CN106055886B - The method for building up of sealing ring decrement detection model based on data-driven - Google Patents
The method for building up of sealing ring decrement detection model based on data-driven Download PDFInfo
- Publication number
- CN106055886B CN106055886B CN201610356721.5A CN201610356721A CN106055886B CN 106055886 B CN106055886 B CN 106055886B CN 201610356721 A CN201610356721 A CN 201610356721A CN 106055886 B CN106055886 B CN 106055886B
- Authority
- CN
- China
- Prior art keywords
- torque
- value
- data
- corner
- zero
- 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
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
Abstract
The method for building up of the present invention relates to a kind of sealing ring decrement detection model based on data-driven, includes the following steps：1) relevant parameter and data of the history corner value and torque value and operational process of record cylinder rundown process；2) recorded data is filtered, and data are converted to rotation angle of torque relational graph by treated；3) data normalization processing is carried out to the rotation angle of torque relational graph of transformed different cylinder types；4) single order and second dervative that rear torque and angle relation is normalized are sought；5) the zero feature of sealing ring compression process in cylinder rundown process is analyzed, establish the discriminant function of zero, and machine training and study are carried out according to the single order and second derivative values of rotation angle of torque relationship and the torque and angle relation that are calculated, determine the discriminant function of zero；6) on the basis of determining zero discriminant function, according to torque value and corner value and relevant parameter, the detection model of sealing ring decrement is established.
Description
Technical field
The invention belongs to sealing ring decrement detection technique fields, are related to a kind of sealing ring decrement based on data-driven
The method for building up of detection model.
Background technology
Production process is tightened in cylinder, the technology of tightening is to ensure that it tightens a key technology of quality.And in cylinder
In the rundown process of body, sealing ring is the key that ensure leakproofness auxiliary original paper, plays very important effect.However sealing ring
As a kind of elastic material, sealing performance is strictly constrained by its effectively sealing section, in rundown process, if close
The deflection of seal exceeds its plastic deformation range, can directly result in ring structure and be destroyed, so as to cause leakage, sealing
Failure；If deformation requirement is not achieved in the deflection of sealing ring, and can cause accessory surface and sealing ring insufficient contact, to
The problem of seal request is not achieved, seal failure and component joint surface is caused to loosen.Therefore, right in the rundown process of cylinder
It is played a key role in the good control of sealed ring deformation amount, it is extremely important.
In order to preferably meet the needs of control, the control accuracy of sealed ring deformation amount is improved, so establishing sealing ring change
Shape measurement model tightens quality to offer cylinder and is very helpful.
Invention content
In view of this, the purpose of the present invention is to provide a kind of sealing ring decrement detection model based on data-driven
Method for building up, the history torque and corner value of cylinder rundown process of this method using record in the database, and to being recorded
Data carry out pretreatment and normalized seek the single order and second order of torque-corner after the completion of to data processing
Derivative analyzes the feature of sealing ring rundown process zero, establishes the discriminant function of zero, and root according to the data after handled
Machine training and study are carried out according to historical data, determines the parameter of its discriminant function；On the basis of determining zero, according to torsion
Square and corner value and relevant parameter, so that it is determined that going out the detection model of sealing ring decrement.The accurate meter of sealing ring decrement
It calculates and provides foundation for the reliable control of cylinder.
In order to achieve the above objectives, the present invention provides the following technical solutions：
A kind of method for building up of the sealing ring decrement detection model based on data-driven, this approach includes the following steps：
1) relevant parameter and data of the history corner value and torque value and operational process of record cylinder rundown process；
2) to the data processings such as the corner value and torque value that are recorded are filtered, and will treated that data are converted to torsion
Square-angle relation figure；
3) further data normalization processing is carried out to the torque-corner relational graph of transformed different cylinder types；
4) single order and second dervative that rear torque and angle relation is normalized are sought；
5) the zero feature for analyzing sealing ring compression process in cylinder rundown process, establishes the discriminant function of zero, and root
Machine training is carried out according to the single order and second derivative values of torque-corner relationship and the torque and angle relation that are calculated and is learned
It practises, determines the discriminant function of zero；
6) it on the basis of determining zero discriminant function, according to torque value and corner value and relevant parameter, establishes out close
The detection model of seal decrement.
