CN109750152A - Dominant frequency for lathe welded machine tool bad folds vibration aging treatment method - Google Patents
Dominant frequency for lathe welded machine tool bad folds vibration aging treatment method Download PDFInfo
- Publication number
- CN109750152A CN109750152A CN201910231525.9A CN201910231525A CN109750152A CN 109750152 A CN109750152 A CN 109750152A CN 201910231525 A CN201910231525 A CN 201910231525A CN 109750152 A CN109750152 A CN 109750152A
- Authority
- CN
- China
- Prior art keywords
- vibration
- dominant frequency
- data
- machine tool
- frequency
- 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
Abstract
The invention discloses a kind of folded vibration aging treatment methods of dominant frequency for lathe welded machine tool bad, establish the welded machine tool bad Vibration Aging Process database, the production process database record sheet of ageing treatment parameter, timeliness quality, residual stress size, welded machine tool bad parameters of structural dimension is formed, expert database is constructed；It is then based on the Vibration Aging Process parameter of the expert database, the folded vibration ageing treatment of dominant frequency is carried out to welded machine tool bad.The invention has the advantages that system can generate master oscillator frequenc and vibration time according to the essential characteristic and ageing treatment basic technology demand of workpiece without carrying out vibrating live Frequency sweep experiments to welded machine tool bad；In conjunction with the online dynamic strain fitting evaluation and test static stress for obtaining timeliness workpiece in real time, and according to aging treatment process target, dynamic corrections vibration frequency improves oscillating aging and removes residual stress one-time success rate.
Description
Technical field
The present invention relates to welding structural elements to connect internal stress aging treatment method, is especially for lathe welded machine tool bad
The folded vibration aging treatment method of dominant frequency.
Background technique
Ageing treatment is to reduce one of lathe welding structural element internal stress, the key link for improving machine tool accuracy retentivity.
For the lathe welded machine tool bad of high-precision retentivity, high-quality, energy-saving and environmental protection, safe and reliable heat aging technique, master are studied and defined
The combination of frequency vibration aging technique and various ageing techniques, being to reduction welded machine tool bad residual stress and raising precision stability must
It wants.For example, first carrying out oscillating aging after lathe bed welding is completed, high peak stress is eliminated, answering in welded machine tool bad is made
Power is distributed more uniformization, guarantees the dimensional stability of welded machine tool bad, so that will not change under actual working conditions, most
Welded machine tool bad is finished again afterwards.By actual measurement, comparative study difference aging technique before and after the processing answer by the interior of lathe bed
The changing rule of power and dimensional accuracy gives full play to oscillating aging and increases dimensional stability, improve lathe bed precision stability, formed
High-precision retentivity cast-weld construction lathe bed combination aging technical method selects optimal vibration timeliness parameter to have high-grade, digitally controlled machine tools
There is important directive significance.
Currently, the technological parameter of traditional oscillating aging instrument mainly has exciting force, excited frequency, exciting time and support
Point, impacting point, to the formulation of the technological parameter, there is also biggish blindness, mainly sweeping by experience and vibration scene
Frequency is tested.Such as exciting time, support are determined as excited frequency, according to the size of workpiece weight using the frequency in subresonance area
Point is generally located on the lesser position of amplitude, in maximum place setting impacting point of amplitude etc..It can be seen that oscillating aging is joined
Several choosing all has centainly empirical, and practical application difficulty is big, therefore strict demand is just proposed to device operator, behaviour
Author must be veteran professional.In oscillating aging application, residual stress eradicating efficacy often occurs and pays no attention to
Think, to find out its cause, it is its main cause that Parameters of Vibration Aging, which is formulated improper,.
Summary of the invention
It is an object of that present invention to provide a kind of folded vibration aging treatment methods of dominant frequency for lathe welded machine tool bad.
To achieve the above object, the present invention takes following technical proposals:
The folded vibration aging treatment method of dominant frequency of the present invention for lathe welded machine tool bad, establishes the welded machine tool bad oscillating aging
Technological data bank forms the production of ageing treatment parameter, timeliness quality, residual stress size, welded machine tool bad parameters of structural dimension
Process database record sheet constructs expert database；It is then based on the Vibration Aging Process parameter of the expert database, butt welding
It connects lathe bed and carries out the folded vibration ageing treatment of dominant frequency, specific steps are as follows:
The first step establishes workpiece (welded machine tool bad) feature, the technological parameter of oscillating aging and residual stress quality assessment index
Relevant database；
Second step, from the structuring table extracted one by one in the relevant database in data write-in calculator memory, shape
At " the model training sample table " of the workpiece information, technological parameter and stress data；
Third step carries out the training of dominant frequency prediction model to the screening of vibration frequency according to " the model training sample table "；Into
When row vibration stress relief treatment, user may be selected to allow system according to new empirical data re -training dominant frequency prediction model, or selection
Use original dominant frequency prediction model；When user selects the re -training dominant frequency prediction model, system is first from relationship type
Database is successively read empirical data, and data are classified according to " input ", " output " then, are generated in calculator memory
" input ", " output " tables of data generates dominant frequency screening model, saves the dominant frequency screening model data in the form of a file；
4th step, when user setting is good after the master data of vibration stress relief treatment workpiece, system calls dominant frequency screening model number
Optimal master oscillator frequenc is generated according to dominant frequency screening model and time of vibration technique is joined using master data as parameter according to file
Number；
5th step, after filtering out master oscillator frequenc, starting vibration direct current generator, and vibrating sensor is combined, constantly adjust the vibration
The voltage output of dynamic DC motor speed-regulating device, so that vibration frequency reaches target value；Vibration stress relief treatment is carried out to workpiece；It is shaking
In dynamic ag(e)ing process, industrial control all-in-one machine will acquire by interacting with the wireless strain sensor real time data being mounted on workpiece
To real-time dynamic stress data be input to stress evaluation and test model in, the stress evaluation and test model dominant frequency frequency is corrected in real time,
So that oscillating aging and dynamic stress on-line testing constitute closed loop feedback, to effectively eliminate the residual stress of workpiece.
The expert database content established in the first step includes workpiece weight, size, material, structure feature, exciting
Power, master oscillator frequenc, Support Position, clip position, time of vibration technological parameter and the residual stress test data of vibration front and back.
In third step, when generating the dominant frequency screening model, needs to pre-process input, output data, reject
Abnormal data, and data are described to discrete date or text and are digitized, then by treated, data carry out normalizing
Change processing carries out linear transformation, linear transformation formula are as follows: X '=(x-min to initial data using min-max standardization
(x))/(max (x)-min (x)), so that the dimension of data is consistent.
In third step, the dominant frequency screening model refers to according to the difference using data set, in conjunction with decision tree step-by-step recursion
Binary tree list structure is established, each binary tree represents the master oscillator frequenc of this history sample, in every dominant frequency frequency two
During fork tree generates, when each node carries out branch, randomly extraction section feature participates in the branch of binary tree, so
Recursive branch afterwards is to randomly select Partial Feature from remaining feature to participate in branch every time, most during recursive branch
After generate more dominant frequency prediction binary trees；When carrying out dominant frequency prediction to the oscillating aging basic parameter newly inputted, dominant frequency screens mould
Every one tree in type can all generate a prediction result, determine new input sample eventually by the principle that the minority is subordinate to the majority
Optimal master oscillator frequenc.
The invention has the advantages that system can be according to the base of workpiece without carrying out vibrating live Frequency sweep experiments to welded machine tool bad
Eigen and ageing treatment basic technology demand generate master oscillator frequenc and vibration time；Timeliness workpiece is obtained in real time in conjunction with online
Dynamic strain fitting evaluation and test static stress, and according to aging treatment process target, dynamic corrections vibration frequency improves oscillating aging and goes
Except residual stress one-time success rate.Realize the effect of Online Judge oscillating aging, simultaneously so as to avoid secondary vibration timeliness
Process saves the power consumption of oscillating aging.Using the technical process data management based on database, so that historical empirical data
System database can be automatically write, not only solves the problems, such as data filing, but also intelligent dominant frequency screening model, static stress
Evaluating and testing model can be with on-line automatic training.
Detailed description of the invention
Fig. 1 is to implement hardware layout schematic diagram of the invention.
Fig. 2 is intelligence learning algorithm flow chart of the invention.
Fig. 3 is that the present invention is based on the dominant frequency frequencies of dynamic stress on-line testing to correct flow chart.
Fig. 4 is the data normalization algorithm flow chart in model of mind training of the present invention.
Fig. 5 is dominant frequency screening model algorithm flow chart of the present invention.
Specific embodiment
It elaborates with reference to the accompanying drawing to the embodiment of the present invention, the present embodiment before being with technical solution of the present invention
It puts and is implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to down
State embodiment.
As shown in Figs. 1-5, the folded vibration aging treatment method of the dominant frequency of the present invention for lathe welded machine tool bad, described in foundation
Welded machine tool bad Vibration Aging Process database forms ageing treatment parameter, timeliness quality, residual stress size, welded machine tool bad knot
The production process database record sheet of structure dimensional parameters constructs expert database；It is then based on the vibration of the expert database
Aging technique parameter carries out the folded vibration ageing treatment of dominant frequency, specific steps to welded machine tool bad are as follows:
The first step establishes workpiece (welded machine tool bad) feature, the technological parameter of oscillating aging and residual stress quality assessment index
Relevant database；
Second step, from the structuring table extracted one by one in the relevant database in data write-in calculator memory, shape
At " the model training sample table " of the workpiece information, technological parameter and stress data；
Third step carries out the training of dominant frequency prediction model to the screening of vibration frequency according to " the model training sample table "；Into
When row vibration stress relief treatment, user may be selected to allow system according to new empirical data re -training dominant frequency prediction model, or selection
Use original dominant frequency prediction model；When user selects the re -training dominant frequency prediction model, system is first from relationship type
Database is successively read empirical data, and data are classified according to " input ", " output " then, are generated in calculator memory
" input ", " output " tables of data, as shown in table 1；
Table 1
And dominant frequency screening model is generated, the dominant frequency screening model data are saved in the form of a file；
4th step, when user setting is good after the master data of vibration stress relief treatment workpiece, system calls dominant frequency screening model number
Optimal master oscillator frequenc is generated according to dominant frequency screening model and time of vibration technique is joined using master data as parameter according to file
Number；
5th step, after filtering out master oscillator frequenc, starting vibration direct current generator, and vibrating sensor is combined, constantly adjust the vibration
The voltage output of dynamic DC motor speed-regulating device, so that vibration frequency reaches target value；Vibration stress relief treatment is carried out to workpiece；It is shaking
In dynamic ag(e)ing process, industrial control all-in-one machine will acquire by interacting with the wireless strain sensor real time data being mounted on workpiece
To real-time dynamic stress data be input to stress evaluation and test model in, the stress evaluation and test model dominant frequency frequency is corrected in real time,
So that oscillating aging and dynamic stress on-line testing constitute closed loop feedback, to effectively eliminate the residual stress of workpiece.
The expert database content established in the first step includes workpiece weight, size, material, structure feature, exciting
Power, master oscillator frequenc, Support Position, clip position, time of vibration technological parameter and the residual stress test data of vibration front and back.
In third step, when generating the dominant frequency screening model, needs to pre-process input, output data, reject
Abnormal data, and data are described to discrete date or text and are digitized, then by treated, data carry out normalizing
Change processing carries out linear transformation, linear transformation formula are as follows: X '=(x-min to initial data using min-max standardization
(x))/(max (x)-min (x)), so that the dimension of data is consistent；After being pre-processed to input, output data, in order to
The generality and repeatability for guaranteeing instruction dominant frequency screening model are carried out upsetting processing at random, then be taken to input, output data record
80% data record is as training data, and in addition 20% data are used to verify the accuracy of dominant frequency screening model prediction.
In third step, the dominant frequency screening model refers to according to the difference using data set, in conjunction with decision tree step-by-step recursion
Binary tree list structure is established, each binary tree represents the master oscillator frequenc of this history sample, in every dominant frequency frequency two
During fork tree generates, when each node carries out branch, randomly extraction section feature participates in the branch of binary tree, so
Recursive branch afterwards is to randomly select Partial Feature from remaining feature to participate in branch every time, most during recursive branch
After generate more dominant frequency prediction binary trees；When carrying out dominant frequency prediction to the oscillating aging basic parameter newly inputted, dominant frequency screens mould
Every one tree in type can all generate a prediction result, determine new input sample eventually by the principle that the minority is subordinate to the majority
Optimal master oscillator frequenc.
