CN110456732B - Punch press fault monitoring system with learning function - Google Patents
Punch press fault monitoring system with learning function Download PDFInfo
- Publication number
- CN110456732B CN110456732B CN201910721454.0A CN201910721454A CN110456732B CN 110456732 B CN110456732 B CN 110456732B CN 201910721454 A CN201910721454 A CN 201910721454A CN 110456732 B CN110456732 B CN 110456732B
- Authority
- CN
- China
- Prior art keywords
- module
- data
- unit
- sample
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
- G05B19/406—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by monitoring or safety
- G05B19/4065—Monitoring tool breakage, life or condition
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/37—Measurements
- G05B2219/37616—Use same monitoring tools to monitor tool and workpiece
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Manufacturing & Machinery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention relates to the technical field of punch fault detection, in particular to a punch fault monitoring system with a learning function, which comprises a punch unit and a sensing module, wherein a historical data storage module is connected with a data backup module, a network module is respectively connected with a sample marking module and a fault prediction module, the fault prediction module is simultaneously connected with a data processing module, the punch unit is connected with the fault prediction module through a remote operation module, a model training module is in signal connection with a model matrix parameter downloading module, and the model matrix parameter downloading module is connected with the fault prediction module through the network module. The invention reflects the abnormity of the punching process by collecting the stress signal of the punch in the punching process and identifying the abnormal change of the stress signal, accurately detects the punching fault, can rapidly solve the punching fault and improves the production efficiency.
Description
Technical Field
The invention relates to the technical field of punch fault monitoring, in particular to a punch fault monitoring system with a learning function.
Background
The mainstream solution in the industry is a graphic processing system for online stamping, the system is installed near a stamping machine through 1-4 cameras, after the stamping of a die, whether the surface of the die is normal or not is detected, and whether residual materials are left on the surface of the die or not, and the mode meets a lot of problems when applied on site, such as that when a die core is shielded, the machine cannot shoot all the general appearances of the surface of the die, and in addition, due to the problem of light source adjustment, a visual system easily causes misidentification and misinformation, thereby reducing the production efficiency.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides a punch press fault monitoring system with a learning function.
In order to achieve the purpose, the invention adopts the following technical scheme:
a punch press fault monitoring system with a learning function is designed, which comprises a punch press unit and a sensing module, wherein the punch press unit is in signal connection with a data processing unit through the sensing module, the data processing unit is respectively connected with a machine learning model design module and a historical data storage module through a network module, the historical data storage module is connected with a data backup module, the network module is respectively connected with a sample marking module and a fault prediction module, the fault prediction module is simultaneously connected with the data processing module, the punch press unit is connected with the fault prediction module through a remote control module, the machine learning model design module comprises a time sequence point input unit and a neural network model frame, the neural network model frame is in signal connection with a Sigmoid binary output unit, and the Sigmoid binary output unit is in signal connection with a model training module, the historical data storage module is connected with the model training module through the time sequence point input unit, the model training module is in signal connection with the model matrix parameter downloading module, and the model matrix parameter downloading module is connected with the fault prediction module through the network module.
Preferably, the data processing unit comprises a data receiver, the data is divided by the data dividing unit, the data is transmitted to the fault prediction module and the data sample sending module after being divided, the connection with the network module is realized by the data sample sending module, and meanwhile, the network module is in signal connection with the fault prediction module, so that the model is downloaded and replaced.
Preferably, the fault prediction module comprises a prediction model matrix parameter receiving module, the network module is connected with the prediction model matrix module through the prediction model matrix parameter receiving module, the data processing unit is connected with the prediction model matrix module to realize the collection and processing of real-time data samples, the prediction model directly outputs a prediction result, and when data fails, the control of the punch unit is realized through the remote control module.
Preferably, the sample marking module monitors the field HMI operation screen through a marking worker, generates a result and sample association module after data recording, sends the sample to the network module through the sample sending module, and sends the sample to the historical data storage module for backup and storage through the network module.
Preferably, the remote control module comprises a model output logic conversion module, an output signal electrical conversion module and an electrical connection module, the model output logic conversion module receives the signal sent by the fault prediction module, and the model output logic conversion module converts the signal through the output signal electrical conversion module and is in signal connection with the punch press unit through the electrical connection module, so that remote control of the punch press unit can be realized.
Preferably, the machine learning model design module needs to ensure that the total recorded stamping times reach more than 1000 times when collecting sample data, and the proportion of the fault state data to the normal state data in the machine learning model design module accounts for about 50% respectively.
The punch press fault monitoring system with the learning function has the beneficial effects that: this punch press fault monitoring system with learning function reflects the unusual of punching press process through the unusual change of discernment stress signal through the stress signal's of collection punch press in the punching press process, and is accurate to the detection of punching press trouble, and solution punching press trouble that simultaneously can be quick has improved the efficiency of production, and this system can arrange in order the data sample of gathering simultaneously and make the data model of this punch press when the punching press.
Drawings
Fig. 1 is a system block diagram of a punch press fault monitoring system with a learning function according to the present invention.
Fig. 2 is a system block diagram of a data processing unit of a punch press fault monitoring system with a learning function according to the present invention.
Fig. 3 is a system block diagram of a fault prediction module of a punch fault monitoring system with a learning function according to the present invention.
Fig. 4 is a system block diagram of a sample marking module of a punch fault monitoring system with a learning function according to the present invention.
Fig. 5 is a system block diagram of a remote control module of a punch press fault monitoring system with a learning function according to the present invention.
Fig. 6 is a system block diagram of a machine learning model design unit of a punch press fault monitoring system with a learning function according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
Referring to fig. 1-6, a punch press failure monitoring system with learning function comprises a punch press unit and a sensing module, the punch press unit is in signal connection with a data processing unit through the sensing module, the data processing unit is respectively connected with a machine learning model design module and a historical data storage module through a network module, the historical data storage module is connected with a data backup module, the network module is respectively connected with a sample marking module and a failure prediction module, the failure prediction module is simultaneously connected with the data processing module, the punch press unit is connected with the failure prediction module through a remote control module, the machine learning model design module comprises a time sequence point input unit and a neural network model frame, the neural network model frame is in signal connection with a Sigmoid binary output unit, and the Sigmoid binary output unit is in signal connection with a model training module, the historical data storage module is connected with the model training module through the time sequence point input unit, the model training module is in signal connection with the model matrix parameter downloading module, the model matrix parameter downloading module is connected with the fault prediction module through the network module, the data segmentation and sample marking module can intercept collected data according to a fixed interception period to form a stamping process data sample and mark the sample, the sample only contains two marking results of fault and normal, the sample is marked and accumulated to a certain number, the prediction model can be designed and trained through the machine learning design unit to obtain a model matrix capable of normally predicting the fault state, and when a stamping new product or a die is replaced, the marking and training process is repeated to obtain a new fault prediction model.
The fault prediction module comprises a prediction model matrix parameter receiving module, the network module is connected with the prediction model matrix module through the prediction model matrix parameter receiving module, meanwhile, the data processing unit is connected with the prediction model matrix module, the real-time data sample collection and processing are realized, when data fails, the control of the punch press unit is realized through the remote control module, the punching force of the punch press during working is detected, foreign matters exist in the punch press and the foreign matters do not exist in the punch press, the generated punching curve characteristic difference is large, and the judgment of whether the punch press is in a normal state or not can be effectively monitored.
The failure prediction module comprises a prediction model matrix parameter receiving module, the network module is connected with the prediction model matrix module through the prediction model matrix parameter receiving module, the data processing unit is connected with the prediction model matrix module to realize the collection and processing of real-time data samples, the prediction model matrix module can output prediction results according to the real-time sample data, when data fails, the punch press unit is controlled through the remote control module, the data is calculated through the stack self-coding algorithm unit, the accuracy of the calculation results of the data can reach more than 99%, the collected data can be rapidly calculated and judged, and the monitoring on the process quality of punching of the punching machine is improved.
Output module includes high-speed IO command unit, the model design unit is after comparing data, when data is the fault data, then can pass through high-speed IO command unit with signal transmission to signal converter on, signal converter can pass through the mode transmission of order to the control system of punching press unit after with data processing simultaneously, make control system carry out scram to the punching press switch, when data is normal data, the transmission of signal can't be carried out to the control system of punching press unit inside to the command that produces, through this design, the model design unit can be in the control system that 0.1S will detect the result through high-speed IO command unit transmission to the punch press, thereby can realize the quick shut down to the punch press.
The remote control module comprises a model output logic conversion module, an output signal electrical conversion module and an electrical connection module, the model output logic conversion module receives signals sent by the fault prediction module, and the output signal electrical conversion module converts the signals and is in signal connection with the punch unit through the electrical connection module, so that remote control of the punch unit can be realized.
The machine learning model design module needs to ensure that the total stamping times of record reaches more than 1000 times when collecting sample data, meanwhile, the proportion of fault state data and normal state data accounts for 50 percent respectively, and the statistics of stamping data samples for more than 1000 times can comprehensively show the difference of two kinds of sample data in the stamping die in the states of foreign matters and normal states, so that whether stamping is normal can be judged according to real-time stamping data, the production efficiency is improved, and the cost of manual inspection is reduced.
