CN112894882A - Robot fault detection system based on industrial internet - Google Patents
Robot fault detection system based on industrial internet Download PDFInfo
- Publication number
- CN112894882A CN112894882A CN202011610327.2A CN202011610327A CN112894882A CN 112894882 A CN112894882 A CN 112894882A CN 202011610327 A CN202011610327 A CN 202011610327A CN 112894882 A CN112894882 A CN 112894882A
- Authority
- CN
- China
- Prior art keywords
- vibration
- neural network
- fault
- robot
- artificial neural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 14
- 238000013528 artificial neural network Methods 0.000 claims abstract description 26
- 238000003745 diagnosis Methods 0.000 claims abstract description 13
- 238000000605 extraction Methods 0.000 claims abstract description 13
- 239000000284 extract Substances 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 abstract description 6
- 238000004458 analytical method Methods 0.000 abstract description 3
- 238000000354 decomposition reaction Methods 0.000 description 7
- 238000012360 testing method Methods 0.000 description 7
- 238000012549 training Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 210000002569 neuron Anatomy 0.000 description 5
- 230000004913 activation Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000000034 method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J19/00—Accessories fitted to manipulators, e.g. for monitoring, for viewing; Safety devices combined with or specially adapted for use in connection with manipulators
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J19/00—Accessories fitted to manipulators, e.g. for monitoring, for viewing; Safety devices combined with or specially adapted for use in connection with manipulators
- B25J19/0095—Means or methods for testing manipulators
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a robot fault detection system based on industrial internet, comprising: the diagnosis platform diagnoses the fault type of the robot based on the vibration signal; the diagnosis platform is integrated with a vibration feature extraction unit and an artificial neural network model, and the output end of the vibration feature unit is connected with the input end of the artificial neural network. The robot monitoring signal is subjected to fault feature extraction by using a discrete wavelet analysis method, and the fault is classified and analyzed by using an artificial neural network, so that complete extraction and fault classification of the fault features can be efficiently realized, and the reliability is good.
Description
Technical Field
The invention belongs to the technical field of fault detection, and particularly relates to a robot fault detection system based on an industrial internet.
Background
According to the test result statistics of a CR test part of the detection and authentication center of the Saidi robot: the functional safety of the domestic robot is great, the average risk failure rate is 3-5 times higher than the standard requirement, wherein the service robot is particularly serious and individually reaches 10 times or more. The electrical key parts adopted by some domestic robots are not authenticated or not in accordance with standard requirements, so that serious potential safety hazards exist, and the safety and reliability of the whole robot are greatly reduced.
The traditional robot monitoring and fault diagnosis system is mostly based on a certain fixed area or a factory workshop, and lacks a full industrial chain cooperative mechanism and a full life cycle management system, so that information islands exist in equipment monitoring and fault diagnosis, data cannot be shared, data safety cannot be guaranteed, and remote measurement and control cannot be realized. Therefore, the device has the advantages of slow maintenance response of the industrial robot, high operation and maintenance cost of the industrial robot, low effective utilization rate of the industrial robot and high loss of the industrial robot.
Disclosure of Invention
The invention provides a robot fault detection system based on an industrial internet, aiming at improving the problems.
The invention is realized in such a way that an industrial internet-based robot fault detection system comprises:
the diagnosis platform diagnoses the fault type of the robot based on the vibration signal;
the diagnosis platform is integrated with a vibration feature extraction unit and an artificial neural network model, and the output end of the vibration feature unit is connected with the input end of the artificial neural network.
Furthermore, the vibration feature extraction unit is composed of a multi-stage high-pass filter and a multi-stage low-pass filter, the vibration signal is input into the high-pass filter and the low-pass filter of one stage, and the input ends of the high-pass filter and the low-pass filter of the next stage are connected with the output end of the low-pass filter of the previous stage.
Further, the vibration feature extraction unit extracts vibration features of the input vibration signals based on discrete wavelet transformation, namely, extracts vibration frequencies changing along with time, inputs the extracted vibration features to the artificial neural network model, and the artificial neural network model outputs fault types corresponding to the vibration features.
The robot fault detection system based on the industrial internet has the following beneficial effects:
1) the robot monitoring signal is subjected to fault feature extraction by using a discrete wavelet analysis method, and the fault is classified and analyzed by using an artificial neural network, so that complete extraction and fault classification of the fault features can be efficiently realized, and the reliability is good.
2) The industrial internet platform is used for efficiently realizing the robot monitoring and fault diagnosis algorithm, and efficiently realizing the sharing and cooperative processing of the robot monitoring and fault diagnosis analysis information.
Drawings
Fig. 1 is a schematic structural diagram of a robot fault detection system of an industrial internet according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a DWT decomposition robot vibration signal provided by an embodiment of the present invention;
fig. 3 is a schematic diagram of a weight-based artificial neuron network according to an embodiment of the present invention.
Detailed Description
The following detailed description of the embodiments of the present invention will be given in order to provide those skilled in the art with a more complete, accurate and thorough understanding of the inventive concept and technical solutions of the present invention.
Fig. 1 is a schematic structural diagram of a robot fault detection system of an industrial internet according to an embodiment of the present invention, and for convenience of description, only a part related to the embodiment of the present invention is shown.
The system comprises:
the diagnosis platform diagnoses the fault type of the robot based on the vibration signal;
in the embodiment of the invention, a vibration feature extraction unit and an Artificial Neural Network (ANN) model are integrated on a diagnosis platform, and the output end of the vibration feature unit is connected with the input end of the artificial neural network;
the vibration feature extraction unit consists of a plurality of stages of high-pass filters and low-pass filters, the high-pass filter and the low-pass filter of the next stage are connected with the low-pass filter of the previous stage, and the decomposition stage number is determined based on the pulling-up of the lowest frequency band required by tracking; respectively inputting the received vibration signal x [ n ] into a first-level high-pass filter and a first-level low-pass filter, extracting the vibration characteristics of the vibration signal based on Discrete Wavelet Transform (DWT), namely extracting the vibration frequency changing along with time, inputting the extracted vibration characteristics into an artificial neural network model, and outputting the fault type corresponding to the vibration characteristics by the artificial neural network model.
The vibration signal is analyzed for high frequencies by a High Pass (HP) filter and then for low frequencies by a Low Pass (LP) filter. In the first stage, the original signal x [ n ]]Decomposing the two signals y through a high-pass filter and a low-pass filter to obtain two signalshigh[k],ylow[k]The signal coefficients extracted from the HP filter, i.e., the detail coefficients of the first stage (cD1), are double down sampled. These coefficients contain the high frequency information of the original signal, while the coefficients extracted from the LP filter are called the approximation coefficients of the first stage (cA 1).
yhigh[k]=∑nx[n]*g[2k-n]
ylow[k]=∑nx[n]*h[2k-n]
Wherein, x [ n ]]For n discrete input signals, the functions g and h are transfer functions of a high-pass filter and a low-pass filter, respectively, k is a tap coefficient of the filter, and cAl has a frequency range of [0, Fs/2 ]l+1]cDl has a frequency range of [ Fs/2 ]l +1,Fs/2l]L is the decomposition order, Fs is the sampling frequency.
After obtaining the first stage of decomposition, cA1 was further decomposed into cA2 and cD2, and then after decomposing cA2 into cA3 and cD3, the above process was repeated cyclically until the decomposition reached a preset level. Fig. 2 is a schematic diagram of a multilevel decomposition of a vibration signal. The number of decomposition levels is determined by the lowest frequency band to be tracked.
In the embodiment of the invention, before the artificial neural network is used, training is required, and the training method is as follows:
to have an associated weight w1,w2,w3,…wnNeuron input vector x of1,x2,x3,…xnThe deviation b (sometimes referred to as a threshold) is usually associated with a neuron and introduces an activation function to it, so that even if the input is zero, the neuron will still have an output. Typically, the deviation is set to a value of 1. However, the input to the neuron (called net) will be the sum of the multiplied inputs and their corresponding weights, as shown in fig. 3, which can be written as:
where n is the number of neuronal inputs. The neurons are then input to an activation function (f) (net) to produce an output, which can be expressed as:
ANN is generally divided into feedforward (e.g., multi-layer perceptron neural network (MLPN)) and feedback (or recursive) types, and most fault diagnosis applications that utilize artificial neural networks employ a feedforward architecture, with MLPN being the most prevalent, and thus the present invention employs MLPN.
Firstly, training a neural network by selecting 100 sample data sets monitored by a robot, wherein the sample data consists of fault types and vibration characteristics corresponding to the fault types, initializing the weight and deviation of the network according to a random number generating function of Matlab after setting a neural network structure and loading data, and generating training set, verification set and test set data. Since the data values of the ANN are in different ranges, the data needs to be normalized to achieve efficient quantization of the data. Separating the training data set and the test data set, training the neural network by using the training data set, testing the trained neural network by using the test data set, and repeating iteration until the accurate recognition rate of the test data set reaches a set value.
The invention has been described above with reference to the accompanying drawings, it is obvious that the invention is not limited to the specific implementation in the above-described manner, and it is within the scope of the invention to apply the inventive concept and solution to other applications without substantial modification.
Claims (3)
1. An industrial internet-based robot fault detection system, characterized in that the system comprises:
the diagnosis platform diagnoses the fault type of the robot based on the vibration signal;
the diagnosis platform is integrated with a vibration feature extraction unit and an artificial neural network model, and the output end of the vibration feature unit is connected with the input end of the artificial neural network.
2. The industrial internet-based robot failure detection system according to claim 1, wherein the vibration characteristic extraction unit is composed of a plurality of stages of high pass filters and low pass filters, the vibration signal is inputted to the high pass filter and the low pass filter of one stage, and the input terminals of the high pass filter and the low pass filter of the subsequent stage are connected to the output terminal of the low pass filter of the previous stage.
3. The industrial internet-based robot fault detection system according to claim 1, wherein the vibration feature extraction unit extracts vibration features of the input vibration signal based on discrete wavelet transform, that is, extracts vibration frequencies varying with time, inputs the extracted vibration features to the artificial neural network model, and the artificial neural network model outputs fault types corresponding to the vibration features.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011610327.2A CN112894882A (en) | 2020-12-30 | 2020-12-30 | Robot fault detection system based on industrial internet |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011610327.2A CN112894882A (en) | 2020-12-30 | 2020-12-30 | Robot fault detection system based on industrial internet |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112894882A true CN112894882A (en) | 2021-06-04 |
Family
ID=76112045
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011610327.2A Pending CN112894882A (en) | 2020-12-30 | 2020-12-30 | Robot fault detection system based on industrial internet |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112894882A (en) |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101173870A (en) * | 2007-09-27 | 2008-05-07 | 上海交通大学 | Identification and detecting method for short time acoustical feature signal time-frequency domain under complex background noise |
CN101435799A (en) * | 2008-12-19 | 2009-05-20 | 清华大学 | Failure diagnosis method and apparatus of hydroturbine based on acoustic emission technology |
CN101612735A (en) * | 2009-07-24 | 2009-12-30 | 哈尔滨工业大学 | Mobile robotic vision system anti-shake apparatus and anti-shake compensating control method |
CN102183366A (en) * | 2011-03-08 | 2011-09-14 | 上海大学 | Device and method for vibration measurement and failure analysis of rolling bearing |
US20120277912A1 (en) * | 2011-04-28 | 2012-11-01 | Seiko Epson Corporation | Robot controller, simple installation-type robot, and method of controlling simple installation-type robot |
CN102840979A (en) * | 2012-09-04 | 2012-12-26 | 国电联合动力技术有限公司 | Method and device for detecting transmission chain faults of wind turbine generator set |
CN104198184A (en) * | 2014-08-11 | 2014-12-10 | 中国人民解放军空军工程大学 | Bearing fault diagnostic method based on second generation wavelet transform and BP neural network |
CN105372087A (en) * | 2015-11-30 | 2016-03-02 | 南通大学 | Fault diagnosis method based on multi-sensor signal analysis |
CN106203359A (en) * | 2016-07-15 | 2016-12-07 | 重庆邮电大学 | Fault Diagnosis of Internal Combustion Engine method based on wavelet packet analysis and k nearest neighbor algorithm |
CN106599794A (en) * | 2016-11-21 | 2017-04-26 | 南京熊猫电子股份有限公司 | Six-axis industrial robot fault diagnosis method and system based on AR model |
CN109213305A (en) * | 2017-06-29 | 2019-01-15 | 沈阳新松机器人自动化股份有限公司 | A kind of gesture identification method based on surface electromyogram signal |
US20200282575A1 (en) * | 2019-03-05 | 2020-09-10 | The Boeing Company | Automatic calibration for a robot optical sensor |
-
2020
- 2020-12-30 CN CN202011610327.2A patent/CN112894882A/en active Pending
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101173870A (en) * | 2007-09-27 | 2008-05-07 | 上海交通大学 | Identification and detecting method for short time acoustical feature signal time-frequency domain under complex background noise |
CN101435799A (en) * | 2008-12-19 | 2009-05-20 | 清华大学 | Failure diagnosis method and apparatus of hydroturbine based on acoustic emission technology |
CN101612735A (en) * | 2009-07-24 | 2009-12-30 | 哈尔滨工业大学 | Mobile robotic vision system anti-shake apparatus and anti-shake compensating control method |
CN102183366A (en) * | 2011-03-08 | 2011-09-14 | 上海大学 | Device and method for vibration measurement and failure analysis of rolling bearing |
US20120277912A1 (en) * | 2011-04-28 | 2012-11-01 | Seiko Epson Corporation | Robot controller, simple installation-type robot, and method of controlling simple installation-type robot |
CN102840979A (en) * | 2012-09-04 | 2012-12-26 | 国电联合动力技术有限公司 | Method and device for detecting transmission chain faults of wind turbine generator set |
CN104198184A (en) * | 2014-08-11 | 2014-12-10 | 中国人民解放军空军工程大学 | Bearing fault diagnostic method based on second generation wavelet transform and BP neural network |
CN105372087A (en) * | 2015-11-30 | 2016-03-02 | 南通大学 | Fault diagnosis method based on multi-sensor signal analysis |
CN108931387A (en) * | 2015-11-30 | 2018-12-04 | 南通大学 | The method for diagnosing faults based on multiple sensor signals analysis of Accurate Diagnosis decision is provided |
CN106203359A (en) * | 2016-07-15 | 2016-12-07 | 重庆邮电大学 | Fault Diagnosis of Internal Combustion Engine method based on wavelet packet analysis and k nearest neighbor algorithm |
CN106599794A (en) * | 2016-11-21 | 2017-04-26 | 南京熊猫电子股份有限公司 | Six-axis industrial robot fault diagnosis method and system based on AR model |
CN109213305A (en) * | 2017-06-29 | 2019-01-15 | 沈阳新松机器人自动化股份有限公司 | A kind of gesture identification method based on surface electromyogram signal |
US20200282575A1 (en) * | 2019-03-05 | 2020-09-10 | The Boeing Company | Automatic calibration for a robot optical sensor |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111585948B (en) | Intelligent network security situation prediction method based on power grid big data | |
Guh et al. | A neural network approach to characterize pattern parameters in process control charts | |
CN114740361B (en) | Fuel cell voltage prediction method based on long-short-term memory neural network model | |
CN112507915B (en) | Bolt connection structure loosening state identification method based on vibration response information | |
CN110866604A (en) | Cleaning method for power transformer state monitoring data | |
Chiu et al. | Identification of process disturbance using SPC/EPC and neural networks | |
CN113657622B (en) | Multi-dimensional state data fusion method, device, terminal and storage medium for power equipment | |
CN116316591A (en) | Short-term photovoltaic power prediction method and system based on hybrid bidirectional gating cycle | |
CN111428201A (en) | Prediction method for time series data based on empirical mode decomposition and feedforward neural network | |
CN111678679A (en) | Circuit breaker fault diagnosis method based on PCA-BPNN | |
CN111079647A (en) | Circuit breaker defect identification method | |
CN112894882A (en) | Robot fault detection system based on industrial internet | |
Addeh et al. | Control chart pattern recognition using adaptive back-propagation artificial Neural networks and efficient features | |
CN111582132B (en) | Improved EEMD and PCNN-based gas leakage signal noise reduction method | |
CN112686182A (en) | Partial discharge mode identification method and terminal equipment | |
Jin et al. | A review of fault diagnosis research on on-load tap-changers | |
CN112287862A (en) | Fault diagnosis method based on wavelet convolution neural network | |
CN117473411A (en) | Bearing life prediction method based on improved transducer model | |
Li et al. | A robust fault diagnosis method for rolling bearings based on deep convolutional neural network | |
CN116559728A (en) | Power transmission line fault diagnosis method based on wavelet transformation-multilayer perceptron | |
CN116405139A (en) | Spectrum prediction model and method based on Informar | |
CN116304940A (en) | Analog circuit fault diagnosis method based on long-short-term memory neural network | |
CN114966224A (en) | Improved neural network-based modular multilevel converter capacitance monitoring method | |
CN114139577A (en) | Fault arc identification method and device based on one-dimensional convolution self-coding network | |
CN113722659B (en) | Fault mode-based substation main equipment diagnosis method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210604 |