CN109782167B - Gear reduction motor quality inspection device and method based on convolutional network depth model - Google Patents
Gear reduction motor quality inspection device and method based on convolutional network depth model Download PDFInfo
- Publication number
- CN109782167B CN109782167B CN201811633174.6A CN201811633174A CN109782167B CN 109782167 B CN109782167 B CN 109782167B CN 201811633174 A CN201811633174 A CN 201811633174A CN 109782167 B CN109782167 B CN 109782167B
- Authority
- CN
- China
- Prior art keywords
- motor
- gear reduction
- model
- reduction motor
- signal
- 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.)
- Active
Links
- 230000009467 reduction Effects 0.000 title claims abstract description 52
- 238000000034 method Methods 0.000 title abstract description 13
- 238000007689 inspection Methods 0.000 title abstract description 10
- 230000001133 acceleration Effects 0.000 claims abstract description 20
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 238000003908 quality control method Methods 0.000 claims abstract description 11
- 230000008901 benefit Effects 0.000 claims abstract description 6
- 238000010586 diagram Methods 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 14
- 238000013145 classification model Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 238000009432 framing Methods 0.000 claims description 8
- 239000002184 metal Substances 0.000 claims description 8
- 238000012795 verification Methods 0.000 claims description 6
- 229910000831 Steel Inorganic materials 0.000 claims description 5
- 239000010959 steel Substances 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 239000003292 glue Substances 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 239000003638 chemical reducing agent Substances 0.000 claims 1
- 238000013135 deep learning Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 210000005069 ears Anatomy 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 229910001220 stainless steel Inorganic materials 0.000 description 1
- 239000010935 stainless steel Substances 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a gear reduction motor quality control device and method based on a convolutional network depth model, wherein the device comprises the following components: the direct-current stabilized power supply is used for supplying power to the miniature direct-current gear reduction motor to be detected; the constant current adapter is used for supplying power to the acceleration sensor and amplifying motor vibration signals acquired by the acceleration sensor and then transmitting the motor vibration signals to the data acquisition device; the data collector is used for generating the amplified motor vibration signal into a digital signal and transmitting the digital signal to the PC; the acceleration sensor is used for collecting motor vibration signals under the condition that the miniature direct-current gear reduction motor to be detected idles; and the PC is used for judging the advantages and disadvantages of the miniature direct-current gear reduction motor to be detected according to the digital signals. The invention solves the problems of huge labor cost, fatigue detection and the like caused by the manual detection method widely used in the motor quality inspection field at present, and improves the quality identification efficiency of the miniature direct-current gear reduction motor while ensuring the accuracy.
Description
Technical Field
The invention relates to a mechanical fault diagnosis technology, in particular to a miniature direct-current gear reduction motor quality control device and method based on a convolutional network depth model.
Background
With the increasing development of human science and technology, automation is one of the main melodies of the 21 st century production development. As an electric motor capable of converting electric energy into mechanical energy, it has been an essential component of an automation system in various fields. Among them, in the occasion of low speed, high torque, gear motor is the most economical, practical preferred scheme all the time.
The gear reduction motor is a motor in which a gear reduction box is attached to an output shaft of a motor, and the output rotation speed is reduced from a high speed to a low speed by reduction of a gear, and the output torque is increased. Due to the characteristics, the miniature motor product is widely applied to precision instruments such as automatic production lines, medical equipment and the like and intelligent power output of related equipment such as intelligent industry, intelligent agriculture, intelligent home, intelligent robots and the like. Under the strong market competition pressure of the products, how to ensure the quality of the products while producing the products on a large scale becomes one of the important problems of whether gear reduction motor enterprises can create considerable economic benefits.
At present, quality detection of the miniature direct-current gear reduction motor is carried out, and besides rotating speed, torque, temperature rise and the like of a plurality of hard indexes, noise and gear quality are identified. In the domestic miniature gear reduction motor factory, the detection is generally carried out manually, namely, the product quality is comprehensively judged by sensing the no-load vibration of the motor through both hands and listening to the no-load noise of the motor through ears. The lagging low-efficiency method not only greatly increases labor cost in production, but also causes fatigue judgment errors of workers due to repeated labor, thereby leading inferior products to be added into the market and causing irrecoverable losses to the reputation and subsequent economy of enterprises.
Disclosure of Invention
Aiming at the technical problems, the invention aims to apply the deep learning technology to the quality detection of the miniature direct-current gear reduction motor, and the deep learning method can remarkably improve the precision and efficiency of motor quality inspection and reduce the manpower cost and precision efficiency of enterprises.
The invention is realized by adopting the following technical scheme:
A gear reduction motor quality control device based on a convolutional network depth model, comprising:
the direct-current stabilized power supply is used for supplying power to the miniature direct-current gear reduction motor to be detected;
The constant current adapter is used for supplying power to the acceleration sensor and amplifying motor vibration signals acquired by the acceleration sensor and then transmitting the motor vibration signals to the data acquisition device;
The data collector is used for generating the amplified motor vibration signal into a digital signal and transmitting the digital signal to the PC;
The acceleration sensor is used for collecting motor vibration signals under the condition that the miniature direct-current gear reduction motor to be detected idles;
And the PC is used for judging the advantages and disadvantages of the miniature direct-current gear reduction motor to be detected according to the digital signals.
Further, the micro direct current gear reduction motor to be detected is in rigid contact with the surface of the acceleration sensor through a rectangular metal steel sheet with the thickness of 0.1-0.25 mm.
Furthermore, the measuring surface of the acceleration sensor is rigidly connected with the metal steel sheet through metal glue.
Further, the direct-current stabilized power supply adopts a numerical control type linear direct-current stabilized power supply; the constant current adapter adopts a single-channel constant current adapter; the data acquisition device adopts a usb multifunctional data acquisition card.
A gear reduction motor quality control method based on a convolutional network depth model adopts the quality control device, and the method comprises the following steps:
1) Extracting a vibration signal of a miniature direct-current gear reduction motor to be detected, converting the vibration signal into a digital voltage signal, and transmitting the digital voltage signal into a PC;
2) The PC acquires a vibration signal of the motor, and preprocesses the vibration signal to acquire a three-channel time-frequency diagram corresponding to the vibration signal;
3) And inputting the three-channel time-frequency diagram into a selected convolutional network depth classification model trained by the corresponding motor model, and identifying the classification information of the motor quality, thereby reducing the burden of artificial quality inspection.
Further, the preprocessing to obtain the three-channel time-frequency diagram corresponding to the vibration signal specifically includes: and obtaining a five-second motor time domain signal, and then carrying out windowing, framing, fast Fourier transformation and image standardization processing on the signal to obtain a three-channel signal five-second time-frequency diagram.
Further, the training of the convolutional network depth classification model specifically includes:
1) Establishing a signal database of the miniature direct-current gear reduction motor, namely collecting hundreds of motors of a specific model, processing to obtain five-second time-frequency data signals of each motor, and dividing a training set, a test set and a verification set according to a certain proportion;
2) Windowing and framing the five-second time-frequency data signal, performing fast Fourier transform and image standardization processing to obtain a three-channel signal five-second time-frequency diagram;
3) And building a convolutional network deep classification model through a tensorflow deep learning library, initializing model super parameters including learning rate, batch size, CNN kernel size and number, network frame position and number, adding batch normalization processing in a convolutional layer and a full-connection layer, adding a dropout layer after the full-connection layer, performing L2 regularization on weights, and putting the model super parameters into LOSS calculation. The model has the characteristics of low error rate of a test set and capability of detecting the quality of the motor in real time.
4) Inputting the data of the training set into an initialized convolutional network deep model for training, and adjusting parameters such as weight of the convolutional network deep model through a BP algorithm when twenty periods or when the relative change of LOSS value of the network is smaller than a threshold value;
5) Inputting the data of the verification set into the trained model in the step six to verify the accuracy, if the accuracy is larger than the threshold, maintaining the network parameters of the model, otherwise, returning to the step 3), and adjusting the network frame and the super parameters of the model.
Further, the first layer of the convolutional network deep classification model adopts a one-dimensional convolutional network with a convolutional kernel of 7*1, and then uses one layer maxpool, two layers of two-dimensional convolutional networks and two layers of full-connection layer networks to classify by softmax.
Further, the model hyper-parameters are specifically:
the input features are a three-channel time-frequency diagram of 3 x 129 x 92, the batchsize of the network is 40, the learning rate is 1e-4 and then 1e-5 is changed in the 20 th period.
Further, the classification information for identifying the motor quality specifically includes: the five-second motor signals are divided and respectively judged, and if the number of the excellent results is large, the motor is judged to be an excellent motor; if the ratio of the obtained softmax results exceeds a certain threshold, the motor is repositioned and measured.
Compared with the prior art, the invention can accurately extract the vibration signal of the miniature direct-current gear reduction motor, judge the quality of the miniature direct-current gear reduction motor by combining with the established depth model library based on the convolution network, improve the quality identification efficiency of the miniature direct-current gear reduction motor, solve the problems of huge labor cost, fatigue detection and the like caused by the manual detection method widely used in the field at present, further reduce the workload of manual inspection and improve the motor detection precision, thereby improving the motor production efficiency.
Drawings
Fig. 1 is a schematic diagram of a quality control system of a gear reduction motor based on a convolution network.
Fig. 2 is a flowchart of the identification of the quality control system of the gear reduction motor.
Fig. 3 is a model building training flow chart of a miniature direct current gear reduction motor based on a convolutional deep network.
FIG. 4 is a framework and partial parameters of a convolutional network deep model.
In the figure: 1-a direct current stabilized power supply; 2-a miniature direct-current gear reduction motor to be detected; 3-metal steel sheet; a 4-acceleration sensor; 5-constant current adapter; 6-a data collector; 7-PC.
Detailed Description
The invention will be further described with reference to the accompanying drawings and specific examples.
As shown in fig. 1, a gear reduction motor quality control device based on a convolutional network depth model includes:
the direct-current stabilized power supply 1 is used for supplying power to the micro direct-current gear reduction motor 2 to be detected, and MOTECH numerical control type linear direct-current stabilized power supply LPS-305 is adopted in the embodiment;
the constant current adapter 5 is used for supplying power to the acceleration sensor 4, amplifying motor vibration signals acquired by the acceleration sensor 4 and then transmitting the motor vibration signals to the data acquisition unit 6, and the embodiment adopts a CT5201 single-channel constant current adapter;
The data collector 6 is configured to generate a digital signal from the amplified motor vibration signal, and send the digital signal to the PC 7, where the MCC1608G usb multifunctional data collection card DAQ is adopted in the embodiment; after the amplified motor vibration signal is transmitted to an analog input port of the MCC1608G usb multifunctional data acquisition card DAQ, the data acquisition device can convert the analog signal into a digital signal and transmit the digital signal to the PC 7;
The acceleration sensor 3 is used for collecting motor vibration signals under the condition that the micro direct-current gear reduction motor 2 to be detected idles, the CT1050LC acceleration sensor is adopted in the embodiment, and the micro direct-current gear reduction motor to be detected is in rigid contact with the surface of the acceleration sensor 3 through metal glue by a rectangular stainless steel sheet with the thickness of 0.25 mm;
and the PC 7 is used for judging the advantages and disadvantages of the micro direct-current gear reduction motor 2 to be detected according to the digital signals.
As shown in fig. 2, a gear reduction motor quality inspection method based on a convolutional network depth model adopts the quality inspection device, and the method comprises the following steps:
1) Extracting a five-second motor vibration signal of a miniature direct-current gear reduction motor to be detected, dividing the five-second motor vibration signal into five parts of one-second motor vibration signals, converting the five parts of one-second motor vibration signals into digital voltage signals, and transmitting the digital voltage signals into a PC;
2) And the PC acquires the vibration signal of the motor, and then performs windowing, framing, fast Fourier transformation and image standardization processing on five parts of one-second motor vibration signals to obtain a three-channel signal five-second time-frequency diagram.
3) Inputting the three-channel time-frequency diagram into a selected convolutional network depth classification model which is trained by the corresponding motor model to obtain five softmax results, and if the ratio of a certain softmax is smaller than a certain threshold value, indicating that the model cannot accurately separate the motor quality, and returning to the first step for re-measurement; and then judging five softmax results, if most results belong to a certain class, judging the quality of the motor as the class, and identifying the classification information of the motor quality, thereby reducing the burden of artificial quality inspection.
FIG. 3 is a training flow chart of the convolutional network depth classification model, comprising the steps of:
step one: and collecting 100 motors of good and bad quality respectively according to a manual judgment method of a factory technician, and marking.
Step two: and collecting 5s motor vibration signals for each sample motor through the quality inspection device, wherein 20 groups of motors are respectively arranged. The training set, the test set and the verification set are divided according to the ratio of 6:2:2, and the 5s motor signals are divided into 5 parts of 1s motor signals in sequence.
Step three: because the motor signal belongs to a non-stationary signal, the windowing and framing process is needed to perform Fourier transform, wherein a Hamming window is selected as a window function:
in order to offset the two-end weakening caused by windowing, the frames need to be overlapped when framing, the frame length is 256 sampling points, and the frame is moved to 128 sampling points.
Step four: and performing fast Fourier transform on the windowed and framed motor signals to obtain a corresponding three-channel motor time-frequency diagram. Then, carrying out image 2 standardization processing on each picture:
where x represents the image matrix, μ is the same size matrix, and the elements are the mean of the image. Σ represents the standard deviation and N represents the number of pixel points of the image x. The image normalization is the process of centering the data through the de-mean value, so that the model can obtain the generalization effect more easily.
Step five: the deep model of the convolutional network is built through tensorflow deep learning libraries, the trained model framework is shown in figure 3, and model super parameters including learning rate, batch size, CNN kernel size and number, network framework position and number and the like are initialized at present. And then adding batch normalization processing in the convolution layer and the full connection layer, adding a dropout layer after the full connection layer, performing L2 regularization on the weight value, and putting the weight value into final LOSS calculation. This is in part to increase the learning speed of the model and the generalization ability of the model.
Step six: inputting the data of the training set into the initialization model in the fifth step for training, and adjusting parameters such as weight of the deep model of the convolutional network through a BP algorithm when twenty periods or when the relative change of LOSS value of the network is smaller than a certain threshold value.
Step seven: inputting the data of the verification set into the trained model in the step six to verify the accuracy, if the accuracy is larger than the threshold, maintaining the network parameters of the model, otherwise, returning to the step five again, and adjusting the network frame and the super parameters of the model.
As shown in fig. 4, the first layer of the deep classification model of the convolutional network in this embodiment adopts a one-dimensional convolutional network with a convolutional kernel 7*1, and then uses one-layer maxpool, two-dimensional convolutional networks and two-layer fully-connected layer networks to classify by softmax. The input features are a three-channel time-frequency diagram of 3 x 129 x 92, batchsize of the network is 40, the learning rate is 1e-4, and then 1e-5 is changed in the 20 th period.
In summary, the embodiments of the present invention mainly include a hardware portion and a software portion, where:
Hardware part: the device mainly comprises an acceleration sensor, a direct-current stabilized power supply, a constant-current adapter, a data acquisition unit and a PC. The miniature direct-current gear reduction motor is powered by a direct-current stabilized power supply and is in rigid contact with the surface of the acceleration sensor under no-load condition; the invention can accurately extract the vibration signal of the miniature DC gear reduction motor by using the algorithm designed by the invention, and judge the advantages and disadvantages of the miniature DC gear reduction motor by combining the established depth model library based on the convolution network, thereby reducing the workload of manual qualification and improving the motor detection precision, and further improving the motor production efficiency.
Software part: mainly relates to the following contents:
1. The frequency domain signal of the vibration signal is that the window function is used for windowing and framing the signal, and the fast Fourier transform is carried out on each frame signal to generate a five-second signal time-frequency diagram corresponding to three channels;
2. The standardization of images, namely, centralizing each image through a mean value, and according to the convex optimization theory and the data probability distribution related knowledge, the data centralization accords with the data distribution rule, so that the generalization effect after training is easier to obtain;
3. and constructing and learning a convolutional network, namely setting super parameters of the convolutional network, and positions and numbers of maxpool and full-connection layers according to the deep convolutional neural classification network with proper acquired signal characteristic design parameters. The generalization capability of the model is ensured while the precision is satisfied;
4. Optimization of convolutional network: adding batch standardization, L2 regularization, dropout and other technologies to improve the learning rate and generalization capability of the model;
5. identification of telecommunication signals: the recognition rate of the network is improved by dividing the five-second motor signal and respectively recognizing, and the reliability degree of classification is measured by judging the ratio of softmax.
The embodiment of the invention aims at the problems of quality detection efficiency, precision and the like of the miniature direct-current gear reduction motor, and integrates a deep learning method into the quality detection efficiency, the precision and the like of the miniature direct-current gear reduction motor. The quality detection task of the miniature direct-current gear reduction motor is realized by carrying out feature extraction and model learning on the vibration signals of the gear reduction motor through the improved convolutional neural network. The quality identification efficiency of the miniature direct-current gear reduction motor is improved while the accuracy is ensured.
The foregoing is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto, and any person skilled in the art can substitute or change the technical solution and the inventive conception of the present invention equally within the scope of the present invention disclosed by the present invention.
Claims (2)
1. Gear reducer motor quality control device based on convolutional network degree of depth model, characterized by comprising:
the direct-current stabilized power supply is used for supplying power to the miniature direct-current gear reduction motor to be detected;
The constant current adapter is used for supplying power to the acceleration sensor and amplifying motor vibration signals acquired by the acceleration sensor and then transmitting the motor vibration signals to the data acquisition device;
the data acquisition device is used for generating digital signals from the amplified motor vibration signals and transmitting the digital signals to the PC;
The acceleration sensor is used for collecting motor vibration signals under the condition that the miniature direct-current gear reduction motor to be detected idles;
The PC is used for judging the advantages and disadvantages of the miniature direct-current gear reduction motor to be detected according to the digital signals; the micro direct-current gear reduction motor to be detected is in rigid contact with the surface of the acceleration sensor through a rectangular metal steel sheet with the thickness of 0.1-0.25 mm; the measuring surface of the acceleration sensor is rigidly connected with the metal steel sheet through metal glue; the direct-current stabilized power supply adopts a numerical control type linear direct-current stabilized power supply; the constant current adapter adopts a single-channel constant current adapter; the data acquisition device adopts a usb multifunctional data acquisition card;
The detection method of the quality control device comprises the following steps:
1) Extracting a vibration signal of a miniature direct-current gear reduction motor to be detected, converting the vibration signal into a digital voltage signal, and transmitting the digital voltage signal into a PC;
2) The PC acquires a vibration signal of the motor, and preprocesses the vibration signal to acquire a three-channel time-frequency diagram corresponding to the vibration signal;
3) Inputting the three-channel time-frequency diagram into a selected convolutional network depth classification model trained by the corresponding motor model, and identifying classification information of motor quality; the three-channel time-frequency diagram corresponding to the vibration signal obtained by preprocessing specifically comprises the following steps: obtaining a five-second motor time domain signal, and then carrying out windowing, framing, fast Fourier transformation and image standardization processing on the signal to obtain a three-channel signal five-second time-frequency diagram;
the training of the convolutional network depth classification model specifically comprises the following steps:
1) Establishing a signal database of the miniature direct-current gear reduction motor, namely collecting hundreds of motors of a specific model, processing to obtain five-second time-frequency data signals of each motor, and dividing a training set, a test set and a verification set according to a certain proportion;
2) Windowing and framing the five-second time-frequency data signal, performing fast Fourier transform and image standardization processing to obtain a three-channel signal five-second time-frequency diagram;
3) Setting up a convolutional network depth classification model through a tensorflow depth learning library, initializing model super parameters including learning rate, batch size, CNN kernel size and number, network frame position and number, adding batch normalization processing in a convolutional layer and a full-connection layer, adding a dropout layer after the full-connection layer, performing L2 regularization on weights, and putting the model super parameters into LOSS calculation;
4) Inputting the data of the training set into an initialized convolutional network deep model for training, and adjusting weight parameters of the convolutional network deep model through a BP algorithm when twenty periods or the relative change of LOSS values of the network is smaller than a threshold value;
5) Inputting the data of the verification set into the trained model in the step 4), if the accuracy is larger than the threshold, maintaining the network parameters of the model, otherwise, returning to the step 3), and adjusting the network frame and the super parameters of the model; the classification information for identifying the motor quality specifically comprises: the five-second motor signals are divided and respectively judged, and if the number of the excellent results is large, the motor is judged to be an excellent motor; if the ratio of the obtained softmax results exceeds a certain threshold, repositioning the motor and measuring; the first layer of the convolutional network deep classification model adopts a one-dimensional convolutional network with a convolutional kernel of 7*1, and is classified by softmax by using one layer maxpool, two layers of two-dimensional convolutional networks and two layers of full-connection layer networks.
2. The gear reduction motor quality control device based on the convolutional network depth model of claim 1, wherein the model hyper-parameters are specifically:
the input features are a three-channel time-frequency diagram of 3 x 129 x 92, the batchsize of the network is 40, the learning rate is 1e-4 and then 1e-5 is changed in the 20 th period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811633174.6A CN109782167B (en) | 2018-12-29 | 2018-12-29 | Gear reduction motor quality inspection device and method based on convolutional network depth model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811633174.6A CN109782167B (en) | 2018-12-29 | 2018-12-29 | Gear reduction motor quality inspection device and method based on convolutional network depth model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109782167A CN109782167A (en) | 2019-05-21 |
| CN109782167B true CN109782167B (en) | 2024-05-10 |
Family
ID=66498890
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811633174.6A Active CN109782167B (en) | 2018-12-29 | 2018-12-29 | Gear reduction motor quality inspection device and method based on convolutional network depth model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109782167B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110109015B (en) * | 2019-05-31 | 2020-06-26 | 中南大学 | Asynchronous motor fault monitoring and diagnosing method based on deep learning |
| CN110705456A (en) * | 2019-09-29 | 2020-01-17 | 华南理工大学 | Micro motor abnormity detection method based on transfer learning |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107421741A (en) * | 2017-08-25 | 2017-12-01 | 南京信息工程大学 | A kind of Fault Diagnosis of Roller Bearings based on convolutional neural networks |
| CN209280134U (en) * | 2018-12-29 | 2019-08-20 | 华南理工大学 | A kind of minisize dc toothed gearing electric motor product examine device based on acceleration transducer |
-
2018
- 2018-12-29 CN CN201811633174.6A patent/CN109782167B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107421741A (en) * | 2017-08-25 | 2017-12-01 | 南京信息工程大学 | A kind of Fault Diagnosis of Roller Bearings based on convolutional neural networks |
| CN209280134U (en) * | 2018-12-29 | 2019-08-20 | 华南理工大学 | A kind of minisize dc toothed gearing electric motor product examine device based on acceleration transducer |
Non-Patent Citations (2)
| Title |
|---|
| Real-Time Motor Fault Detection by 1-D Convolutional Neural Networks;Turker Ince 等;《IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS》;第第63卷卷(第第11期期);第7067-7075页 * |
| 小型电机生产过程在线品质检测方法研究;罗炳聪等;《测控技术》;第第34卷卷(第第12期期);第34-37 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109782167A (en) | 2019-05-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | A novel convolutional neural network based fault recognition method via image fusion of multi-vibration-signals | |
| CN109555566B (en) | Steam turbine rotor fault diagnosis method based on LSTM | |
| Kong et al. | Attention recurrent autoencoder hybrid model for early fault diagnosis of rotating machinery | |
| CN114235415B (en) | Wind driven generator variable pitch bearing fault diagnosis method and device based on neural network | |
| CN109782167B (en) | Gear reduction motor quality inspection device and method based on convolutional network depth model | |
| CN108435819B (en) | Energy consumption abnormity detection method for aluminum profile extruder | |
| CN111950396B (en) | Meter reading neural network identification method | |
| CN114818774A (en) | Intelligent gearbox fault diagnosis method based on multi-channel self-calibration convolutional neural network | |
| CN116034904B (en) | Pig health monitoring system and method based on track type inspection robot | |
| CN110427943A (en) | A kind of intelligent electric meter technique for partitioning based on R-CNN | |
| CN111402249B (en) | Image evolution analysis method based on deep learning | |
| CN117009916A (en) | Actuator fault diagnosis method based on multi-sensor information fusion and transfer learning | |
| CN115822943A (en) | Hydraulic pump leakage fault diagnosis method | |
| CN114714145A (en) | Method for enhancing, comparing, learning and monitoring tool wear state by using Gelam angular field | |
| Xue et al. | A novel framework for motor bearing fault diagnosis based on multi-transformation domain and multi-source data | |
| CN114626625A (en) | Bearing life prediction method based on multi-module U-BiLSTM network | |
| CN113705887A (en) | Data-driven photovoltaic power generation power prediction method and system | |
| CN108734692A (en) | A kind of bridge section model system | |
| CN109975702B (en) | Direct-current gear reduction motor quality inspection method based on circulation network classification model | |
| CN101825621A (en) | Device for detecting activity of stored grain and method thereof | |
| CN116522237A (en) | Rolling bearing fault diagnosis method based on parallel convolution neural network | |
| CN110852189A (en) | Low-complexity dense crowd analysis method based on deep learning | |
| CN115239952A (en) | Method for reading industrial table based on YOLOV5 and U2-Net framework assistance | |
| CN112257773B (en) | Mechanical equipment fault diagnosis method based on multiple measurement vectors of wireless sensor network | |
| CN114548295A (en) | Bearing fault classification system and method based on multi-scale domain adaptive network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |