutional neural network model, and calculating to obtain a two-classification probability value;
and comparing the two classification probability values with a preset threshold value to determine the motor state, wherein the motor state comprises normal vibration and abnormal vibration. Specifically, the two-dimensional vibration characteristics are input into the convolutional neural network model to obtain a binary probability value output by the convolutional neural network model, and when the probability value exceeds a preset value, the motor is indicated to be abnormal; and when the probability value does not reach the preset value, indicating that the motor is normal. Taking one possible embodiment as an example, the method converts the acquired vibration signal into two-dimensional vibration characteristics, inputs the two-dimensional vibration characteristics into the convolutional neural network model, and obtains the two-classification probability value output by the convolutional neural network model. For example: a probability value smaller than 0.5 can be set to indicate that the motor sounds normally; the probability value is larger than 0.5, which indicates that the motor sound is abnormal.
Preferably, as shown in fig. 2, a schematic structural diagram of a contact type vibration tool of a motor vibration detection method according to an embodiment of the present invention is shown, where the contact type vibration tool includes:
the device comprises a base 1, an upper supporting top plate 2 and a lower supporting top plate 3, wherein the upper supporting top plate and the lower supporting top plate are arranged on the base 1; a limiting hole (not shown in the figure) is formed in the upper supporting top plate 2; the upper supporting top plate 2 is provided with a fixing structure 5 for fixing the motor 4 so that the motor 4 is tightly clamped on the limiting hole; and a spring 7 is arranged at the position of the lower supporting top plate 3 corresponding to the limiting hole and used for providing supporting force for the vibration sensor 6 so that the vibration sensor 6 is tightly attached to the motor 4 through the limiting hole. The contact type vibration tool provided by the invention has the advantages that the vibration sensor 6 can be directly contacted with the motor 4 and is firmly fixed, and the accuracy of the vibration signal of the motor to be measured is greatly improved.
Preferably, as shown in fig. 2, the fixing structure 5 comprises at least two quick clamps; one end of the rapid clamp is arranged as a fixed part and is fixed on the upper supporting top plate 2; the other end sets up to joint portion, and the joint is in on the motor 4, support motor 4 through the joint portion of quick anchor clamps promptly, make it fix on spacing hole to guarantee motor 4's fastness, and motor 4 and vibration sensor 6's relative stability.
Preferably, as shown in fig. 2, the vibration sensor 6 is sleeved in the spring 7, and the spring 7 is limited by the lower supporting top plate 3 and can only move in the vertical direction. Specifically, the vibration sensor 6 is sleeved in the spring 7 and directly props against the bottom of the motor 4, and can record vibration signals when the motor runs. The spring 7 supports the vibration sensor 6 and is limited by the lower supporting top plate 3 to move only in the vertical direction.
Preferably, as shown in fig. 2, the upper supporting top plate 2 and the lower supporting top plate 3 are connected with the base 1 through a supporting rib plate 8; the upper supporting top plate 2 and the lower supporting top plate 3 are provided with corresponding connecting holes with the supporting rib plates 8; the bolts 9 are connected with the upper supporting top plate 2 and the supporting rib plate 8 through the connecting holes, and connected with the upper supporting top plate 2 and the supporting rib plate 8, and are used for adjusting the height of the upper supporting top plate 2 and the lower supporting top plate 3 relative to the base 1.
Preferably, as shown in fig. 3, a schematic structural diagram of a convolutional neural network of a motor vibration detection method according to an embodiment of the present invention is shown; the convolutional neural network model includes: the system comprises at least one convolution layer block, at least one pooling layer and at least one full-connection layer which are connected in sequence; fig. 4 is a schematic diagram illustrating a convolution block structure of a motor vibration detection method according to an embodiment of the present invention, where the convolution block includes a 3 × 3 convolution layer branch, a 1 × 1 convolution layer branch, and a residual branch; the three branches are respectively connected with a batch of standard layers and then linearly added, new two-dimensional vibration characteristics are output after the ReLU activation function operation, and the binary probability value is calculated.
