CN109029937A - A kind of mechanical arm track method for monitoring abnormality based on data - Google Patents
A kind of mechanical arm track method for monitoring abnormality based on data Download PDFInfo
- Publication number
- CN109029937A CN109029937A CN201810511936.9A CN201810511936A CN109029937A CN 109029937 A CN109029937 A CN 109029937A CN 201810511936 A CN201810511936 A CN 201810511936A CN 109029937 A CN109029937 A CN 109029937A
- Authority
- CN
- China
- Prior art keywords
- mechanical arm
- real
- characteristic
- data
- abnormality
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing And Monitoring For Control Systems (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a kind of mechanical arm track method for monitoring abnormality based on data, malfunction monitoring is carried out to the characteristic signal under the real-time working condition of mechanical arm including the use of malfunction monitoring model, real-time characteristic related coefficient is obtained, judges the real-time working condition of mechanical arm for normal condition or abnormality using real-time characteristic related coefficient；The training of the malfunction monitoring model includes: the characteristic signal using six axis gyro sensor collection machinery arms under normal condition and abnormality, carries out data cleansing to characteristic signal, obtains training set；Using training set training neural network, malfunction monitoring model is obtained.The present invention has the advantages that the signal of detection is easy to get, sensor arrangement is simple, calculates simple, algorithm simply easily to run.
Description
Technical field
The invention belongs to mechanical arm failure exceptions to monitor field, more particularly, to a kind of manipulator based on data
Arm track method for monitoring abnormality.
Background technique
Prominent with modern science and technology flies to fiercely attack, existing mechanical equipment increasingly complex, precise treatment, systematization and from
Dynamicization；Role and influence of the equipment in modern enterprise are increasing；The main force of production is also gradually turned from manpower to mechanical equipment
It moves.Therefore it is necessary to guarantee the normal operation of various equipment, assessment, real-time status of the operation monitoring technology are made to the state of equipment
Equipment operating condition is detected, early prediction and in time diagnosis are carried out to wherein implicit failure.
In the past few decades, various mechanical arms are widely used in industrial production, bring enormous benefits.But it is previous
Fault diagnosis be often mostly based on sense organ control (vision, sense of hearing etc.), trigger-type alarming signal, for example be based on machine vision
Industrial robot intelligent control method, position, angle, motion profile and movement velocity are judged by vision, then pass through
After central processing module carries out data processing, control information is sent to industrial robot.
Common mechanical equipment fault exception monitoring parameter is generally measured with corresponding sensor, main detection parameters
There are electromagnetic parameter, vibration parameters, parameters,acoustic, mechanics parameter, kinematic parameter, hydraulic parameter etc..But often data acquire
On need to obtain certain signals for being difficult to detect, the problems such as sensor arrangement trouble and method to calculate complicated or algorithm complicated
It spends high-leveled and difficult with operation.
Summary of the invention
Aiming at the above defects or improvement requirements of the prior art, the present invention provides a kind of mechanical arm rail based on data
Mark method for monitoring abnormality, the signal that thus the solution prior art has detection are difficult to obtain, sensor arrangement trouble, calculate again
Miscellaneous, algorithm complexity it is high-leveled and difficult to run the technical issues of.
To achieve the above object, the present invention provides a kind of mechanical arm track method for monitoring abnormality based on data, packet
It includes:
Malfunction monitoring is carried out to the characteristic signal under the real-time working condition of mechanical arm using malfunction monitoring model, is obtained in real time
Characteristic correlation coefficient judges the real-time working condition of mechanical arm for normal condition or abnormal shape using real-time characteristic related coefficient
State；
The training of the malfunction monitoring model includes:
(1) it is clear to carry out data to characteristic signal for characteristic signal of the collection machinery arm under normal condition and abnormality
It washes, obtains training set；
(2) using training set training neural network, malfunction monitoring model is obtained.
Further, when real-time characteristic related coefficient is less than 0.2, the real-time working condition of mechanical arm is normal condition, otherwise
The real-time working condition of mechanical arm is abnormality.
Further, step (2) includes:
When using training set training neural network, the depth characteristic of training set, benefit are extracted using the convolutional layer of neural network
Depth characteristic is merged with the full articulamentum of neural network, obtains characteristic correlation coefficient, the characteristic correlation coefficient corresponds to machinery
The normal condition or abnormality of arm, and then obtain malfunction monitoring model.
Further, data cleansing is mean filter, window filtering or complementary filter.
Further, characteristic signal is three axis angular rate signals and 3-axis acceleration signal.
