CN114873425A - Escalator drive chain fault diagnosis method based on vibration characteristic enhancement - Google Patents
Escalator drive chain fault diagnosis method based on vibration characteristic enhancement Download PDFInfo
- Publication number
- CN114873425A CN114873425A CN202210562832.7A CN202210562832A CN114873425A CN 114873425 A CN114873425 A CN 114873425A CN 202210562832 A CN202210562832 A CN 202210562832A CN 114873425 A CN114873425 A CN 114873425A
- Authority
- CN
- China
- Prior art keywords
- fault
- feature
- vibration
- drive chain
- amplified
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B29/00—Safety devices of escalators or moving walkways
- B66B29/005—Applications of security monitors
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Algebra (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a fault diagnosis method for an escalator drive chain based on vibration characteristic enhancement. The method comprises the following steps: step 1: and collecting vibration signals of the driving chain. Step 2: enhancing the vibration signal of the driving chain by using a vibration feature enhancement method to obtain an amplified fault feature, and inputting the amplified fault feature into a fault classification model for judgment; and step 3: and when the judgment result of the fault classification model is that a fault influencing the operation of the drive chain is met, early warning is carried out. According to the invention, the amplified fault characteristics are obtained after the vibration signals are subjected to characteristic enhancement, the escalator drive chain can be subjected to timely fault diagnosis according to the amplified fault characteristics, the escalator can be maintained in time, and the maintenance cost is reduced.
Description
Technical Field
The invention belongs to a method for detecting faults of an escalator in the field of escalators, and particularly relates to a method for diagnosing faults of an escalator drive chain based on vibration characteristic enhancement.
Background
The risk judgment is carried out on the driving device of the escalator according to the principle and the procedure about the risk evaluation of the elevator in GB/T20900 2007 method for evaluating and reducing the risk of the elevator, the escalator and the moving sidewalk. The driving chain wheel drives the driving chain to operate, so that the step chain wheel is driven to control the operation of the step chain and the steps, and therefore the driving chain can be considered as a core component in the driving device.
The escalator is one of the most common devices in public places and is closely related to the traveling safety of people, and the safety accidents are mostly caused by the fatigue failure, breakage and the like of the parts of the escalator. In the statistics of escalator accidents, accidents caused by the failure of escalator components account for about 51%, and the reason for the failure is mainly due to the failure of the drive chain, the step chain and the steps. Once the life threatening events of the people caused by the failure of the parts of the escalator happen, the life threatening events inevitably cause great loss. The driving chain, which is the most important component of the escalator component, is often difficult to distinguish against the operating conditions during operation, so that accidents are difficult to prevent, and because of large noise in practical application, fault vibration caused by defects is often submerged in noise, so that faults are difficult to judge. Therefore, there is a need for a method that enhances the defective failure characteristics, which is particularly important for drive chain vibration condition detection and failure diagnosis.
Disclosure of Invention
The invention aims to solve the problems mentioned above, and provides a fault diagnosis method for an escalator drive chain based on vibration characteristic enhancement aiming at fault analysis of the drive chain in the escalator, wherein different fault types such as chain wheel abrasion, chain roller abrasion and the like of the drive chain can occur in the operation process of the escalator, so that different signals are generated, the different fault type signals transmitted by an acceleration sensor are processed, the characteristics of the different fault type signals are extracted, and the fault types are judged through a neural network, so that fault information is reflected to an early warning system to timely carry out fault early warning on a drive device, the occurrence of accidents is prevented in advance, and the loss is avoided.
The invention is realized by the following technical scheme:
the invention comprises the following steps:
step 1: and collecting vibration signals of the driving chain.
Step 2: enhancing the vibration signal of the driving chain by using a vibration feature enhancement method to obtain an amplified fault feature, and inputting the amplified fault feature into a fault classification model for judgment;
and step 3: when the judgment result of the fault classification model is normal, the fault classification model is not processed; otherwise, early warning is carried out according to the judgment result.
In the step 2, a vibration characteristic enhancement method is used for enhancing the vibration signal of the driving chain to obtain an amplification fault characteristic, and the method specifically comprises the following steps:
s1: carrying out modal decomposition on the vibration signal by using a variational decomposition method to obtain a plurality of corresponding modes, then removing a first mode from the plurality of modes and reconstructing the rest modes to obtain a reconstructed vibration signal;
s2: s transformation is carried out on the reconstructed vibration signal to obtain a feature matrix, time, frequency domain and time-frequency domain features are extracted from the feature matrix to serve as a plurality of features to be screened, the accumulated contribution rate of each feature to be selected is calculated according to principal component analysis, and the features to be screened with the accumulated contribution rate larger than a preset threshold value are screened to obtain the features to be screened and recorded as the features to be spliced;
s3: and after one-dimensional splicing is carried out on each feature to be spliced, the spliced feature is obtained, then the two-dimensional conversion is carried out on the spliced feature by utilizing the gram angular field, a feature diagram is obtained, and the feature diagram is used as an amplified fault feature.
And the fault classification model in the step 2 is a convolutional neural network.
The activation function of the activation layer in the convolutional neural network is ReLU.
A pooling layer in the convolutional neural network takes a maximum pooling operation.
Compared with the prior art, the invention has the following beneficial effects:
the driving device applied to the escalator collects vibration signals through the acceleration sensor on the driving chain wheel, the vibration signals are processed in the industrial personal computer by the method, so that different fault categories are judged, fault classification is difficult to perform due to the fact that fault characteristics can be submerged in noise, and fault characteristics are enhanced, so that the fault classification model is high in input learning training efficiency, high in fault identification accuracy and high in accuracy. And then the judgment result is transmitted into a fault early warning system for signal early warning, so that the escalator is timely maintained by fault diagnosis and warning of a driving device of the escalator, the maintenance cost is reduced, and the escalator accident is avoided.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
In the figure: 1-chain wheel, 2-driving chain, 4-acceleration sensor and 5-industrial personal computer.
Detailed Description
The invention is described in further detail below with reference to the following figures and specific embodiments:
as shown in fig. 1, the present invention comprises the steps of:
step 1: in the concrete implementation, the acceleration sensor 4 is installed at one side of the chain wheel 1 and used for capturing the vibration signal of the driving chain 2 when the escalator runs, so as to collect the vibration signal of the driving chain.
Step 2: the acceleration sensor 4 is connected with an interface of an industrial personal computer 5, a vibration signal is input into the industrial personal computer 5, the vibration signal of a driving chain output by the industrial personal computer 5 is enhanced by a vibration feature enhancement method, an amplified fault feature is obtained, and the amplified fault feature is one-dimensional. Inputting the amplified fault characteristics into a fault classification model for judgment, namely judging the corresponding fault category;
in the step 2, a vibration characteristic enhancement method is used for enhancing a vibration signal of a driving chain output by the industrial personal computer 5 to obtain an amplified fault characteristic, and the method specifically comprises the following steps:
s1: carrying out modal decomposition on the vibration signal by using a variational decomposition method to obtain a plurality of corresponding modes, then removing a first mode from the plurality of modes and reconstructing the rest modes to obtain a reconstructed vibration signal;
s2: s transformation is carried out on the reconstructed vibration signal to obtain a feature matrix, time, frequency domain and time-frequency domain features are extracted from the feature matrix to serve as a plurality of features to be screened, the accumulated contribution rate of each feature to be selected is calculated according to principal component analysis, and the features to be screened with the accumulated contribution rate larger than a preset threshold value are screened to obtain the features to be screened and recorded as the features to be spliced;
s3: and after one-dimensional splicing is carried out on each feature to be spliced, the spliced feature is obtained, then the two-dimensional conversion is carried out on the spliced feature by utilizing the gram angular field, a feature diagram is obtained, and the feature diagram is used as an amplified fault feature.
More specifically, S1: signals collected by the acceleration sensor are transmitted into the industrial personal computer, and fault vibration signals are submerged in noise to cause unobvious fault characteristics, so that the signal reconstruction module is firstly utilized for denoising. Assuming that the signal is decomposed into a plurality of modes,to obtain a denoised signal, the normalization method by L2:
wherein u is k The secondary mode is a sub-mode, and alpha is a secondary penalty factor of L2;
since the L2 norm is equidistant in the time-frequency domain, (1) is converted to the frequency domain:
and by developing (2) into functional pairs u k Solving for
Adding a center frequency at the center of a sub-mode so that it is in the time domainAfter Fourier transform conversion to frequency domainThereby expanding into functional
To find the center frequency, an alternate direction update method is used, since u k The modal update formula can be obtained in the same way as (4), fixed u k Thereby centering the center frequency omega k Direction searching
To obtain omega k Is updated to
The final update formula is obtained as:
control two variables fixed at a time to update the other value, toAs a determination condition for the iteration stop, epsilon is a sufficiently small value, and if it cannot be stopped, updating is performed again. Since each component is different in the center of frequency, the number of decompositions is selected by first selecting a range of 2-15, and decomposing the signal in the range of 2 to 15, and when the number of decompositions is such that the center frequency of a component of a certain value is close, the signal is considered to be over-decomposed, and then the number of decompositions-1 at that time is selected as the desired number of decompositions. Since the decomposed first component contains a large amount of high-frequency noise, the first modal component is rejected and the remaining modal components are added to reconstruct the signal.
S2: because the drive chain is connected with the chain wheel, the mutual influence of vibration signals is caused by the mutual connection between the structures, the fault characteristics are not obvious, in order to further amplify the signal characteristics, the signal characteristics are extracted by the characteristic extraction module, because the time domain of the signal has a large amount of information reflecting the signal characteristics, firstly, a Gaussian window function is added in the Fourier transform of the time sequence data after passing through the signal reconstruction module to obtain:
So at this time the transformation formula becomes:
wherein τ is the time shift of the gaussian window, σ represents the width of the gaussian function, a feature matrix is obtained after the transformation, the columns of the matrix represent the sampling time, the rows represent the frequency values, and the amplitude and phase information can be extracted from the feature matrix. And extracting p characteristics of the variance, the kurtosis, the skewness, the root mean square, the peak-to-peak value, the standard deviation, the mean square error, the root mean square error, the maximum frequency, the minimum frequency, the average value and the characteristic value of the transformed matrix of the signals from the characteristic matrix as new characteristics of the new representative signals. And combining all the extracted signal features into a feature space. F ═ F ij ] n×p Where n is a sample point and p is a feature number, normalizing the feature data to eliminate the influence of the feature dimension, and then calculating a correlation coefficient matrix R ═ R ij ] p×p Wherein
Calculating the eigenvalues λ of the matrix R i And calculating the accumulated contribution rate of the characteristic value through the characteristic vector, selecting the characteristics with the accumulated contribution rate of the first 85 percent, and reconstructing the signal characteristics by using the characteristic vector. Wherein the cumulative contribution rate is calculated as
S3: converting the one-dimensional vector subjected to feature screening into a two-dimensional matrix, wherein the conversion formula is as follows:
Thereby converting the one-dimensional vector into a two-dimensional space so that the fault characteristics of the vibration signal are enhanced and output in the form of a two-dimensional picture.
In specific implementation, the same operation is carried out on the faults of different types, so that fault pictures of different types are formed and input into the fault classification model for feature learning, and the faults are classified. The fault classification model in step 2 is a convolutional neural network. As the fault characteristics are enhanced, experiments show that the simple convolutional neural network structure can achieve the effects of learning the fault characteristics and better classifying. Where the convolutional neural network can be expressed as:where k is the k-th tier network,for the output of the k-th one,is the kth input, W i j Is a weight matrix of the convolution kernel,for biasing, f is the activation function. Through convolution and pooling operations, the classification probability output is:
the activation function of the activation layer in the convolutional neural network is ReLU, and the pooling layer takes maximum pooling operation.
And step 3: when the judgment result of the fault classification model is normal, the fault classification model is not processed; otherwise, early warning is carried out according to the judgment result. Specifically, a signal is sent to the fault early warning system according to the judgment result, and the fault early warning system carries out early warning corresponding to the judgment result, so that the effect of preventing danger in advance is achieved.
The above embodiments are not limited to the technical solutions of the embodiments themselves, and the embodiments may be combined with each other into a new embodiment. The above embodiments are only for illustrating the technical solutions of the present invention and are not limited thereto, and any modification or equivalent replacement without departing from the spirit and scope of the present invention should be covered within the technical solutions of the present invention.
Claims (5)
1. A fault diagnosis method for an escalator drive chain based on vibration characteristic enhancement is characterized by comprising the following steps:
step 1: and collecting vibration signals of the driving chain.
Step 2: enhancing the vibration signal of the driving chain by using a vibration feature enhancement method to obtain an amplified fault feature, and inputting the amplified fault feature into a fault classification model for judgment;
and step 3: when the judgment result of the fault classification model is normal, the fault classification model is not processed; otherwise, early warning is carried out according to the judgment result.
2. The escalator drive chain fault diagnosis method based on vibration characteristic enhancement according to claim 1, wherein in step 2, a vibration signal of the drive chain is enhanced by using a vibration characteristic enhancement method to obtain an amplified fault characteristic, specifically:
s1: carrying out modal decomposition on the vibration signal by using a variational decomposition method to obtain a plurality of corresponding modes, then removing a first mode from the plurality of modes and reconstructing the rest modes to obtain a reconstructed vibration signal;
s2: s transformation is carried out on the reconstructed vibration signal to obtain a feature matrix, time, frequency domain and time-frequency domain features are extracted from the feature matrix to serve as a plurality of features to be screened, the accumulated contribution rate of each feature to be selected is calculated according to principal component analysis, and the features to be screened with the accumulated contribution rate larger than a preset threshold value are screened to obtain the features to be screened and recorded as the features to be spliced;
s3: and after one-dimensional splicing is carried out on each feature to be spliced, the spliced feature is obtained, then the two-dimensional conversion is carried out on the spliced feature by utilizing the gram angular field, a feature diagram is obtained, and the feature diagram is used as an amplified fault feature.
3. The escalator drive chain fault diagnosis method based on vibration characteristic enhancement according to claim 1, wherein the fault classification model in step 2 is a convolutional neural network.
4. The escalator drive chain fault diagnosis method based on vibration characteristic enhancement according to claim 3, characterized in that the activation function of the activation layer in the convolutional neural network is ReLU.
5. The escalator drive chain fault diagnosis method based on vibration characteristic enhancement according to claim 3, characterized in that the pooling layer in the convolutional neural network takes maximum pooling operation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210562832.7A CN114873425B (en) | 2022-05-23 | 2022-05-23 | Escalator driving chain fault diagnosis method based on vibration characteristic enhancement |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210562832.7A CN114873425B (en) | 2022-05-23 | 2022-05-23 | Escalator driving chain fault diagnosis method based on vibration characteristic enhancement |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114873425A true CN114873425A (en) | 2022-08-09 |
CN114873425B CN114873425B (en) | 2023-06-23 |
Family
ID=82677500
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210562832.7A Active CN114873425B (en) | 2022-05-23 | 2022-05-23 | Escalator driving chain fault diagnosis method based on vibration characteristic enhancement |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114873425B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117786560A (en) * | 2024-02-28 | 2024-03-29 | 通用电梯股份有限公司 | Elevator fault classification method based on multi-granularity cascade forest and electronic equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111325159A (en) * | 2020-02-25 | 2020-06-23 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Fault diagnosis method and device, computer equipment and storage medium |
CN111751133A (en) * | 2020-06-08 | 2020-10-09 | 南京航空航天大学 | Intelligent fault diagnosis method of deep convolutional neural network model based on non-local mean embedding |
CN112633132A (en) * | 2020-12-18 | 2021-04-09 | 合肥工业大学 | Bearing fault diagnosis method and system based on two-dimensional vibration image enhancement |
CN114044431A (en) * | 2021-10-08 | 2022-02-15 | 上海三菱电梯有限公司 | Method and device for monitoring abnormality of step roller of passenger conveyor, and passenger conveyor |
CN114492518A (en) * | 2022-01-13 | 2022-05-13 | 北京科技大学 | Equipment fault diagnosis method and device based on multi-scale 2D expansion convolution network |
-
2022
- 2022-05-23 CN CN202210562832.7A patent/CN114873425B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111325159A (en) * | 2020-02-25 | 2020-06-23 | 中国电子产品可靠性与环境试验研究所((工业和信息化部电子第五研究所)(中国赛宝实验室)) | Fault diagnosis method and device, computer equipment and storage medium |
CN111751133A (en) * | 2020-06-08 | 2020-10-09 | 南京航空航天大学 | Intelligent fault diagnosis method of deep convolutional neural network model based on non-local mean embedding |
CN112633132A (en) * | 2020-12-18 | 2021-04-09 | 合肥工业大学 | Bearing fault diagnosis method and system based on two-dimensional vibration image enhancement |
CN114044431A (en) * | 2021-10-08 | 2022-02-15 | 上海三菱电梯有限公司 | Method and device for monitoring abnormality of step roller of passenger conveyor, and passenger conveyor |
CN114492518A (en) * | 2022-01-13 | 2022-05-13 | 北京科技大学 | Equipment fault diagnosis method and device based on multi-scale 2D expansion convolution network |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117786560A (en) * | 2024-02-28 | 2024-03-29 | 通用电梯股份有限公司 | Elevator fault classification method based on multi-granularity cascade forest and electronic equipment |
CN117786560B (en) * | 2024-02-28 | 2024-05-07 | 通用电梯股份有限公司 | Elevator fault classification method based on multi-granularity cascade forest and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN114873425B (en) | 2023-06-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Eftekhar Azam et al. | Damage detection in structural systems utilizing artificial neural networks and proper orthogonal decomposition | |
CN110555273B (en) | Bearing life prediction method based on hidden Markov model and transfer learning | |
US20190285517A1 (en) | Method for evaluating health status of mechanical equipment | |
Tibaduiza et al. | Damage classification in structural health monitoring using principal component analysis and self‐organizing maps | |
Kankar et al. | Rolling element bearing fault diagnosis using wavelet transform | |
Kwan et al. | A novel approach to fault diagnostics and prognostics | |
CN113642754B (en) | Complex industrial process fault prediction method based on RF noise reduction self-coding information reconstruction and time convolution network | |
CN113469060A (en) | Multi-sensor fusion convolution neural network aeroengine bearing fault diagnosis method | |
CN111170103B (en) | Equipment fault identification method | |
US11780452B2 (en) | Method and system for fault diagnoses of intelligent vehicles | |
CN114873425A (en) | Escalator drive chain fault diagnosis method based on vibration characteristic enhancement | |
CN116578940A (en) | Bearing fault new type identification and diagnosis method based on mixed depth self-coding | |
CN114913606A (en) | YOLO-based violation detection method for deep learning industrial field production work area | |
Shi et al. | A novel unsupervised real‐time damage detection method for structural health monitoring using machine learning | |
CN114660180A (en) | Sound emission and 1D CNNs-based light-weight health monitoring method and system for medium and small bridges | |
CN112418460A (en) | Fault diagnosis method and fault diagnosis device for engineering vehicle | |
Manghai et al. | Brake fault diagnosis through machine learning approaches–a review | |
CN111091554B (en) | Railway wagon swing bolster fracture fault image identification method | |
CN117315365A (en) | Camshaft surface damage detecting system based on visual analysis | |
CN114580472B (en) | Large-scale equipment fault prediction method with repeated cause and effect and attention in industrial internet | |
Li et al. | A robust fault diagnosis method for rolling bearings based on deep convolutional neural network | |
CN115630278A (en) | Network vibration damage detection method based on channel-space-time attention mechanism | |
CN112850408A (en) | Elevator emergency stop trapped person fault detection method based on multi-model fusion | |
CN118114186B (en) | Equipment vibration fault diagnosis method based on artificial intelligence | |
CN112578794B (en) | AGV fault detection method, storage medium and system based on machine learning |
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 |