Specifically, the convolutional neural network model is formed by sequentially connecting at least one convolutional layer block, at least one pooling layer and at least one fully-connected layer, as shown in fig. 2. The convolution layer block is provided with a convolution layer and is used for extracting input data characteristics and outputting a characteristic diagram. The number of convolutional blocks is determined experimentally. The pooling layer is used to sample or aggregate the feature map, for example, by selecting the maximum (or average) value of the region instead of the region, which greatly reduces the sensitivity of the feature while also reducing the computational complexity of the feature while preserving feature information. If we say that convolutional layers, pooling layers, etc. are operations that map raw data to hidden layer feature space, fully-connected layers play a role in mapping the learned "distributed feature representation" to sample label space. The invention adopts the ReLU activation function to lead the convolutional neural network model to train more quickly, increase the nonlinearity of the convolutional neural network model and prevent the gradient from disappearing.
It should be noted that the residual branch of the convolutional layer block of the present invention refers to the structure of the ResNet model. However, the convolution layer block of the present invention is different from the ResNet model convolution layer block in that:
1) the invention increases 1 x 1 convolutional layer branches;
2) the residual branch of the ResNet model is a cross-layer connection, while the convolutional layer block of the present invention is a single connection layer.
Preferably, the convolution layer block includes convolution kernels having a convolution kernel size of 3 × 3.
Preferably, the batch specification layer is used for forcibly pulling back the distribution of any neuron of each layer of the neural network to a standard normal distribution with a mean value of 0 and a variance of 1. The specific batch of standard layers can forcibly pull back the distribution of any neuron of each layer of neural network to the standard normal distribution with the mean value of 0 and the variance of 1 through a certain standardization means, so that the problem of gradient disappearance is avoided, and the training speed of the convolutional neural network model can be accelerated.
Preferably, the training of the convolutional neural network model comprises the following steps:
s401, initializing the weight value of the convolutional neural network model;
s402, extracting two-dimensional vibration characteristics from the motor detection training set, inputting the two-dimensional vibration characteristics into the initialized convolutional neural network model, and performing iterative training on the convolutional neural network model; wherein the weight value is updated once per iteration;
s403, extracting two-dimensional vibration characteristics from a pre-established motor detection test set, inputting the two-dimensional vibration characteristics into the convolutional neural network model obtained through the training in the step S402, and outputting a test result;
s404, judging whether the detection precision of the convolutional neural network model obtained by the training in the step S402 on the motor state meets the preset requirement or not according to the test result; if the result is consistent, the training is finished; if not, returning to step S402 to continue training until the convolutional neural network model meeting the preset requirements is obtained by training.
Preferably, the motor detection training set includes a pre-collected motor vibration signal, the collected motor being pre-flagged as "normal" or "fault". Specifically, in the embodiment of the present invention, a batch of motors that have been pre-marked with "normal" or "failure" tags are used, powered on to run, and vibration signals are collected in real time and then saved as a file. And preprocessing the file, namely extracting two-dimensional vibration characteristics to form a motor detection training set, and inputting the motor detection training set into a convolutional neural network model for training. After the model training is finished, vibration signals of the motor to be tested can be collected in real time, and the running state of the motor to be tested is predicted through the convolutional neural network model.
Preferably, the extracting the two-dimensional vibration feature comprises the following steps;
s501, collecting vibration signals of the motor;
s502, framing the vibration signal according to a preset frame time length to obtain N frames of vibration fragments; calculating a Mel filter bank according to the vibration signals of each frame, and obtaining logarithmic Mel feature vectors after logarithm;
and S503, combining the logarithmic Mel feature vectors of all the frames to generate a two-dimensional vibration feature matrix.
Compared with the prior art, the motor multi-signal deep learning detection method provided by the embodiment of the invention has the following beneficial effects:
according to the invention, in the vibration signal acquisition process, the contact type vibration tool is arranged, the vibration sensor is fixed on the motor with small volume or uneven surface, the accuracy of the acquired vibration signal is ensured, meanwhile, the abnormal sound of the motor is predicted through the three-branch convolution neural network model by extracting the two-dimensional vibration characteristics in the vibration signal, and the accuracy of judging the abnormal sound of the motor is greatly improved.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are given by way of illustration of the principles of the present invention, and that various changes and modifications may be made without departing from the spirit and scope of the invention as defined by the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (10)
1. A motor vibration deep learning detection method is characterized by comprising the following steps:
arranging a contact type vibration tool on the motor, wherein the contact type vibration tool is used for limiting the motor so that the motor is tightly attached to the vibration sensor;
building a convolutional neural network model, and training the convolutional neural network model by using a pre-built motor detection training set;
the vibration sensor collects motor vibration signals in real time;
preprocessing the collected motor vibration signals and extracting two-dimensional vibration characteristics;
inputting the two-dimensional vibration characteristics into the convolutional neural network model, and calculating to obtain a two-classification probability value;
and comparing the difference value of the two classification probability values with a preset threshold value to determine the motor state, wherein the motor state comprises normal vibration and abnormal vibration.
2. The motor vibration deep learning detection method according to claim 1, wherein the contact type vibration tool comprises:
the device comprises a base, an upper supporting top plate and a lower supporting top plate, wherein the upper supporting top plate and the lower supporting top plate are arranged on the base; a limiting hole is formed in the upper supporting top plate; the upper supporting top plate is provided with a fixing structure for fixing the motor so that the motor is tightly clamped on the limiting hole; and a spring is arranged at the position of the lower supporting top plate corresponding to the limiting hole and used for providing supporting force for the vibration sensor so that the vibration sensor is tightly attached to the motor through the limiting hole.
3. The method as claimed in claim 2, wherein the fixing structure comprises at least two fast clamps; one end of the rapid clamp is provided with a fixing part and is fixed on the upper supporting top plate; the other end is arranged as a clamping part which is clamped on the motor.
4. The motor vibration deep learning detection method according to claim 2, wherein the vibration sensor is sleeved in the spring, and the spring is limited by the lower supporting top plate and can only move in a vertical direction.
5. The motor vibration deep learning detection method according to claim 2, characterized in that the upper supporting top plate and the lower supporting top plate are connected with the base through supporting rib plates; the upper supporting top plate and the lower supporting top plate are provided with corresponding connecting holes on the supporting rib plate; and the bolt is connected with the upper supporting top plate and the supporting rib plate through the connecting hole, is connected with the upper supporting top plate and the supporting rib plate, and is used for adjusting the height of the upper supporting top plate and the lower supporting top plate relative to the base.
6. The method according to claim 1, wherein the convolutional neural network model comprises: the system comprises at least one convolution layer block, at least one pooling layer and at least one full-connection layer which are connected in sequence; the convolutional layer block comprises a 3 × 3 convolutional layer branch, a 1 × 1 convolutional layer branch and a residual error branch; the three branches are respectively connected with a batch of standard layers and then linearly added, and the two classification probability values are output after the ReLU activation function operation.
7. The method as claimed in claim 6, wherein the batch of standard layers is used to pull back the distribution of any neuron in each layer of neural network to a standard normal distribution with a mean value of 0 and a variance of 1.
8. The method for detecting deep learning of motor vibration according to claim 1, wherein the training of the convolutional neural network model comprises the following steps:
s401, initializing the weight value of the convolutional neural network model;
s402, extracting two-dimensional vibration characteristics from the motor detection training set, inputting the two-dimensional vibration characteristics into the initialized convolutional neural network model, and performing iterative training on the convolutional neural network model; wherein the weight value is updated once per iteration;
s403, extracting two-dimensional vibration characteristics from a pre-established motor detection test set, inputting the two-dimensional vibration characteristics into the convolutional neural network model obtained through the training in the step S402, and outputting a test result;
s404, judging whether the detection precision of the convolutional neural network model obtained by the training in the step S402 on the motor state meets the preset requirement or not according to the test result; if the result is consistent, the training is finished; if not, returning to step S402 to continue training until the convolutional neural network model meeting the preset requirements is obtained by training.
9. The method as claimed in claim 8, wherein the motor detection training set includes a pre-collected motor vibration signal, and the collected motor is pre-marked as "normal" or "fault".
10. The motor vibration deep learning detection method according to claim 1, wherein the extracting two-dimensional vibration features comprises the following steps;
s501, collecting vibration signals of the motor;
s502, framing the vibration signal according to a preset frame time length to obtain N frames of vibration fragments; calculating a Mel filter bank according to the vibration signals of each frame, and obtaining logarithmic Mel feature vectors after logarithm;
and S503, combining the logarithmic Mel feature vectors of all the frames to generate a two-dimensional vibration feature matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111678049.9A CN114358078A (en) | 2021-12-31 | 2021-12-31 | Deep learning detection method for motor vibration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111678049.9A CN114358078A (en) | 2021-12-31 | 2021-12-31 | Deep learning detection method for motor vibration |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114358078A true CN114358078A (en) | 2022-04-15 |
Family
ID=81104820
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111678049.9A Pending CN114358078A (en) | 2021-12-31 | 2021-12-31 | Deep learning detection method for motor vibration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114358078A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115856628A (en) * | 2023-02-28 | 2023-03-28 | 宁波慧声智创科技有限公司 | Micro-special motor acoustic quality detection method based on PSO-SVM detection model |
-
2021
- 2021-12-31 CN CN202111678049.9A patent/CN114358078A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115856628A (en) * | 2023-02-28 | 2023-03-28 | 宁波慧声智创科技有限公司 | Micro-special motor acoustic quality detection method based on PSO-SVM detection model |
| CN115856628B (en) * | 2023-02-28 | 2023-06-27 | 宁波慧声智创科技有限公司 | Micro-special motor acoustic quality detection method based on PSO-SVM detection model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111507233B (en) | Multi-mode information fusion intelligent vehicle pavement type identification method | |
| CN110704801B (en) | Bridge cluster structure operation safety intelligent monitoring and rapid detection complete method | |
| CN111562108A (en) | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC | |
| CN111695598A (en) | Monitoring data abnormity diagnosis method based on generation countermeasure network | |
| CN113091872B (en) | Method and device for diagnosing fault sensor | |
| CN112487890B (en) | Bearing acoustic signal fault diagnosis method based on parallel sparse filtering | |
| CN114358078A (en) | Deep learning detection method for motor vibration | |
| CN113029327B (en) | Tunnel fan embedded foundation damage identification method based on metric attention convolutional neural network | |
| CN111695452B (en) | RBF neural network-based parallel reactor internal aging degree assessment method | |
| CN110018322B (en) | Rotating speed detection method and system based on deep learning | |
| CN113406524B (en) | Inconsistent fault diagnosis method and system for power battery system | |
| CN113761649A (en) | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network | |
| JP6876589B2 (en) | Anomaly detection device, anomaly detection method, and anomaly detection program | |
| CN115112372A (en) | Bearing fault diagnosis method and device, electronic equipment and storage medium | |
| CN110082106B (en) | Bearing fault diagnosis method based on Yu norm deep measurement learning | |
| CN113392874B (en) | Abnormal state diagnosis method and device for rail vehicle and terminal equipment | |
| CN114861349A (en) | Rolling bearing RUL prediction method based on model migration and wiener process | |
| CN111062093A (en) | Automobile tire service life prediction method based on image processing and machine learning technology | |
| CN114330430A (en) | Elevator fault judgment method and system based on big data characteristic analysis | |
| CN113758709A (en) | Rolling bearing fault diagnosis method and system combining edge calculation and deep learning | |
| Chen et al. | Unsupervised damage identification scheme using PCA-reduced frequency response function and waveform chain code analysis | |
| CN112362214A (en) | Method and system for online identification of belt tension | |
| CN114386572A (en) | Motor multi-signal deep learning detection method | |
| CN114084764B (en) | Elevator transportation quality detection method and detection system | |
| CN113672859B (en) | Fault acoustic diagnosis system for switch machine |
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 |