Further, characteristic signal is collected by six axis gyro sensors.
In general, through the invention it is contemplated above technical scheme is compared with the prior art, can obtain down and show
Beneficial effect:
(1) present invention proposes a kind of failure exception monitoring method of mechanical arm based on data, it is intended to by machine learning
It is applied to the failure exception monitoring problem of mechanical arm with deep learning method, collection machinery arm is in normal condition and abnormal shape
Characteristic signal under state carries out data cleansing to characteristic signal, obtains training set；Using training set training neural network, obtain
Malfunction monitoring model.Real-time failure exception monitoring is carried out using malfunction monitoring model.So that the present invention has the signal of detection
Be easy to get, sensor arrangement is simple, calculate simple, algorithm simply easily to run the advantages of.
(2) the method for the present invention use deep learning theoretical algorithm simply easily realize and be suitable in line computation, it can be achieved that
The mechanical arm failure exception of industry spot is predicted in real time, avoids generating industrial production further loss, and this method is only
Equipment surface characteristic signal in the collection machinery arm course of work, six axis gyro sensors of arrangement can easily acquire spy
Reference number, the arrangement of sensor do not influence the course of normal operation of mechanical arm, do not change mechanical arm physical structure itself, work
It is easy to accomplish in journey.Production process characteristic is excavated using deep learning, it is new to disclose production process using characteristic correlation coefficient
Causality and correlation mechanism, the deep physical essence for annotating production process, so to manufacturing process future evolution process into
Row prediction, realizes life prediction and failure predication, greatly improves the reliability and efficiency of process unit, sufficiently excavation intelligence manufacture
The value chain of big data, so that big data really becomes the enabling tool of the following intelligence manufacture.
Detailed description of the invention
Fig. 1 is a kind of process of mechanical arm track method for monitoring abnormality based on data provided in an embodiment of the present invention
Figure；
Fig. 2 is mechanical arm normal trace operating characteristic signal waveforms provided in an embodiment of the present invention；
Fig. 3 is abnormal work characteristic signal waveform diagram in mechanical arm track provided in an embodiment of the present invention；
Fig. 4 (a) is the wavy curve of x-axis angular speed provided in an embodiment of the present invention；
Fig. 4 (b) is the wavy curve of y-axis angular speed provided in an embodiment of the present invention；
Fig. 4 (c) is the wavy curve of z-axis angular speed provided in an embodiment of the present invention；
Fig. 4 (d) is the wavy curve of x-axis acceleration provided in an embodiment of the present invention；
Fig. 4 (e) is the wavy curve of y-axis acceleration provided in an embodiment of the present invention；
Fig. 4 (f) is the wavy curve of z-axis acceleration provided in an embodiment of the present invention；
Fig. 5 is the characteristic correlation coefficient characteristic curve diagram of characterization mechanical arm track exception provided in an embodiment of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.As long as in addition, technical characteristic involved in the various embodiments of the present invention described below
Not constituting a conflict with each other can be combined with each other.
As shown in Figure 1, a kind of mechanical arm track method for monitoring abnormality based on data, comprising:
Malfunction monitoring is carried out to the characteristic signal under the real-time working condition of mechanical arm using malfunction monitoring model, is obtained in real time
Characteristic correlation coefficient judges the real-time working condition of mechanical arm for normal condition or abnormal shape using real-time characteristic related coefficient
State；When real-time characteristic related coefficient is less than 0.2, the real-time working condition of mechanical arm is normal condition, and otherwise mechanical arm is real-time
Operating condition is abnormality.
The training of the malfunction monitoring model includes:
(1) it is clear to carry out data to characteristic signal for characteristic signal of the collection machinery arm under normal condition and abnormality
It washes, obtains training set；
(2) using training set training neural network when, the depth characteristic of training set is extracted using the convolutional layer of neural network,
Depth characteristic is merged using the full articulamentum of neural network, obtains characteristic correlation coefficient, the characteristic correlation coefficient corresponds to machine
The normal condition or abnormality of tool arm, and then obtain malfunction monitoring model.
Data cleansing is mean filter, window filtering or complementary filter, and characteristic signal is three axis angular rate signals and three
Axle acceleration signal, characteristic signal are collected by six axis gyro sensors.
As shown in Fig. 2, the running track under setting mechanical arm normal condition, six axis gyro sensors are acquired in normal shape
The 3-axis acceleration signal of mechanical arm fixed point and three axis angular rate signals under state obtain 3-axis acceleration under normal condition and believe
Number data changed over time with three axis angular rate signals and the characteristic signal waveform diagram for generating corresponding normal work.