Claims (4)
1. a kind of folded vibration aging treatment method of dominant frequency for lathe welded machine tool bad, it is characterised in that: establish the welded machine tool bad
Vibration Aging Process database forms ageing treatment parameter, timeliness quality, residual stress size, welded machine tool bad structure size ginseng
Several production process database record sheets constructs expert database；It is then based on the Vibration Aging Process of the expert database
Parameter carries out the folded vibration ageing treatment of dominant frequency, specific steps to welded machine tool bad are as follows:
The first step establishes the relational data of workpiece features, the technological parameter of oscillating aging and residual stress quality assessment index
Library；
Second step, from the structuring table extracted one by one in the relevant database in data write-in calculator memory, shape
At " the model training sample table " of the workpiece information, technological parameter and stress data；
Third step carries out the training of dominant frequency prediction model to the screening of vibration frequency according to " the model training sample table "；Into
When row vibration stress relief treatment, user may be selected to allow system according to new empirical data re -training dominant frequency prediction model, or selection
Use original dominant frequency prediction model；When user selects the re -training dominant frequency prediction model, system is first from relationship type
Database is successively read empirical data, and data are classified according to " input ", " output " then, are generated in calculator memory
" input ", " output " tables of data generates dominant frequency screening model, saves the dominant frequency screening model data in the form of a file；
4th step, when user setting is good after the master data of vibration stress relief treatment workpiece, system calls dominant frequency screening model number
Optimal master oscillator frequenc is generated according to dominant frequency screening model and time of vibration technique is joined using master data as parameter according to file
Number；
5th step, after filtering out master oscillator frequenc, starting vibration direct current generator, and vibrating sensor is combined, constantly adjust the vibration
The voltage output of dynamic DC motor speed-regulating device, so that vibration frequency reaches target value；Vibration stress relief treatment is carried out to workpiece；It is shaking
In dynamic ag(e)ing process, industrial control all-in-one machine will acquire by interacting with the wireless strain sensor real time data being mounted on workpiece
To real-time dynamic stress data be input to stress evaluation and test model in, the stress evaluation and test model dominant frequency frequency is corrected in real time,
So that oscillating aging and dynamic stress on-line testing constitute closed loop feedback, to effectively eliminate the residual stress of workpiece.
2. according to claim 1 for the folded vibration aging treatment method of the dominant frequency of lathe welded machine tool bad, it is characterised in that: first
The expert database content established in step include workpiece weight, size, material, structure feature, exciting force, master oscillator frequenc,
Residual stress test data before and after Support Position, clip position, time of vibration technological parameter and vibration.
3. according to claim 1 for the folded vibration aging treatment method of the dominant frequency of lathe welded machine tool bad, it is characterised in that: third
In step, when generating the dominant frequency screening model, need to pre-process input, output data, the data of rejecting abnormalities, and
It describes data to discrete date or text to digitize, then by treated, data are normalized, that is, are used
Min-max standardization carries out linear transformation, linear transformation formula are as follows: X '=(x-min (x))/(max (x)-to initial data
Min (x)) so that the dimension of data is consistent.
4. according to claim 1 for the folded vibration aging treatment method of the dominant frequency of lathe welded machine tool bad, it is characterised in that: third
In step, the dominant frequency screening model refers to according to the difference using data set, establishes binary tree chain in conjunction with decision tree step-by-step recursion
Table structure, each binary tree represent the master oscillator frequenc of this history sample, in the mistake that every dominant frequency frequency binary tree generates
Cheng Zhong, when each node carries out branch, randomly extraction section feature participates in the branch of binary tree, and then recursive branch, is passed
Return in ramifying, is to randomly select Partial Feature from remaining feature to participate in branch every time, ultimately produces more masters
Frequency prediction binary tree；When carrying out dominant frequency prediction to the oscillating aging basic parameter that newly inputs, each in dominant frequency screening model
Tree can all generate a prediction result, and the optimal main vibration frequency of new input sample is determined eventually by the principle that the minority is subordinate to the majority
Rate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910231525.9A CN109750152B (en) | 2019-03-26 | 2019-03-26 | Main frequency superposition vibration aging treatment method for machine tool welding bed body |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910231525.9A CN109750152B (en) | 2019-03-26 | 2019-03-26 | Main frequency superposition vibration aging treatment method for machine tool welding bed body |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109750152A true CN109750152A (en) | 2019-05-14 |
| CN109750152B CN109750152B (en) | 2020-11-06 |
Family
ID=66409359
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910231525.9A Active CN109750152B (en) | 2019-03-26 | 2019-03-26 | Main frequency superposition vibration aging treatment method for machine tool welding bed body |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109750152B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110578050A (en) * | 2019-10-23 | 2019-12-17 | 郑州机械研究所有限公司 | Composite aging treatment process for reducing residual stress of welding structure lathe bed |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101994001A (en) * | 2010-09-14 | 2011-03-30 | 上海海事大学 | Support vector machine algorithm based method for predicting vibration aging effect |
-
2019
- 2019-03-26 CN CN201910231525.9A patent/CN109750152B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101994001A (en) * | 2010-09-14 | 2011-03-30 | 上海海事大学 | Support vector machine algorithm based method for predicting vibration aging effect |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110578050A (en) * | 2019-10-23 | 2019-12-17 | 郑州机械研究所有限公司 | Composite aging treatment process for reducing residual stress of welding structure lathe bed |
| CN110578050B (en) * | 2019-10-23 | 2021-05-18 | 郑州机械研究所有限公司 | Composite aging treatment process for reducing residual stress of welding structure lathe bed |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109750152B (en) | 2020-11-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| DE102017002607B4 (en) | Processing machine system that determines the acceptance / rejection of workpieces | |
| EP2169491B1 (en) | Support system and method for optimising process parameters and/or regulating parameters | |
| DE102017108169A1 (en) | Production system that specifies a determination value of a variable in relation to a product deviation | |
| DE102016011528A1 (en) | A machine learning apparatus for optimizing the cycle processing time of a processing machine, motor control device, processing machine, and machine learning method | |
| DE102017010799A1 (en) | A machine learning apparatus and robot system for learning a machining order of a laser machining robot and machine learning method therefor | |
| JP2016130908A (en) | Parameter automatic adjustment apparatus for adjusting parameters in response to processing condition | |
| DE10341764A1 (en) | Integrated model prediction control and optimization within a process control system | |
| US10635091B2 (en) | Machining condition adjustment device and machine learning device | |
| CN101055468A (en) | Robust process model identification in model based control techniques | |
| JP2011183455A (en) | Method for monitoring quality of laser machining process and system corresponding to the same | |
| CN110308658B (en) | PID parameter setting method, device and system and readable storage medium | |
| DE102020204854A1 (en) | Machine learning apparatus, numerical control system and machine learning method | |
| CN110991495B (en) | Method, system, medium, and apparatus for predicting product quality in manufacturing process | |
| DE102018117423A1 (en) | MACHINING MACHINE SYSTEM AND MANUFACTURING SYSTEM | |
| DE102016007404A1 (en) | A servo control device with online control gain enhancement function using a weighting function | |
| DE102019106939A1 (en) | WORKING CONDITION ADJUSTMENT DEVICE AND MACHINE LEARNING DEVICE | |
| CN109750152A (en) | Dominant frequency for lathe welded machine tool bad folds vibration aging treatment method | |
| DE102010030691A1 (en) | Dialogue system and method for examining a machining process | |
| DE102019125486A1 (en) | Machining support system and cutting device | |
| DE102020200635A1 (en) | Machine learning system, control device and machine learning method for optimizing filter coefficients | |
| WO2021185909A1 (en) | Computer-implemented method for creating control data sets, cad-cam system and production system | |
| JP3200104B2 (en) | Automatic welding system | |
| CN117620391A (en) | Welding equipment control method and device | |
| EP3717170B1 (en) | Ultrasonic machining device, method for configuring an ultrasonic machining device, and system having an ultrasonic machining device of this type | |
| WO2021028111A1 (en) | Method and apparatus for optimizing an inductive heat-treatment process with the aid of a bayesian optimization process |
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 |