Further, in step 1), the history corner value and torque value of the record cylinder rundown process, and operation
The relevant parameter of process and the method for data are：The corner value of operation overall process is tightened to cylinder using data-base recording method
A and torque value T carry out real-time and synchronous recording, and its data is stored respectively, while to other ginsengs in operational process
Number is also recorded and is stored, including the diameter R of the type cylinder, is tightened maximum and born torque Tmax。
Further, in step 2), the corner value and torque value to being recorded is filtered equal data processings
Method is：The effective torque value and corner value of storage are filtered according to the following formula：
Wherein：T′nWith A 'nFor the torque value and corner value after being filtered；
TnAnd AnThe effective torque value and corner value recorded by database；
The serial number for the torque and corner virtual value that n is recorded by database.
Further, in step 2), described by treated, method that data are converted to torque-corner relational graph is：
Torque value T ' after being filterednWith corner value A 'nAccording to the principle and corner value A ' of corner value roundingnChange rate be 1
Requirement, to torque value T 'nWith corner value A 'nSorted out and screened, is converted to the relational graph of torque-corner.
Further, in step 3), the torque-corner relational graph to transformed different cylinder types is into traveling
One step data normalization processing method be：By treated, torsion curve figure carries out integral transformation, after making transformation
Curve graph is seated under same coordinate, and tighten terminate torque value range it is roughly the same, transforming function transformation function is：
Wherein：TsetFor standard curve target torque value after conversion；
TmaxMaximum is tightened for cylinder to bear torque or be target torque；
K is standard curve target torque TsetWith the type cylinder target torque TmaxRatio；
A is constant；
Screw-down torque normalized function can be then determined as：
Wherein：Y is the torque value after being normalized.
Further, the single order and second dervative for seeking being normalized rear torque and angle relation described in step 4)
Method be：The single order and second order for acquiring torque-corner using difference formula derivation to the torque Value Data after normalization are led
Number, is shown below：
Further, in step 5), the zero feature of sealing ring compression process, root in the analysis cylinder rundown process
Machine training is carried out according to the single order and second derivative values of torque-corner relationship and the torque and angle relation that are calculated and is learned
It practises, the method for determining the discriminant function of zero is：Physics and data characteristics first to sealing ring rundown process zero carry out
Analysis, and establish discriminant vector using torsion curve figure and required single order and second derivative values
Then, the discriminant function of zero is established according to discriminant vector：
On the basis of the zero discriminant function of foundation, according to the principle of error amount minimum, recorded in database
Torque value and corner Value Data to discriminant functionIt is trained, so that it is determined that going out corresponding solution vector
Finally, it can will meet angle value corresponding to the torque value of zero and position zero angle Aset。
Further, described on the basis of determining zero discriminant function in step 6), according to torque value and corner value
And relevant parameter, the method for establishing out the detection model of sealing ring decrement are：According to identified zero point angle Aset, with
And corresponding angle value A, the decrement of corresponding sealing ring can be found out according to the following formula：
Wherein：For the decrement of sealing ring.
The beneficial effects of the present invention are：1) according to the system operation torque and corner value recorded, sealing ring change is established
The detection model of shape amount realizes the detectability of sealed ring deformation amount；2) according to the detection mould for the sealed ring deformation amount established
The online and real-time detection to sealed ring deformation amount in cylinder rundown process may be implemented in type；3) sealed ring deformation amount is established
Detection model, provide reliable support to improve the control accuracy tightened of cylinder and quality.
Description of the drawings
In order to keep the purpose of the present invention, technical solution and advantageous effect clearer, the present invention provides following attached drawing and carries out
Explanation：
Fig. 1 is the flow diagram of the method for the invention；
Fig. 2 is the relational graph of torque-corner.
Specific implementation mode
Below in conjunction with attached drawing, the preferred embodiment of the present invention is described in detail.
The present invention first records the historical data in cylinder rundown process, and is pre-processed to recorded data
And normalized, will treated that data are converted to torque-corner relational graph, seek torque-corner relationship single order and
Second dervative, while according to the data after handled, analyzing the feature of sealing ring rundown process zero, establishing the differentiation letter of zero
Number；On the basis of determining zero, according to torque and corner value and relevant parameter, the detection mould of sealing ring decrement is determined
Type.
Fig. 1 is the flow diagram of the method for the invention, as shown, the inspection of the sealing ring decrement based on data-driven
The method for building up for surveying model is as follows shown：
1) relevant parameter and data of the history corner value and torque value and operational process of record cylinder rundown process；
2) to the data processings such as the corner value and torque value that are recorded are filtered, and will treated that data are converted to torsion
Square-angle relation figure；
3) further data normalization processing is carried out to the torque-corner relational graph of transformed different cylinder types；
4) single order and second dervative that rear torque and angle relation is normalized are sought；
5) the zero feature for analyzing sealing ring compression process in cylinder rundown process, establishes the discriminant function of zero, and root
Machine training is carried out according to the single order and second derivative values of torque-corner relationship and the torque and angle relation that are calculated and is learned
It practises, determines the discriminant function of zero；
6) it on the basis of determining zero discriminant function, according to torque value and corner value and relevant parameter, establishes out close
The detection model of seal decrement.
Wherein, the history corner value and torque value and operational process of cylinder rundown process are recorded described in step 1)
The method of relevant parameter and data is：
Cylinder is tightened using data-base recording method operation overall process corner value A and torque value T carry out in real time and
Synchronous recording, and its data is stored respectively, while the other parameters in operational process are also recorded and stored, example
As the type cylinder diameter R, tighten maximum and bear torque TmaxDeng；
Wherein, it is to the method for the data processings such as the corner value and torque value that are recorded are filtered described in step 2)：
Storage effective torque value in the database and corner value are filtered according to the following formula：
Wherein：T′nWith A 'nFor the torque value and corner value after being filtered；
TnAnd AnThe effective torque value and corner value recorded by database；
The serial number for the torque and corner virtual value that n is recorded by database；
Wherein, by treated, method that data are converted to torque-corner relational graph is described in step 2)：
Torque value T ' after being filterednWith corner value A 'nAccording to the principle and corner value A ' of corner value roundingnChange
The requirement that rate is 1, to torque value T 'nWith corner value A 'nSorted out and screened, is converted to the relational graph of torque-corner, such as
Shown in Fig. 2.
Wherein, the torque-corner relational graph of transformed different cylinder types is carried out described in step 3) further
Data normalization processing method be：
By treated, torsion curve figure carries out integral transformation, and the curve graph after transformation is made to be seated same coordinate
Under, and tighten terminate torque value range it is roughly the same, transforming function transformation function is：
Wherein：TsetFor standard curve target torque value after conversion；
TmaxMaximum is tightened for cylinder to bear torque or be target torque；
K is standard curve target torque TsetWith the type cylinder target torque TmaxRatio；
A is constant；
Screw-down torque normalized function can be then determined as：
Wherein：Y is the torque value after being normalized；
Wherein, the single order that rear torque and angle relation is normalized and second dervative are sought described in step 4)
Method is：
The single order and second order for acquiring torque-corner using difference formula derivation to the torque Value Data after normalization are led
Number, is shown below：
Wherein, the zero feature that sealing ring compression process in cylinder rundown process is analyzed described in step 5), according to torque-
The single order and second derivative values of angle relation and the torque and angle relation that are calculated carry out machine training and study, determine
The method of the discriminant function of zero is：
The physics and data characteristics of sealing ring rundown process zero are analyzed first, and bent using torque-corner
Line chart and required single order and second derivative values establish discriminant vector
Then, the discriminant function of zero is established according to discriminant vector：
On the basis of the zero discriminant function of foundation, according to the principle of error amount minimum, recorded in database
Torque value and corner Value Data to discriminant functionIt is trained, so that it is determined that going out corresponding solution vector
Finally, it can will meet angle value corresponding to the torque value of zero and position zero angle Aset；
Wherein, described in step 6) on the basis of determining zero discriminant function, according to torque value and corner value and phase
Related parameter, the method for establishing out the detection model of sealing ring decrement are：
According to identified zero point angle Aset, and corresponding angle value A, corresponding sealing ring can be found out according to the following formula
Decrement：
Wherein：For the decrement of sealing ring.
Finally illustrate, preferred embodiment above is merely illustrative of the technical solution of the present invention and unrestricted, although logical
It crosses above preferred embodiment the present invention is described in detail, however, those skilled in the art should understand that, can be
Various changes are made to it in form and in details, without departing from claims of the present invention limited range.
Claims (5)
1. a kind of method for building up of the sealing ring decrement detection model based on data-driven, it is characterised in that：This method includes
Following steps：
1) relevant parameter and data of the history corner value and torque value and operational process of record cylinder rundown process；
2) data processing is filtered to the corner value and torque value that are recorded, and data are converted to torque-turn by treated
Angular dependence figure；
3) further data normalization processing is carried out to the torque-corner relational graph of transformed different cylinder types；
4) single order and second dervative that rear torque and angle relation is normalized are sought；
5) the zero feature for analyzing sealing ring compression process in cylinder rundown process, establishes the discriminant function of zero, and according to torsion
The single order and second derivative values of square-angle relation and the torque and angle relation that are calculated carry out machine training and study, really
Make the discriminant function of zero；
6) on the basis of determining zero discriminant function, according to torque value and corner value and relevant parameter, sealing ring is established out
The detection model of decrement；
In step 2), the method that the corner value and torque value to being recorded is filtered data processing is：It will storage
Effective torque value and corner value be filtered according to the following formula：
Wherein：T′nWith A 'nFor the torque value and corner value after being filtered；
TnAnd AnThe effective torque value and corner value recorded by database；
The serial number for the torque and corner virtual value that n is recorded by database；
In step 3), the torque-corner relational graph to transformed different cylinder types carries out further data
The method of normalized is：By treated, torsion curve figure carries out integral transformation, and the curve graph after transformation is made to be located
Under same coordinate, and tighten terminate torque value range it is roughly the same, transforming function transformation function is：
Wherein：TsetFor standard curve target torque value after conversion；
TmaxMaximum is tightened for cylinder to bear torque or be target torque；
K is standard curve target torque TsetWith the type cylinder target torque TmaxRatio；
A is constant；
Screw-down torque normalized function can be then determined as：
Wherein：Y is the torque value after being normalized；
In step 5), the zero feature of sealing ring compression process in the analysis cylinder rundown process, according to torque-corner
The single order and second derivative values of relationship and the torque and angle relation that are calculated carry out machine training and study, determine zero
The method of discriminant function be：The physics and data characteristics of sealing ring rundown process zero are analyzed first, and utilized
Torsion curve figure and required single order and second derivative values establish discriminant vector
Then, the discriminant function of zero is established according to discriminant vector：
On the basis of the zero discriminant function of foundation, according to the principle of error amount minimum, the torsion recorded in database is utilized
Square value and corner Value Data are to discriminant functionIt is trained, so that it is determined that going out corresponding solution vectorMost
Afterwards, it can will meet angle value corresponding to the torque value of zero and position zero angle Aset。
2. a kind of method for building up of sealing ring decrement detection model based on data-driven according to claim 1,
It is characterized in that：In step 1), the history corner value and torque value of the described record cylinder rundown process and operational process
The method of relevant parameter and data is：Corner value A and the torsion of operation overall process are tightened to cylinder using data-base recording method
Square value T is carried out in real time and synchronous recording, and its data is stored respectively, at the same to the other parameters in operational process also into
Row record and storage, including the diameter R of the type cylinder, tighten maximum and bear torque Tmax。
3. a kind of method for building up of sealing ring decrement detection model based on data-driven according to claim 1,
It is characterized in that：In step 2), described by treated, method that data are converted to torque-corner relational graph is：It will filtering
Treated torque value T 'nWith corner value A 'nAccording to the principle and corner value A ' of corner value roundingnChange rate be 1 requirement,
To torque value T 'nWith corner value A 'nSorted out and screened, is converted to the relational graph of torque-corner.
4. a kind of method for building up of sealing ring decrement detection model based on data-driven according to claim 1,
It is characterized in that：The method of the single order for seeking being normalized rear torque and angle relation and second dervative described in step 4)
For：The single order and second dervative of torque-corner are acquired using difference formula derivation to the torque Value Data after normalization, it is as follows
Shown in formula：
5. a kind of method for building up of sealing ring decrement detection model based on data-driven according to claim 1,
It is characterized in that：It is described on the basis of determining zero discriminant function in step 6), according to torque value and corner value and phase
Related parameter, the method for establishing out the detection model of sealing ring decrement are：According to identified zero point angle Aset, and it is corresponding
Angle value A, the decrement of corresponding sealing ring can be found out according to the following formula：