The sample marking module monitors the field HMI operation screen through a marking worker, generates a result and sample association module after data recording, sends the sample to the network module through the sample sending module, and simultaneously sends the sample to the historical data storage module for backup and storage, backups the data, and can effectively prevent the data loss from causing the whole system to be paralyzed through backups and storage of the data, thereby improving the strength of the system.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (6)
1. A punch press fault monitoring system with a learning function comprises a punch press unit and a sensing module, and is characterized in that the punch press unit is in signal connection with a data processing unit through the sensing module, the data processing unit is respectively connected with a machine learning model design module and a historical data storage module through a network module, the historical data storage module is connected with a data backup module, the network module is respectively connected with a sample marking module and a fault prediction module, the fault prediction module is simultaneously connected with the data processing module, the punch press unit is connected with the fault prediction module through a remote control module, the machine learning model design module comprises a time sequence point input unit and a neural network model frame, the neural network model frame is in signal connection with a Sigmoid binary output unit, and the Sigmoid binary output unit is in signal connection with a model training module, the historical data storage module is connected with the model training module through the time sequence point input unit, the model training module is in signal connection with the model matrix parameter downloading module, and the model matrix parameter downloading module is connected with the fault prediction module through the network module.
2. The system of claim 1, wherein the data processing unit comprises a data receiver, the data is divided by the data dividing unit, the divided data is transmitted to the failure prediction module and the data sample sending module, the data sample sending module is connected to the network module, the network module and the failure prediction module are in signal connection with each other, so as to download and replace the model, the data dividing and marking module can intercept the collected data according to a fixed interception period, form the data sample of the stamping process and mark the sample, and the historical data storage unit can store the marked sample data.
3. The system of claim 1, wherein the failure prediction module comprises a prediction model matrix parameter receiving module, the network module is connected to the prediction model matrix module through the prediction model matrix parameter receiving module, the data processing unit is connected to the prediction model matrix module to collect and process real-time data samples, and the control of the punch unit is realized through the remote control module when data monitored is abnormal.
4. The system for monitoring the faults of the punch press with the learning function as claimed in claim 1, wherein the sample marking module comprises a module for obtaining sample marking information from a field HMI operation screen through a marking worker, generating a result and associating the result with the sample after data recording, and sending the sample to a historical data storage module for backup and storage through a network module by a sample sending module.
5. The system of claim 1, wherein the remote control module comprises a model output logic conversion module, an output signal electrical conversion module and an electrical connection module, the model output logic conversion module converts the signal received from the failure prediction module through the output signal electrical conversion, and is connected with the punching unit through the electrical connection module, so as to realize remote control of the punching unit.
6. The system of claim 1, wherein the machine learning model design module is configured to ensure that the total punching frequency of the record is greater than 1000 times when collecting the sample data, and the ratio of the fault status data to the normal status data is about 50%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910721454.0A CN110456732B (en) | 2019-08-06 | 2019-08-06 | Punch press fault monitoring system with learning function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910721454.0A CN110456732B (en) | 2019-08-06 | 2019-08-06 | Punch press fault monitoring system with learning function |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110456732A CN110456732A (en) | 2019-11-15 |
| CN110456732B true CN110456732B (en) | 2021-09-28 |
Family
ID=68485060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910721454.0A Active CN110456732B (en) | 2019-08-06 | 2019-08-06 | Punch press fault monitoring system with learning function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110456732B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116428984A (en) * | 2023-05-24 | 2023-07-14 | 湖南涛淼实业有限公司 | Hardware mould stamping processing intelligent detection system |
| CN117261343B (en) * | 2023-11-21 | 2024-02-09 | 山东迪格重工机械有限公司 | Punch press fault monitoring system based on thing networking |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1878623A (en) * | 2003-11-11 | 2006-12-13 | 新日本制铁株式会社 | Press forming device, press forming method, computer program, and recording medium |
| CN101859128A (en) * | 2010-07-05 | 2010-10-13 | 北京信息科技大学 | Knowledge-based fault prediction expert system for complex milling machine tool |
| CN104573237A (en) * | 2015-01-08 | 2015-04-29 | 湖南大学 | Frictional wear CAE (Computer Aided Engineering) analysis-based mold optimization method |
| CN104616033A (en) * | 2015-02-13 | 2015-05-13 | 重庆大学 | Fault diagnosis method for rolling bearing based on deep learning and SVM (Support Vector Machine) |
| KR20180014364A (en) * | 2016-07-29 | 2018-02-08 | 한국전자통신연구원 | Apparatus and method for tool wear compensation of cnc |
| KR20180048218A (en) * | 2016-10-31 | 2018-05-10 | 현대위아 주식회사 | Machine tool breakdown diagnosis system based on Machine Learning, and method thereof |
| CN108107838A (en) * | 2017-12-27 | 2018-06-01 | 山东大学 | A kind of numerical control equipment tool wear monitoring method based on cloud knowledge base and machine learning |
| CN108227625A (en) * | 2016-12-22 | 2018-06-29 | 发那科株式会社 | Life tools apparatus for predicting |
| CN108334033A (en) * | 2018-02-28 | 2018-07-27 | 中国科学院重庆绿色智能技术研究院 | Punching machine group failure prediction method and its system based on Internet of Things and machine learning |
| CN108803552A (en) * | 2018-08-31 | 2018-11-13 | 承德建龙特殊钢有限公司 | A kind of the monitoring system and monitoring method of equipment fault |
-
2019
- 2019-08-06 CN CN201910721454.0A patent/CN110456732B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1878623A (en) * | 2003-11-11 | 2006-12-13 | 新日本制铁株式会社 | Press forming device, press forming method, computer program, and recording medium |
| CN101859128A (en) * | 2010-07-05 | 2010-10-13 | 北京信息科技大学 | Knowledge-based fault prediction expert system for complex milling machine tool |
| CN104573237A (en) * | 2015-01-08 | 2015-04-29 | 湖南大学 | Frictional wear CAE (Computer Aided Engineering) analysis-based mold optimization method |
| CN104616033A (en) * | 2015-02-13 | 2015-05-13 | 重庆大学 | Fault diagnosis method for rolling bearing based on deep learning and SVM (Support Vector Machine) |
| KR20180014364A (en) * | 2016-07-29 | 2018-02-08 | 한국전자통신연구원 | Apparatus and method for tool wear compensation of cnc |
| KR20180048218A (en) * | 2016-10-31 | 2018-05-10 | 현대위아 주식회사 | Machine tool breakdown diagnosis system based on Machine Learning, and method thereof |
| CN108227625A (en) * | 2016-12-22 | 2018-06-29 | 发那科株式会社 | Life tools apparatus for predicting |
| CN108107838A (en) * | 2017-12-27 | 2018-06-01 | 山东大学 | A kind of numerical control equipment tool wear monitoring method based on cloud knowledge base and machine learning |
| CN108334033A (en) * | 2018-02-28 | 2018-07-27 | 中国科学院重庆绿色智能技术研究院 | Punching machine group failure prediction method and its system based on Internet of Things and machine learning |
| CN108803552A (en) * | 2018-08-31 | 2018-11-13 | 承德建龙特殊钢有限公司 | A kind of the monitoring system and monitoring method of equipment fault |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110456732A (en) | 2019-11-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110366031B (en) | Vision-based abnormal state monitoring and fault diagnosis method for MES (manufacturing execution system) of digital workshop | |
| CN110456732B (en) | Punch press fault monitoring system with learning function | |
| CN103676868B (en) | A kind of FPC manufactures critical process and automatically monitors and intelligent analysis system | |
| CN112925279A (en) | Fault comprehensive analysis system based on MES system | |
| CN112173636B (en) | Method for detecting faults of belt conveyor carrier roller by inspection robot | |
| CN109141625B (en) | Online state monitoring method for ball screw pair | |
| CN112798042A (en) | Intelligent diagnosis method for working state and supporting quality of hydraulic support | |
| CN111666978B (en) | Intelligent fault early warning system for IT system operation and maintenance big data | |
| CN116880309A (en) | Factory environment monitoring management system and method | |
| CN109051104A (en) | A kind of bottle placer operating status intelligent Detection and method | |
| CN116244765A (en) | Equipment maintenance management method based on industrial Internet | |
| CN111404756A (en) | Fault diagnosis system for communication equipment | |
| CN116880402A (en) | Intelligent factory cooperative control system and method thereof | |
| CN103203325A (en) | System and method of online vision detection | |
| CN117142038A (en) | Belt conveyor conveying method | |
| CN117373478A (en) | Voiceprint recognition system for transformer | |
| CN116796261A (en) | Closed switch equipment mechanical characteristic prediction method based on artificial intelligence | |
| CN111733498A (en) | Intelligent spinning spun yarn fault detection system | |
| CN116245506A (en) | Equipment operation and maintenance management system, fault self-detection method and using method thereof | |
| CN115356970A (en) | Dynamic loop monitoring system and method based on real-time data and operation and maintenance data | |
| CN110096383B (en) | Automatic classification method for maintenance information of signal equipment | |
| KR102306430B1 (en) | Industrial Process Equipment System Management Method using Cloud-based Intergrated Context-Aware Processing Engine | |
| CN112215503A (en) | Reliability monitoring method based on SPC | |
| CN117966313B (en) | Textile equipment monitoring system and method | |
| CN113762913B (en) | User account real-time monitoring 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |