CN109359767A - A kind of intelligence expression method and device of intelligence manufacture procedure fault recognition result - Google Patents
A kind of intelligence expression method and device of intelligence manufacture procedure fault recognition result Download PDFInfo
- Publication number
- CN109359767A CN109359767A CN201811113319.XA CN201811113319A CN109359767A CN 109359767 A CN109359767 A CN 109359767A CN 201811113319 A CN201811113319 A CN 201811113319A CN 109359767 A CN109359767 A CN 109359767A
- Authority
- CN
- China
- Prior art keywords
- layer
- layer network
- recognition result
- manufacturing process
- single layer
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Abstract
The present invention relates to intelligence manufacture failure predication technical fields, more particularly to a kind of intelligence expression method and device of intelligence manufacture procedure fault recognition result, pass through the historical failure signal generated in acquisition manufacturing process, generate the single layer network of fault data, and then generate visual deep layer network parameter, by determine current layer and it is one layer upper between feature space similar characteristic and variation tendency, express the fault identification result based on depth network, the present invention can intuitively, clearly express the health status of manufacturing process, and various aspects show the health status of manufacturing process.
Description
Technical field
The present invention relates to intelligence manufacture failure predication technical fields, and in particular to a kind of intelligence manufacture procedure fault identification knot
The intelligence expression method and device of fruit.
Background technique
Complex Industrial Systems the device is complicated change, system integration, scale the Enlargement Tendency become increasingly conspicuous, so that fault-signal
Have the characteristics that concealment, timely and effectively detection be out of order and take measures, can be improved production efficiency, Improving The Quality of Products,
Reduce economic loss and unnecessary personal injury.Therefore, the fault diagnosis of Complex Industrial Systems safety and reliability is improved
Technology more and more attention has been paid to.
Failure in modern industry system have hidden, random, amplitude is low, fault characteristic is unobvious, easily covered by noise,
The characteristics of easily being flooded by larger normal variation.By research, the other result of malfunction forms the healthy situation map of manufacturing process, and
When predict and diagnosis is safeguards system safety and one of the key factor that inhibits failure to deteriorate.And it how intuitively, clearly expresses
The health status of manufacturing process, various aspects, which show that the health status of manufacturing process becomes, to be worth solving the problems, such as.
Summary of the invention
The present invention provides a kind of intelligence expression method and device of intelligence manufacture procedure fault recognition result, can be intuitive, clear
The health status of manufacturing process is expressed clearly, and various aspects show the health status of manufacturing process.
A kind of intelligent expression of intelligence manufacture procedure fault recognition result provided by the invention, comprising the following steps:
Step A, the historical failure signal generated in manufacturing process is acquired, cancelling noise signal generates sample fault data；
Step B, sample fault data is generated into single layer network data；
Step C, the weight of single layer network training is split as to the combination of multiple basic functions；
Step D, visual deep layer network parameter is generated；
Step E, determine current layer and it is one layer upper between feature space similar characteristic and variation tendency；
Step F, the fault identification result based on depth network is expressed.
Further, historical failure data described in the step A includes: the amplitude of the signal generated in manufacturing process, phase
Position and frequency；The signal generated in the manufacturing process is made of fault-signal and noise signal.
Further, the step C is specifically included: being visualized by the parameter of single layer network, by the power of single layer network training
It is split as the combination of multiple basic functions again, by the point product transformation of original signal and multiple basic functions to carry out single layer network
Feature extraction.
Further, the step D is specifically included:
Step D1, single layer network parameter is visualized；
Step D2, existing activation relationship between current layer neuron and upper one layer of neuron in depth network is obtained；
Step D3, the maximum input pattern for determining activation current layer neuron node is filtered the node as upper one layer
The linear weighted combination of wave device；
Step D4, deep layer network parameter is generated；
Step D5, by the visualization of deep layer network parameter.
Further, the step E is specifically included:
The feature extracted by layer-by-layer visual network, determine current layer and it is one layer upper between feature space similar characteristic with
Variation tendency.
A kind of intelligent expression device of intelligence manufacture procedure fault recognition result provided by the invention, including control module and
The memory module of control store instruction, control module read described instruction and execute following steps:
Step A, the historical failure signal generated in manufacturing process is acquired, cancelling noise signal generates sample fault data；
Step B, sample fault data is generated into single layer network data；
Step C, the weight of single layer network training is split as to the combination of multiple basic functions；
Step D, visual deep layer network parameter is generated；
Step E, determine current layer and it is one layer upper between feature space similar characteristic and variation tendency；
Step F, the fault identification result based on depth network is expressed.
Further, historical failure data described in the step A includes: the amplitude of the signal generated in manufacturing process, phase
Position and frequency；The signal generated in the manufacturing process is made of fault-signal and noise signal.
Further, control module reads described instruction and specifically executes step in the step C are as follows:
It is visualized by the parameter of single layer network, the weight of single layer network training is split as to the combination of multiple basic functions,
By the point product transformation of original signal and multiple basic functions to carry out the feature extraction of single layer network.
Further, control module reads described instruction and specifically executes step in the step D are as follows:
Step D1, single layer network parameter is visualized；
Step D2, existing activation relationship between current layer neuron and upper one layer of neuron in depth network is obtained；
Step D3, the maximum input pattern for determining activation current layer neuron node is filtered the node as upper one layer
The linear weighted combination of wave device；
Step D4, deep layer network parameter is generated；
Step D5, by the visualization of deep layer network parameter.
Further, control module reads described instruction and specifically executes step in the step E are as follows:
The feature extracted by layer-by-layer visual network, determine current layer and it is one layer upper between feature space similar characteristic with
Variation tendency.
The beneficial effects of the present invention are: the present invention discloses a kind of intelligent expression side of intelligence manufacture procedure fault recognition result
Method and device generate the single layer network of fault data, and then generate by the historical failure signal generated in acquisition manufacturing process
Visual deep layer network parameter, by determine current layer and it is one layer upper between feature space similar characteristic and variation tendency, table
It is many-sided up to the fault identification based on depth network as a result, the present invention can intuitively, clearly express the health status of manufacturing process
Show the health status of manufacturing process.
Detailed description of the invention
The invention will be further described with example with reference to the accompanying drawing.
Fig. 1 is a kind of intelligence expression method flow signal of intelligence manufacture procedure fault recognition result of the embodiment of the present invention
Figure；
Fig. 2 is the stream of the intelligence expression method and step D of intelligence manufacture procedure fault recognition result of the embodiment of the present invention a kind of
Journey schematic diagram.
Specific embodiment
With reference to Fig. 1, a kind of intelligent expression of intelligence manufacture procedure fault recognition result provided in an embodiment of the present invention,
The following steps are included:
Step A, the historical failure signal generated in manufacturing process is acquired, cancelling noise signal generates sample fault data；
Step B, sample fault data is generated into single layer network data；
Step C, the weight of single layer network training is split as to the combination of multiple basic functions；
Step D, visual deep layer network parameter is generated；
Step E, determine current layer and it is one layer upper between feature space similar characteristic and variation tendency, thus parse failure letter
The expression pattern of breath from the superficial to the deep, the immanent structure of intuitive reflection monitoring big data, enhances the understanding to monitoring big data essence；
Step F, the fault identification based on depth network is expressed as a result, forming manufacturing process using these recognition results in turn
Healthy situation map, clearly express manufacturing process health status, in conjunction with radar map, rising sun figure etc. from result identification, become
Many-sided health status for showing manufacturing process such as gesture development, confidence assessment.
Further, historical failure data described in the step A includes: the amplitude of the signal generated in manufacturing process, phase
Position and frequency；The signal generated in the manufacturing process is made of fault-signal and noise signal.
Further, the step C is specifically included: being visualized by the parameter of single layer network, by the power of single layer network training
It is split as the combination of multiple basic functions again, by the point product transformation of original signal and multiple basic functions to carry out single layer network
Feature extraction.
With reference to Fig. 2, further, the step D is specifically included:
Step D1, single layer network parameter is visualized；
Step D2, existing activation relationship between current layer neuron and upper one layer of neuron in depth network is obtained；
Step D3, the maximum input pattern for determining activation current layer neuron node is filtered the node as upper one layer
The linear weighted combination of wave device；
Step D4, deep layer network parameter is generated；
Step D5, by the visualization of deep layer network parameter.
Further, the step E is specifically included:
The feature extracted by layer-by-layer visual network, determine current layer and it is one layer upper between feature space similar characteristic with
Variation tendency.
A kind of intelligent expression device of intelligence manufacture procedure fault recognition result provided in an embodiment of the present invention, including control
The memory module of module and control store instruction, control module read described instruction and execute following steps:
Step A, the historical failure signal generated in manufacturing process is acquired, cancelling noise signal generates sample fault data；
Step B, sample fault data is generated into single layer network data；
Step C, the weight of single layer network training is split as to the combination of multiple basic functions；
Step D, visual deep layer network parameter is generated；
Step E, determine current layer and it is one layer upper between feature space similar characteristic and variation tendency, thus parse failure letter
The expression pattern of breath from the superficial to the deep, the immanent structure of intuitive reflection monitoring big data, enhances the understanding to monitoring big data essence；
Step F, the fault identification based on depth network is expressed as a result, forming manufacturing process using these recognition results in turn
Healthy situation map, clearly express manufacturing process health status, in conjunction with radar map, rising sun figure etc. from result identification, become
Many-sided health status for showing manufacturing process such as gesture development, confidence assessment.
Further, historical failure data described in the step A includes: the amplitude of the signal generated in manufacturing process, phase
Position and frequency；The signal generated in the manufacturing process is made of fault-signal and noise signal.
Further, control module reads described instruction and specifically executes step in the step C are as follows:
It is visualized by the parameter of single layer network, the weight of single layer network training is split as to the combination of multiple basic functions,
By the point product transformation of original signal and multiple basic functions to carry out the feature extraction of single layer network.
Further, control module reads described instruction and specifically executes step in the step D are as follows:
Step D1, single layer network parameter is visualized；
Step D2, existing activation relationship between current layer neuron and upper one layer of neuron in depth network is obtained；
Step D3, the maximum input pattern for determining activation current layer neuron node is filtered the node as upper one layer
The linear weighted combination of wave device；
Step D4, deep layer network parameter is generated；
Step D5, by the visualization of deep layer network parameter.
Further, control module reads described instruction and specifically executes step in the step E are as follows:
The feature extracted by layer-by-layer visual network, determine current layer and it is one layer upper between feature space similar characteristic with
Variation tendency.
The above, only presently preferred embodiments of the present invention, the invention is not limited to above embodiment, as long as
It reaches technical effect of the invention with identical means, all should belong to protection scope of the present invention.
Claims (10)
1. a kind of intelligent expression of intelligence manufacture procedure fault recognition result, which comprises the following steps:
Step A, the historical failure signal generated in manufacturing process is acquired, cancelling noise signal generates sample fault data；
Step B, sample fault data is generated into single layer network data；
Step C, the weight of single layer network training is split as to the combination of multiple basic functions；
Step D, visual deep layer network parameter is generated；
Step E, determine current layer and it is one layer upper between feature space similar characteristic and variation tendency；
Step F, the fault identification result based on depth network is expressed.
2. a kind of intelligent expression of intelligence manufacture procedure fault recognition result according to claim 1, feature exist
In historical failure data described in the step A includes: the amplitude of the signal generated in manufacturing process, phase and frequency；It is described
The signal generated in manufacturing process is made of fault-signal and noise signal.
3. a kind of intelligent expression of intelligence manufacture procedure fault recognition result according to claim 1, feature exist
In the step C is specifically included: being visualized, the weight of single layer network training is split as multiple by the parameter of single layer network
The combination of basic function, by the point product transformation of original signal and multiple basic functions to carry out the feature extraction of single layer network.
4. a kind of intelligent expression of intelligence manufacture procedure fault recognition result according to claim 1, feature exist
In the step D is specifically included:
Step D1, single layer network parameter is visualized；
Step D2, existing activation relationship between current layer neuron and upper one layer of neuron in depth network is obtained；
Step D3, the maximum input pattern for determining activation current layer neuron node, using the node as upper one layer of filter
Linear weighted combination；
Step D4, deep layer network parameter is generated；
Step D5, by the visualization of deep layer network parameter.
5. a kind of intelligent expression of intelligence manufacture procedure fault recognition result according to claim 1, feature exist
In the step E is specifically included:
The feature extracted by layer-by-layer visual network, determine current layer and it is one layer upper between feature space similar characteristic and variation
Trend.
6. a kind of intelligent expression device of intelligence manufacture procedure fault recognition result, which is characterized in that including control module and deposit
The memory module of control instruction is stored up, control module reads described instruction and executes following steps:
Step A, the historical failure signal generated in manufacturing process is acquired, cancelling noise signal generates sample fault data；
Step B, sample fault data is generated into single layer network data；
Step C, the weight of single layer network training is split as to the combination of multiple basic functions；
Step D, visual deep layer network parameter is generated；
Step E, determine current layer and it is one layer upper between feature space similar characteristic and variation tendency；
Step F, the fault identification result based on depth network is expressed.
7. a kind of intelligent expression device of intelligence manufacture procedure fault recognition result according to claim 6, feature exist
In historical failure data described in the step A includes: the amplitude of the signal generated in manufacturing process, phase and frequency；It is described
The signal generated in manufacturing process is made of fault-signal and noise signal.
8. a kind of intelligent expression device of intelligence manufacture procedure fault recognition result according to claim 6, feature exist
In control module reads described instruction and specifically executes step in the step C are as follows:
It is visualized by the parameter of single layer network, the weight of single layer network training is split as to the combination of multiple basic functions, is passed through
The point product transformation of original signal and multiple basic functions is to carry out the feature extraction of single layer network.
9. a kind of intelligent expression device of intelligence manufacture procedure fault recognition result according to claim 6, feature exist
In control module reads described instruction and specifically executes step in the step D are as follows:
Step D1, single layer network parameter is visualized；
Step D2, existing activation relationship between current layer neuron and upper one layer of neuron in depth network is obtained；
Step D3, the maximum input pattern for determining activation current layer neuron node, using the node as upper one layer of filter
Linear weighted combination；
Step D4, deep layer network parameter is generated；
Step D5, by the visualization of deep layer network parameter.
10. a kind of intelligent expression device of intelligence manufacture procedure fault recognition result according to claim 6, feature exist
In control module reads described instruction and specifically executes step in the step E are as follows:
The feature extracted by layer-by-layer visual network, determine current layer and it is one layer upper between feature space similar characteristic and variation
Trend.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811113319.XA CN109359767B (en) | 2018-09-25 | 2018-09-25 | Intelligent expression method and device for fault recognition result in intelligent manufacturing process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811113319.XA CN109359767B (en) | 2018-09-25 | 2018-09-25 | Intelligent expression method and device for fault recognition result in intelligent manufacturing process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109359767A true CN109359767A (en) | 2019-02-19 |
| CN109359767B CN109359767B (en) | 2022-03-25 |
Family
ID=65351356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811113319.XA Active CN109359767B (en) | 2018-09-25 | 2018-09-25 | Intelligent expression method and device for fault recognition result in intelligent manufacturing process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109359767B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040030435A1 (en) * | 2002-08-07 | 2004-02-12 | Popp Robert L. | Manufacturing information and troubleshooting system and method |
| CN104914851A (en) * | 2015-05-21 | 2015-09-16 | 北京航空航天大学 | Adaptive fault detection method for airplane rotation actuator driving device based on deep learning |
| CN106953802A (en) * | 2017-03-01 | 2017-07-14 | 浙江工商大学 | A kind of network optimal route selection method based on deep learning |
| US20180284742A1 (en) * | 2016-05-09 | 2018-10-04 | StrongForce IoT Portfolio 2016, LLC | Methods and systems for industrial internet of things data collection for process adjustment in an upstream oil and gas environment |
| WO2018193377A1 (en) * | 2017-04-17 | 2018-10-25 | Cerebras Systems Inc. | Control wavelet for accelerated deep learning |
-
2018
- 2018-09-25 CN CN201811113319.XA patent/CN109359767B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040030435A1 (en) * | 2002-08-07 | 2004-02-12 | Popp Robert L. | Manufacturing information and troubleshooting system and method |
| CN104914851A (en) * | 2015-05-21 | 2015-09-16 | 北京航空航天大学 | Adaptive fault detection method for airplane rotation actuator driving device based on deep learning |
| US20180284742A1 (en) * | 2016-05-09 | 2018-10-04 | StrongForce IoT Portfolio 2016, LLC | Methods and systems for industrial internet of things data collection for process adjustment in an upstream oil and gas environment |
| CN106953802A (en) * | 2017-03-01 | 2017-07-14 | 浙江工商大学 | A kind of network optimal route selection method based on deep learning |
| WO2018193377A1 (en) * | 2017-04-17 | 2018-10-25 | Cerebras Systems Inc. | Control wavelet for accelerated deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 赵光权等: "基于DBN的故障特征提取及诊断方法研究", 《仪器仪表学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109359767B (en) | 2022-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhao et al. | Challenges and opportunities of AI-enabled monitoring, diagnosis & prognosis: A review | |
| Lu et al. | Early fault detection approach with deep architectures | |
| CN102128022B (en) | Drilling engineering early warning method and system thereof | |
| CN103323749B (en) | The partial discharge diagnostic method of multi-classifier information fusion | |
| CN102915447B (en) | Binary tree-based SVM (support vector machine) classification method | |
| CN101718634B (en) | Equipment state comprehensive dynamic alarming method based on multivariate probability model | |
| CN110534118A (en) | Transformer/reactor method for diagnosing faults based on Application on Voiceprint Recognition and neural network | |
| Si et al. | A novel approach for coal seam terrain prediction through information fusion of improved D–S evidence theory and neural network | |
| CN102841131B (en) | Intelligent steel cord conveyer belt defect identification method and intelligent steel cord conveyer belt defect identification system | |
| Oh et al. | Acoustic data condensation to enhance pipeline leak detection | |
| Zhang et al. | Bearing performance degradation assessment based on time-frequency code features and SOM network | |
| Yan et al. | Big-data-driven based intelligent prognostics scheme in industry 4.0 environment | |
| CN106649755A (en) | Threshold self-adaption setting abnormity detection method for multi-dimensional real-time power transformation device data | |
| CN105626502A (en) | Plunger pump health assessment method based on wavelet packet and Laplacian Eigenmap | |
| CN110263944A (en) | A kind of multivariable failure prediction method and device | |
| Barbosa et al. | Classification of multiple and single power quality disturbances using a decision tree-based approach | |
| KR20080074441A (en) | Method for diagnosis and analysis of electric power quality using artificial intelligence | |
| CN110119756B (en) | Automatic trend data feature selection method based on voting method | |
| CN109359767A (en) | A kind of intelligence expression method and device of intelligence manufacture procedure fault recognition result | |
| CN112180315A (en) | Fault feature extraction method, device and system for optical fiber current transformer | |
| Wang et al. | First-order differential filtering spectrum division method and information fusion multi-scale network for fault diagnosis of bearings under different loads | |
| Xu et al. | Multi-attention-based Feature Aggregation Convolutional Networks with Dual Focal Loss for Fault Diagnosis of Rotating Machinery Under Data Imbalance Conditions | |
| Sheriff et al. | Enhanced performance of shewhart charts using multiscale representation | |
| Xia et al. | Research on the abrasive water-jet cutting machine information fusion fault diagnosis system based on fuzzy neural network | |
| Li | Recognition method of non-stationary mechanical vibration signal based on convolution neural 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 |