Wherein：For the decrement of sealing ring.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610356721.5A CN106055886B (en) | 2016-05-25 | 2016-05-25 | The method for building up of sealing ring decrement detection model based on data-driven |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610356721.5A CN106055886B (en) | 2016-05-25 | 2016-05-25 | The method for building up of sealing ring decrement detection model based on data-driven |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106055886A CN106055886A (en) | 2016-10-26 |
| CN106055886B true CN106055886B (en) | 2018-08-17 |
Family
ID=57174497
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610356721.5A Active CN106055886B (en) | 2016-05-25 | 2016-05-25 | The method for building up of sealing ring decrement detection model based on data-driven |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106055886B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110633514B (en) * | 2019-08-28 | 2022-09-27 | 南京理工大学 | Casing pipe screwing control method based on thread interference magnitude |
| CN114812909B (en) * | 2022-03-29 | 2024-03-22 | 东风柳州汽车有限公司 | Method, device, equipment and storage medium for determining battery bolt torque |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103949876A (en) * | 2014-05-15 | 2014-07-30 | 重庆大学 | Elastic body flexible screwing device |
| CN103983390A (en) * | 2014-05-09 | 2014-08-13 | 南京理工大学 | Measuring device for mechanical parameters of powder rotating extruding granulation forming |
| CN104792508A (en) * | 2015-04-15 | 2015-07-22 | 中国航天科技集团公司川南机械厂 | Online compression amount measurement method for seal ring of small solid rocket |
| CN104807432A (en) * | 2015-05-19 | 2015-07-29 | 重庆大学 | Soft measuring method of axial deformation quantity of rectangular rubber sealing ring based on flexible screwing assembly |
-
2016
- 2016-05-25 CN CN201610356721.5A patent/CN106055886B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103983390A (en) * | 2014-05-09 | 2014-08-13 | 南京理工大学 | Measuring device for mechanical parameters of powder rotating extruding granulation forming |
| CN103949876A (en) * | 2014-05-15 | 2014-07-30 | 重庆大学 | Elastic body flexible screwing device |
| CN104792508A (en) * | 2015-04-15 | 2015-07-22 | 中国航天科技集团公司川南机械厂 | Online compression amount measurement method for seal ring of small solid rocket |
| CN104807432A (en) * | 2015-05-19 | 2015-07-29 | 重庆大学 | Soft measuring method of axial deformation quantity of rectangular rubber sealing ring based on flexible screwing assembly |
Non-Patent Citations (1)
| Title |
|---|
| 一种密封圈压缩量在线测量方法;胡韶华等;《制造技术研究》;20150430(第2期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106055886A (en) | 2016-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105740899A (en) | Machine vision image characteristic point detection and matching combination optimization method | |
| CN106055886B (en) | The method for building up of sealing ring decrement detection model based on data-driven | |
| CN110906875B (en) | Visual processing method for aperture measurement | |
| CN109159113B (en) | Robot operation method based on visual reasoning | |
| CN107798326A (en) | A kind of profile visual detection algorithm | |
| CN106371315B (en) | A kind of threaded connector tightening control method based on Compress rate of sealing ring | |
| CN113570631B (en) | Image-based pointer instrument intelligent identification method and device | |
| CN109461135A (en) | The processing method of gluing picture | |
| CN107065559B (en) | A kind of industrial robot increment self-adaptation control method | |
| CN108734722B (en) | Visual tracking error correction method based on PSR | |
| CN110487229A (en) | Circumference equal-dividing hole minimum position degree calculation method | |
| WO2021258498A1 (en) | Ellipse image detection algorithm and system based on contour integration | |
| CN111684462A (en) | Image matching method and visual system | |
| CN112986977A (en) | Method for overcoming radar extended Kalman track filtering divergence | |
| CN105867446B (en) | A kind of cylinder sealing tightening control method based on sealing ring decrement | |
| CN109325487B (en) | Full-category license plate recognition method based on target detection | |
| CN108805909B (en) | Target tracking method based on particle filter redetection under related filter framework | |
| CN110837782A (en) | Method for identifying fracture information according to material stretching process monitoring video | |
| CN101625763A (en) | Method for measuring similarity of spatial color histogram | |
| CN113674323A (en) | Visual target tracking algorithm based on multidimensional confidence evaluation learning | |
| CN113642204A (en) | Method for correcting combustion starting point identification deviation based on combustion excitation contribution degree | |
| US20070185592A1 (en) | Systems and methods for detecting suction valve closure | |
| Morante et al. | Automatic demonstration and feature selection for robot learning | |
| CN105930840A (en) | Medicine bar code defect detection method based on statistical analysis | |
| CN114518116A (en) | Visual navigation method based on tracking guide line |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20191203 Address after: Room 314-1, dantaihu building (Wuluo Science Park), No.9, Taihu East Road, Wuzhong District, Suzhou, Jiangsu Province Patentee after: Suzhou ruipengcheng Technology Co.,Ltd. Address before: 400044 Shapingba street, Shapingba District, Chongqing, No. 174 Patentee before: Chongqing University |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230508 Address after: No. 5, Pacific 1st Road, Xintang Town, Zengcheng District, Guangzhou City, Guangdong Province (factory building) Patentee after: Guangzhou Jingbang Sealing Technology Co.,Ltd. Address before: Room 314-1, dantaihu building (Wuluo Science Park), No.9, Taihu East Road, Wuzhong District, Suzhou City, Jiangsu Province Patentee before: Suzhou ruipengcheng Technology Co.,Ltd. |