As shown in figure 3, slightly changing end effector position in operating normally track, thus analog track deformation event
Barrier uses the 3-axis acceleration signal and three shaft angles of six axis gyro sensors acquisition mechanical arm fixed point under abnormality
Speed signal obtains the data that 3-axis acceleration signal is changed over time with three axis angular rate signals under abnormality and generates phase
The characteristic signal waveform diagram of corresponding abnormal work.
Data cleansing is carried out to characteristic signal, training set is obtained, includes 6 class signals in training set, such as Fig. 4 (a), Fig. 4
(b), shown in Fig. 4 (c), Fig. 4 (d), Fig. 4 (e), Fig. 4 (f), it is respectively as follows: the wavy curve of x-axis angular speed, the wave of y-axis angular speed
Shape curve, the wavy curve of z-axis angular speed, the wavy curve of x-axis acceleration, the wavy curve of y-axis acceleration and z-axis acceleration
Wavy curve.
Malfunction monitoring is carried out to the characteristic signal under the real-time working condition of mechanical arm using malfunction monitoring model, is obtained in real time
Characteristic correlation coefficient judges the real-time working condition of mechanical arm for normal condition or abnormal shape using real-time characteristic related coefficient
State；As shown in figure 5, being track exception in rectangle frame.The purpose of the present invention is to provide a kind of manipulators based on data
Arm track method for monitoring abnormality greatly improves the reliability and production efficiency of process unit, sufficiently excavates intelligence manufacture big data
Value chain so that big data really becomes the enabling tool of the following intelligence manufacture.The present invention is set by collection machinery arm
The characteristic signal on standby surface, then signal is once purged, and the causality and correlation of data after analysis cleaning are found out some implicit
Physics law, the characteristic coefficient of equipment fault exception can be characterized by extracting, and be analyzed the vibration signal acquired in real time, meter
The characteristic signal index of equipment fault exception is calculated, realizes the monitoring of mechanical arm failure exception, termination device is run in advance, reduces life
Produce cost.
As it will be easily appreciated by one skilled in the art that the foregoing is merely illustrative of the preferred embodiments of the present invention, not to
The limitation present invention, any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should all include
Within protection scope of the present invention.
Claims (6)
1. a kind of mechanical arm track method for monitoring abnormality based on data characterized by comprising
Malfunction monitoring is carried out to the characteristic signal under the real-time working condition of mechanical arm using malfunction monitoring model, obtains real-time characteristic
Related coefficient judges the real-time working condition of mechanical arm for normal condition or abnormality using real-time characteristic related coefficient；
The training of the malfunction monitoring model includes:
(1) characteristic signal of the collection machinery arm under normal condition and abnormality carries out data cleansing to characteristic signal, obtains
To training set；
(2) using training set training neural network, malfunction monitoring model is obtained.
2. a kind of mechanical arm track method for monitoring abnormality based on data as described in claim 1, which is characterized in that described
When real-time characteristic related coefficient is less than 0.2, the real-time working condition of mechanical arm is normal condition, otherwise the real-time working condition of mechanical arm
For abnormality.
3. a kind of mechanical arm track method for monitoring abnormality based on data as claimed in claim 1 or 2, which is characterized in that
The step (2) includes:
When using training set training neural network, the depth characteristic of training set is extracted using the convolutional layer of neural network, utilizes mind
Full articulamentum through network merges depth characteristic, obtains characteristic correlation coefficient, and the characteristic correlation coefficient corresponds to mechanical arm
Normal condition or abnormality, and then obtain malfunction monitoring model.
4. a kind of mechanical arm track method for monitoring abnormality based on data as claimed in claim 1 or 2, which is characterized in that
The data cleansing is mean filter, window filtering or complementary filter.
5. a kind of mechanical arm track method for monitoring abnormality based on data as claimed in claim 1 or 2, which is characterized in that
The characteristic signal is three axis angular rate signals and 3-axis acceleration signal.
6. a kind of mechanical arm track method for monitoring abnormality based on data as claimed in claim 1 or 2, which is characterized in that
The characteristic signal is collected by six axis gyro sensors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810511936.9A CN109029937A (en) | 2018-05-24 | 2018-05-24 | A kind of mechanical arm track method for monitoring abnormality based on data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810511936.9A CN109029937A (en) | 2018-05-24 | 2018-05-24 | A kind of mechanical arm track method for monitoring abnormality based on data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109029937A true CN109029937A (en) | 2018-12-18 |
Family
ID=64611260
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810511936.9A Pending CN109029937A (en) | 2018-05-24 | 2018-05-24 | A kind of mechanical arm track method for monitoring abnormality based on data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109029937A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113988202A (en) * | 2021-11-04 | 2022-01-28 | 季华实验室 | Mechanical arm abnormal vibration detection method based on deep learning |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105841961A (en) * | 2016-03-29 | 2016-08-10 | 中国石油大学(华东) | Bearing fault diagnosis method based on Morlet wavelet transformation and convolutional neural network |
| CN106226074A (en) * | 2016-09-22 | 2016-12-14 | 华中科技大学 | Based on convolutional neural networks and the rotary machinery fault diagnosis method of small echo gray-scale map |
| CN106874957A (en) * | 2017-02-27 | 2017-06-20 | 苏州大学 | A kind of Fault Diagnosis of Roller Bearings |
-
2018
- 2018-05-24 CN CN201810511936.9A patent/CN109029937A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105841961A (en) * | 2016-03-29 | 2016-08-10 | 中国石油大学(华东) | Bearing fault diagnosis method based on Morlet wavelet transformation and convolutional neural network |
| CN106226074A (en) * | 2016-09-22 | 2016-12-14 | 华中科技大学 | Based on convolutional neural networks and the rotary machinery fault diagnosis method of small echo gray-scale map |
| CN106874957A (en) * | 2017-02-27 | 2017-06-20 | 苏州大学 | A kind of Fault Diagnosis of Roller Bearings |
Non-Patent Citations (2)
| Title |
|---|
| LUYANG JING等: "A convolutional neural network based feature learning and fault diagnosis method for the condition monitoring of gearbox", 《MEASUREMENT》 * |
| 林焕辉等: "基于力反馈数据手套控制的仿生机械手设计研究", 《机电技术》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113988202A (en) * | 2021-11-04 | 2022-01-28 | 季华实验室 | Mechanical arm abnormal vibration detection method based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhou et al. | Review of tool condition monitoring methods in milling processes | |
| Chen et al. | Unsupervised anomaly detection of industrial robots using sliding-window convolutional variational autoencoder | |
| CN106649727B (en) | Database construction method for fault detection of unmanned aerial vehicle flight control system | |
| Xiang et al. | Fault diagnosis of rolling bearing under fluctuating speed and variable load based on TCO spectrum and stacking auto-encoder | |
| CN111947928B (en) | Multi-source information fusion bearing fault prediction system and method | |
| KR102166649B1 (en) | Machine Diagnosis and Prediction System using Machine Learning | |
| CN104808585B (en) | A kind of quick inspection method of lathe health status | |
| US10839317B2 (en) | Control device and machine learning device | |
| CN108025431A (en) | Actuator diagnostics and prediction | |
| CN112162545B (en) | Automobile fault diagnosis method and system | |
| CN114004256A (en) | Fault diagnosis method for manufacturing equipment main bearing based on digital twin body | |
| CN110375983A (en) | Failsafe valve real-time diagnosis system and diagnostic method based on time series analysis | |
| Zhou et al. | Harmonic reducer in-situ fault diagnosis for industrial robots based on deep learning | |
| CN109029937A (en) | A kind of mechanical arm track method for monitoring abnormality based on data | |
| KR20230159880A (en) | Robot failure precursor detection device and robot failure precursor detection method | |
| Chen et al. | Residual shrinkage transformer relation network for intelligent fault detection of industrial robot with zero-fault samples | |
| CN106123930A (en) | The disconnected fine localization method of a kind of distributed optical fiber sensing system and device | |
| CN113400309B (en) | Motion state identification method of upper limb assistance exoskeleton robot | |
| KR102484471B1 (en) | Method for real-time robot condition diagnosis and apparatus for performing the same | |
| CN109765033A (en) | A kind of flash ranging measuring device camera lens intelligent monitor system and method | |
| Hong et al. | Intelligent anomaly detection of robot manipulator based on energy consumption auditing | |
| Zhong et al. | Industrial robot vibration anomaly detection based on sliding window one-dimensional convolution autoencoder | |
| Kolar et al. | Identification of inability states of rotating machinery subsystems using industrial IoT and convolutional neural network–initial research | |
| CN112091973A (en) | Mechanical arm protective door anti-collision detection method and system | |
| Wang et al. | Milling-tool wear-condition prediction with statistic analysis and echo-state networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |