CN108803486A - Numerical control machining tool heat error prediction based on deep learning network in parallel and compensation method - Google Patents
Numerical control machining tool heat error prediction based on deep learning network in parallel and compensation method Download PDFInfo
- Publication number
- CN108803486A CN108803486A CN201810936815.9A CN201810936815A CN108803486A CN 108803486 A CN108803486 A CN 108803486A CN 201810936815 A CN201810936815 A CN 201810936815A CN 108803486 A CN108803486 A CN 108803486A
- Authority
- CN
- China
- Prior art keywords
- deep learning
- thermal error
- numerical control
- parallel
- error prediction
- 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
- 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
- G05B19/404—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 control arrangements for compensation, e.g. for backlash, overshoot, tool offset, tool wear, temperature, machine construction errors, load, inertia
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Manufacturing & Machinery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Automatic Control Of Machine Tools (AREA)
- Numerical Control (AREA)
Abstract
The invention discloses a kind of numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel, include the following steps：A, collecting sample data：Heat source measurement point is chosen on numerically-controlled machine tool, the temperature value of heat source measurement point, and the Spindle thermal error value at corresponding time point is detected, as sample data；B, the deep learning Thermal Error prediction model based on depth belief network in parallel is established；C, sample data training deep learning Thermal Error prediction model will be collected；D, the temperature value of numerical control lathe heat source measurement point is detected in real time, and inputs the deep learning Thermal Error prediction model after training, predicts thermal error value in real time；E, it using the thermal error value of prediction as the compensation translational movement of NC Machine Tools Coordinate system origin, is deviated by coordinate origin and realizes Thermal Error real-time compensation.The present invention has many advantages, such as mapping relations that can be complicated between monitoring temperature signal and Thermal Error in the case of accurate characterization big data, is conducive to improve Thermal Error prediction and compensation precision.
Description
Technical field
The present invention relates to the Accuracy Control fields of numerically-controlled machine tool industry, in particular to one kind based on depth in parallel
Practise the numerical control machining tool heat error prediction and compensation method of network.
Background technology
According to statistics, the mismachining tolerance caused by process system thermal deformation accounts for total mismachining tolerance in precision machining processes
40%-70%, the proportion that wherein Thermal Deformation of NC Machine Tool error accounts for is very big, in addition account for the 50% of entire workpiece machining error with
On.It is to improve the important guarantee of numerically-controlled machine tool machining accuracy rationally and effectively to carry out Thermal Error control.Error compensation method is exactly it
Middle the most frequently used effective method of one kind.And heat error compensation premise is that establishing machine tool thermal error and temperature as precisely as possible
Mapping relations between degree, to forecast Thermal Error with lathe temperature value during real-time compensation.
Since Thermal Error itself has quasi-static time-varying, non-linear, decaying delay and coupling comprehensive characteristics, so difficult
To establish accurate Thermal Model using theory analysis.Currently used thermal error modeling method is Experimental modeling method,
Correlation analysis is made to thermal error data and lathe temperature value according to statistical theory and is fitted modeling with the principle of least square.Closely
Nian Lai, shallow-layer neural network theory (BP networks, RBF networks), gray system theory etc. have also applied in thermal error modeling.But
The Thermal Model set up with conventional method other than there are compensation precision and robustness problem, it is big there is also two
Defect：First, needing to be grasped a large amount of signal processing technology in conjunction with abundant engineering experience to extract signal characteristic；Second is that
It is difficult to characterize mapping relations complicated between monitoring signals and Thermal Error in the case of big data using shallow Model.
Deep learning network and its learning algorithm, as successful big data analysis method, compared with conventional method, depth
Learning method with data-driven, can be automatically from extracting data feature (knowledge), for analyzing, unstructured, pattern is unknown more
Become, cross-cutting big data has significant advantage.
Invention content
In view of the above shortcomings of the prior art, the technical problem to be solved by the present invention is to：How to provide one kind can be certainly
It moves and extracts numerically-controlled machine tool temperature data further feature, it is multiple between monitoring temperature signal and Thermal Error in the case of accurate characterization big data
Miscellaneous mapping relations are conducive to improve Thermal Error prediction with compensation precision, real-time and adaptability based on deep learning in parallel
The numerical control machining tool heat error of network is predicted and compensation method.
In order to solve the above-mentioned technical problem, present invention employs the following technical solutions：
A kind of numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel, which is characterized in that packet
Include following steps：
A, collecting sample data：Heat source measurement point is chosen on numerically-controlled machine tool, detects the temperature value of heat source measurement point and right
The Spindle thermal error value for answering time point, as sample data；
B, the deep learning Thermal Error prediction model based on depth belief network in parallel is established；
C, deep learning Thermal Error prediction model is trained using the collected sample datas of step A；
D, the temperature value of numerical control lathe heat source measurement point is detected in real time, and inputs the deep learning Thermal Error prediction after training
Model predicts thermal error value in real time；
E, inclined by coordinate origin using the thermal error value of prediction as the compensation translational movement of NC Machine Tools Coordinate system origin
It moves and realizes Thermal Error real-time compensation.
Further, which is characterized in that the temperature samples data of input are normalized to section [0,1], it is right
The thermal error value of neural network forecast output carries out normalized.
Further, the deep learning Thermal Error prediction model is mainly believed by three depth with identical network structure
It reads network to be formed in parallel, is respectively used to predict the DBN1 of the thermal error value of main shaft in the X-axis direction, for predicting main shaft in Y
The DBN2 of thermal error value in axis direction and DBN3 for predicting the thermal error value of main shaft in the Z-axis direction；The DBN1,
DBN2 and DBN3 includes 1 visual input layer, and 1 output layer and 3 limit Boltzmann machine RBM hidden layers, respectively
RBM1, RBM2 and RBM3；Described DBN1, DBN2 and DBN3 are respectively provided with different weighting parameters, and one RBM1 layers shared.
Further, the heat source measurement point quantity chosen in the neuronal quantity of the visual input layer and the step A
Unanimously；The neuronal quantity of the output layer is 1, and output valve is heat of the main shaft in X-direction, Y direction or Z-direction
Error amount.
Further, the neuronal quantity of the limitation Boltzmann machine RBM hidden layers is determined using following steps：First set
Determine neuron initial numberIn formula, r is the heat source measurement point quantity chosen in step A, then with step-length be s by
Step increases, while with the deep learning Thermal Error prediction model predicted root mean square error, and the deep learning Thermal Error is pre-
It is implicit that the neuronal quantity corresponding to lowest mean square root error that survey model prediction obtains is determined as limitation Boltzmann machine RBM
The neuronal quantity of layer.
Further, using logarithm divergence unsupervised learning method, the depth belief network DBN1 in pre-training model is obtained
DBN1 network initial weights are obtained, the DBN1 parameters completed correspondence will be trained to be assigned to DBN2, DBN3, realize that initial weight is shared
Further, the temperature data at each moment forms exemplar data with the Thermal Error at corresponding moment, uses label
Sample data finely tunes the best initial weights for generating 3 depth belief networks using BP algorithm respectively.
Further, in the step A, at least 15 heat source measurement points are chosen, and acquire at least 100 groups of temperature values and right
The thermal error value answered is as sample data.
Further, the heat source measurement point is mainly distributed on main shaft of numerical control machine tool, each feed shaft feed screw nut pair, bed
At body, coolant liquid and operating room.
Further, the thermal error value includes 3 Thermal Errors of the main shaft in X-direction, Y direction and Z-direction
Value, respectively ex、ey、ez。
In conclusion the invention has the advantages that：
1, the present invention is based on depth belief networks in parallel, solve conventional method and are difficult to characterize big data using shallow Model
In the case of mapping relations complicated between monitoring temperature signal and Thermal Error.Compared to traditional Thermal Error forecasting system, carry significantly
While high forecasting accuracy, the real-time and adaptability of heat error compensation are significantly improved.
2, the present invention is based on deep learning principles, can automatically extract numerically-controlled machine tool temperature data further feature, not need
Artificial intervention is intervened, the dependence to a large amount of signal processing technologies and diagnostic experiences can be broken away from, the adaptive of feature is completed and carries
Take with the intelligent predicting of Thermal Error state and compensation, robustness is good, real-time.
3, the present invention improves the Deep Learning ability of Thermal Error prediction network, solves conventional method and needs to be grasped largely
Signal processing technology the problem of signal characteristic is extracted in conjunction with abundant engineering experience.
Description of the drawings
Fig. 1 is numerical control machining tool heat error prediction and compensation method flow chart based on deep learning network in parallel.
Fig. 2 is gantry machining center Spindle thermal error instrumentation plan.
Fig. 3 is numerically-controlled machine tool deep learning Thermal Error prediction model structure.
Fig. 4 is X-direction heat error compensation curve.
Fig. 5 is Y direction heat error compensation curve.
Fig. 6 is Z-direction heat error compensation curve.
Specific implementation mode
With reference to a kind of gantry machining center Thermal Error prediction and compensation the present invention is described in further detail.
When it is implemented, as shown in Figure 1, specifically using following steps：
One, numerically-controlled machine tool key point temperature value and the Spindle thermal error value of corresponding time are acquired as sample data；
(1) it is distributed according to the structure of the gantry machining center, operating mode and heat source, the setting of heat source key point is generated heat in lathe
Larger position or its near, detect temperature datas using 18 temperature sensing.Temperature Key point is numbered from T1 to T18, and arrangement is such as
Under：
1) left and right pillar position
Leading screw upper and lower axle bearing：T1, T2；Feed screw nut：T3, T4；Guide rail：T5,T6.
2) crossbeam position
Leading screw or so bearing block：T7, T8；Feed screw nut T9；Guide rail T10.
3) ram position
Spindle box upper surface：T11；On the left of spindle box：T12；On the right side of spindle box：T13；Spindle flange：T14.
4) lathe bed and workbench position
Lathe bed：T15；Workbench：T16.
5) coolant liquid position：T17 at coolant inlet pipe.
6) environment temperature：Workshop temperature T18.
(2) 3 thermal error values of the measurement gantry machining center main shaft in radial direction and axial direction：ex、ey、ez, measure
Mode is as shown in Figure 2.
(3) the continuous cyclic process state of simulation gantry machining center, main shaft rotation, feed shaft movement, coolant liquid cycle, but
Not actual cut.Data acquisition divides morning and afternoon to carry out, and preheats gantry machining center 0.5 hour before acquisition, and 1 hour noon stopped
Acquisition, but do not shut down, continue gathered data in the afternoon, the numerical value every the primary each temperature sensor of 5 minutes records and main shaft displacement
The numerical value of sensor acquires 150 groups of temperature and thermal error data as sample data altogether.
(4) 150 groups of sample datas are normalized into section [0,1] as follows, for deep learning prediction model
Training and verification：
A in formulai' indicate the value after each sample data normalization, aiFor each sample data original value, amaxAnd aminTable respectively
Show the minimum value and maximum value of all types of sample datas.
Two, the deep learning Thermal Error prediction model based on depth belief network in parallel is established；
(1) numerically-controlled machine tool deep learning Thermal Error prediction model structure is built
Numerically-controlled machine tool deep learning Thermal Error prediction model it is in parallel by 3 depth belief networks (DBN1, DBN2, DBN3) and
At.3 depth belief networks have identical network structure, different weighting parameters, and shared 1 limitation Boltzmann machine is
RBM1 layers.
DBN1, DBN2, DBN3 predict respectively main shaft radial X-direction, the radial direction directions y and axial Z-direction thermal error value
ex、ey、ez.Numerically-controlled machine tool deep learning Thermal Error prediction model is as shown in 3 figures.
(2) depth belief network structure determination
1) each depth belief network includes 1 visual input layer, and 3 limit Boltzmann machine RBM hidden layers and 1
Output layer.
2) visual input layer quantity is identical as the crucial hot source point quantity of setting, is 18.Output layer nerve
First quantity is 1, and output is thermal error value of the main shaft in X-direction, Y direction or Z-direction, specifically, DBN1
Output is the thermal error value of main shaft in the X-axis direction；The output of DBN2 is the thermal error value of main shaft in the Y-axis direction；DBN3's
Output is the thermal error value of main shaft in the Z-axis direction；
3) neuronal quantity of each RBM hidden layers is affected to the precision of prediction and generalization ability of model.Neuron
The very few loss that can cause temperature profile information of quantity, causes feature extraction imperfect, precision of prediction is low.Increase neuron number
Amount, precision of prediction can improve, but neuronal quantity excessively can cause the generalization ability of model poor.The neuron number of RBM hidden layers
Amount determines that method is：
1. the sample data volume L=150 of counted according to heat source key r=18 and acquisition, first set each RBM hidden layers
Neuron initial number isThe formula meaning is to take the integer part of r/2, is made using root-mean-square error RMSE
Model prediction accuracy is evaluated for index, root-mean-square error RMSE is defined as：
Wherein, ytFor Thermal Error actual value,For the Thermal Error predicted value of model, n is Thermal Error total number, and t is Thermal Error
Serial number.In the present embodiment, p0For initial value 9 when, the root-mean-square error that is calculated is 0.102.
2. the neuronal quantity of RBM2 and RBM3 hidden layers is kept to immobilize, it is implicit to gradually increase RBM1 with step-length s=2
The neuronal quantity p of layer1, in the present embodiment, work as p1When=21, the root-mean-square error of model prediction reaches minimum value 0.025, will
p1=21 are determined as the neuronal quantity of RBM1 hidden layers.
3. same method, when determining RBM2, the neuronal quantity of RBM1 is fixed on the neuron number of 21, RBM3
Amount is fixed on 9, and the final neuronal quantity for determining RBM2 is 21.When determining RBM3, the neuronal quantity of RBM1 is fixed
21 are fixed in the neuronal quantity of 21, RBM2, the final neuronal quantity for determining RBM3 is 19.
Determine that the neuronal quantity of RBM2 and RBM3 hidden layers is respectively p2=21, p3=19.
Three, deep learning Thermal Error prediction model is trained with institute's collecting sample data；
(1) logarithm divergence unsupervised learning method is used, wherein 1 depth belief network DBN1 in pre-training model,
Its network initial weight is obtained, depth belief network DBN2, DBN3 share the initial weight.Depth belief network DBN logarithms dissipate
It is as follows to spend unsupervised learning method：
1) it is 0 or 1 by each layer of neuron random initializtion of DBN, the neuron connection weight w between different layersijIt sets
For the arbitrary value in (0,1) range.Pre-training RBM1 first.
2) with a temperature samples data by visual layers neuron viCalculate hidden layer neuron hj, then connection weight wijJust
It is H to gradientij=vi×hj。
3) by hidden layer neuron hjThe visual layers neuron that backwards calculation obtains is vi', then the reversed gradient of connection weight
For Hij'=vi′×hj。
4) weight is updated：wij=wij+ε×(Hij-Hij'), wherein ε is learning rate, is usually set to 0<ε<1.
5) 18 temperature datas of each moment acquisition are a sample data, totally 150 acquisition moment i.e. totally 150 samples
Notebook data.150 temperature samples data of cycle acquisition go to train RBM1 networks, do not stop iteration, until convergence, i.e. Hij-Hij
≤ δ, δ are convergence threshold, and value is less than 0.01.
In the present embodiment, when pre-training RBM1, learning rate ε is set as 0.1, and iteration convergence threshold value is set as 0.005, warp
Pre-training is completed after 990 iteration.Using same method, completed after 860 times and 1100 iteration respectively to RBM2 and
The pre-training of RBM3.The DBN1 parameters correspondence that training is completed is assigned to DBN2, DBN3, realizes that initial weight is shared.
(2) temperature data at each moment forms exemplar data with the Thermal Error at corresponding moment.With exemplar number
According to, using BP algorithm respectively finely tune generate 3 depth belief networks best initial weights.
By BP methods, with label data, the reversed initial weight corrected pre-training and obtained；Then, positive test is new obtains
Each layer weighting parameter arrived；Repeatedly, until the root-mean-square error of model prediction restrains.
Root-mean-square error E restrains discriminate：
Wherein t is frequency of training, and λ is minimum threshold value, is usually set to be less than 0.01.
In the present embodiment, the convergence threshold λ of root-mean-square error E is set as 0.005, and fine tuning learning rate ε is set as 0.1,3 depths
It spends belief network and completes weights fine tuning after 35,46,39 iteration respectively, at this time the prediction root mean square of 3 depth belief networks
Error is respectively 0.01,0.015,0.012.
After above-mentioned pre-training and tuning, prediction model structure is completed, and each model parameter successfully obtains.Then it can incite somebody to action
The series of temperature data input prediction model detected in real time, for predicting corresponding Spindle thermal error value.
Four, detection numerically-controlled machine tool key point temperature value, input deep learning predict neural network forecast thermal error value in real time.
30 groups of gantry machining center key hot source point temperature datas are measured in real time, and it is pre- to input deep learning prediction network
Spindle thermal error is surveyed, obtains very high-precision Thermal Error prediction result, as shown in figures 4-6.The Thermal Error of model prediction is square
Root error is 0.018, and real time data prediction is almost the same with the performance indicator of training sample data prediction, illustrates deep learning heat
Error prediction model has extraordinary generalization ability.
Five, it using the thermal error value of prediction as the compensation translational movement of lathe coordinate system origin, is deviated by coordinate origin
Realize Thermal Error real-time compensation.
Thermal Error of the gantry machining center main shaft on X, Y, Z3 directions is respectively ex、ey、ez, using this as machine
The compensation translational movement of bed coordinate origin, lathe coordinate system origin translation-e respectively on 3 directions of X, Y, Z axisx、-ey、-ez,
The accurate compensation that can realize machine tool thermal error, the maximum residul difference of X, Y, Z-direction after compensation is respectively 2.5 μm, 3 μm, 2.5
μm, the residual error after compensation is as shown in figures 4-6.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not limitation, all essences in the present invention with the present invention
All any modification, equivalent and improvement etc., should all be included in the protection scope of the present invention made by within refreshing and principle.
Claims (10)
1. a kind of numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel, which is characterized in that including
Following steps：
A, collecting sample data：Choose heat source measurement point on numerically-controlled machine tool, detect the temperature value of heat source measurement point, and to it is corresponding when
Between the Spindle thermal error value put, as sample data；
B, the deep learning Thermal Error prediction model based on depth belief network in parallel is established；
C, deep learning Thermal Error prediction model is trained using the collected sample datas of step A；
D, the temperature value of numerical control lathe heat source measurement point is detected in real time, and inputs the deep learning Thermal Error prediction mould after training
Type predicts thermal error value in real time；
E, it using the thermal error value of prediction as the compensation translational movement of NC Machine Tools Coordinate system origin, is deviated by coordinate origin real
Existing Thermal Error real-time compensation.
2. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as described in claim 1,
Be characterized in that, which is characterized in that in the step A, to the temperature samples data of input be normalized to section [0,
1], in the step C and step D, normalized is carried out to the thermal error value of output.
3. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as described in claim 1,
It is characterized in that, the deep learning Thermal Error prediction model mainly has the depth belief networks of identical network structure simultaneously by three
Join, is respectively used to predict the DBN1 of the thermal error value of main shaft in the X-axis direction, for predicting main shaft in the Y-axis direction
Thermal error value DBN2 and DBN3 for predicting the thermal error value of main shaft in the Z-axis direction；Described DBN1, DBN2 and DBN3
Include 1 visual input layer, 1 output layer and 3 limitation Boltzmann machine RBM hidden layers, respectively RBM1, RBM2 and
RBM3；Described DBN1, DBN2 and DBN3 are respectively provided with different weighting parameters, and one RBM1 layers shared.
4. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as claimed in claim 3,
It is characterized in that, the neuronal quantity of the visual input layer is consistent with the heat source measurement point quantity chosen in the step A；It is described
The neuronal quantity of output layer is 1, and output valve is thermal error value of the main shaft in X-direction, Y direction or Z-direction.
5. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as claimed in claim 4,
It is characterized in that, the neuronal quantity of the limitation Boltzmann machine RBM hidden layers is determined using following steps：First set neuron
Initial numberIn formula, r is the heat source measurement point quantity chosen in step A, is then stepped up for s with step-length, together
When with the deep learning Thermal Error prediction model predicted root mean square error, the deep learning Thermal Error prediction model is predicted
The obtained neuronal quantity corresponding to lowest mean square root error is determined as the neuron of the limitation Boltzmann machine RBM hidden layers
Quantity.
6. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as claimed in claim 5,
It is characterized in that, using logarithm divergence unsupervised learning method, the depth belief network DBN1 in pre-training model obtains DBN1 nets
The DBN1 parameters correspondence that training is completed is assigned to DBN2, DBN3, realizes that initial weight is shared by network initial weight.
7. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as claimed in claim 6,
Be characterized in that, the Thermal Error at the temperature data at each moment and corresponding moment forms exemplar data, with exemplar data,
Finely tune the best initial weights for generating 3 depth belief networks respectively using BP algorithm.
8. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as described in claim 1,
It is characterized in that, in the step A, chooses at least 15 heat source measurement points, and acquire at least 100 groups of temperature values and corresponding heat accidentally
Difference is as sample data.
9. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as described in claim 1,
It is characterized in that, the heat source measurement point is mainly distributed on main shaft of numerical control machine tool, each feed shaft feed screw nut pair, lathe bed, coolant liquid
At operating room.
10. numerical control machining tool heat error prediction and compensation method based on deep learning network in parallel as described in claim 1,
It is characterized in that, the thermal error value includes 3 thermal error values of the main shaft in X-direction, Y direction and Z-direction, respectively
ex、ey、ez。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810936815.9A CN108803486B (en) | 2018-08-16 | 2018-08-16 | Numerical control machine tool thermal error prediction and compensation method based on parallel deep learning network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810936815.9A CN108803486B (en) | 2018-08-16 | 2018-08-16 | Numerical control machine tool thermal error prediction and compensation method based on parallel deep learning network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108803486A true CN108803486A (en) | 2018-11-13 |
| CN108803486B CN108803486B (en) | 2021-02-02 |
Family
ID=64080245
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810936815.9A Active CN108803486B (en) | 2018-08-16 | 2018-08-16 | Numerical control machine tool thermal error prediction and compensation method based on parallel deep learning network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108803486B (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109765846A (en) * | 2019-03-11 | 2019-05-17 | 哈尔滨理工大学 | The selection method of double turntable five-axle number control machine tool thermal error modeling temperature sensitive points |
| CN109800537A (en) * | 2019-02-20 | 2019-05-24 | 大连理工大学 | A kind of machine tool thermal error model reliability degree calculation method based on deep neural network and Monte Carlo method |
| CN110058569A (en) * | 2019-05-19 | 2019-07-26 | 重庆理工大学 | A kind of numerical control machining tool heat error modeling method based on Optimization of Fuzzy neural network |
| CN110083125A (en) * | 2019-05-19 | 2019-08-02 | 重庆理工大学 | A kind of machine tool thermal error modeling method based on deep learning |
| CN110568817A (en) * | 2019-09-13 | 2019-12-13 | 深圳市烨嘉为技术有限公司 | machine tool motion temperature difference compensation method based on big data analysis and prejudgment |
| CN111240268A (en) * | 2020-01-14 | 2020-06-05 | 重庆大学 | Axle system thermal error modeling method and thermal error compensation system based on SLSTM neural network |
| CN111273607A (en) * | 2018-12-04 | 2020-06-12 | 沈阳高精数控智能技术股份有限公司 | Spark-based numerical control machine tool running state monitoring method |
| CN113021078A (en) * | 2021-03-01 | 2021-06-25 | 珠海艾诚精密模具有限公司 | Method for detecting dynamic elongation of NC spindle |
| CN113341878A (en) * | 2021-06-23 | 2021-09-03 | 重庆理工大学 | Thermal error measuring method of five-axis numerical control machine tool |
| CN114265365A (en) * | 2021-12-21 | 2022-04-01 | 重庆大学 | Gear grinding machine thermal error dynamic modeling and compensation method based on online measurement |
| CN114609967A (en) * | 2020-12-04 | 2022-06-10 | 迈鑫机械工业股份有限公司 | Real-time space precision compensation intelligent module of numerical control machine |
| CN116224905A (en) * | 2023-03-02 | 2023-06-06 | 广东工业大学 | Machine tool thermal error prediction method and system based on joint distribution self-adaption |
| CN117226599A (en) * | 2023-11-10 | 2023-12-15 | 上海诺倬力机电科技有限公司 | Numerical control machine tool thermal error prediction method, device, equipment and medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104764527A (en) * | 2015-03-31 | 2015-07-08 | 上海大学 | Thermal error measuring device for workpiece machining |
| CN104898562A (en) * | 2015-04-27 | 2015-09-09 | 攀枝花学院 | Modeling method of thermal error compensation of numerically-controlled machine tool |
| CN105607575A (en) * | 2016-01-27 | 2016-05-25 | 电子科技大学 | FA-LSSVM based numerical control machine tool main shaft heat drift modeling method |
| CN106372337A (en) * | 2016-09-05 | 2017-02-01 | 华中科技大学 | Thermal deformation prediction method of preheating stage of numerical control machine tool |
| CN108363870A (en) * | 2018-02-11 | 2018-08-03 | 宁波大学 | A kind of main shaft of numerical control machine tool thermal error modeling method based on deep learning |
| JP6679086B2 (en) * | 2015-11-11 | 2020-04-15 | 国立研究開発法人情報通信研究機構 | Learning device, prediction device, learning method, prediction method, and program |
-
2018
- 2018-08-16 CN CN201810936815.9A patent/CN108803486B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104764527A (en) * | 2015-03-31 | 2015-07-08 | 上海大学 | Thermal error measuring device for workpiece machining |
| CN104898562A (en) * | 2015-04-27 | 2015-09-09 | 攀枝花学院 | Modeling method of thermal error compensation of numerically-controlled machine tool |
| JP6679086B2 (en) * | 2015-11-11 | 2020-04-15 | 国立研究開発法人情報通信研究機構 | Learning device, prediction device, learning method, prediction method, and program |
| CN105607575A (en) * | 2016-01-27 | 2016-05-25 | 电子科技大学 | FA-LSSVM based numerical control machine tool main shaft heat drift modeling method |
| CN106372337A (en) * | 2016-09-05 | 2017-02-01 | 华中科技大学 | Thermal deformation prediction method of preheating stage of numerical control machine tool |
| CN108363870A (en) * | 2018-02-11 | 2018-08-03 | 宁波大学 | A kind of main shaft of numerical control machine tool thermal error modeling method based on deep learning |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111273607A (en) * | 2018-12-04 | 2020-06-12 | 沈阳高精数控智能技术股份有限公司 | Spark-based numerical control machine tool running state monitoring method |
| CN111273607B (en) * | 2018-12-04 | 2022-12-13 | 沈阳高精数控智能技术股份有限公司 | Spark-based numerical control machine tool running state monitoring method |
| CN109800537A (en) * | 2019-02-20 | 2019-05-24 | 大连理工大学 | A kind of machine tool thermal error model reliability degree calculation method based on deep neural network and Monte Carlo method |
| CN109800537B (en) * | 2019-02-20 | 2022-11-18 | 大连理工大学 | Machine tool thermal error model reliability calculation method based on deep neural network and Monte Carlo method |
| CN109765846B (en) * | 2019-03-11 | 2022-03-18 | 哈尔滨理工大学 | Method for selecting thermal error modeling temperature sensitive points of double-turntable five-axis numerical control machine tool |
| CN109765846A (en) * | 2019-03-11 | 2019-05-17 | 哈尔滨理工大学 | The selection method of double turntable five-axle number control machine tool thermal error modeling temperature sensitive points |
| CN110058569A (en) * | 2019-05-19 | 2019-07-26 | 重庆理工大学 | A kind of numerical control machining tool heat error modeling method based on Optimization of Fuzzy neural network |
| CN110083125A (en) * | 2019-05-19 | 2019-08-02 | 重庆理工大学 | A kind of machine tool thermal error modeling method based on deep learning |
| CN110083125B (en) * | 2019-05-19 | 2021-05-07 | 重庆理工大学 | Machine tool thermal error modeling method based on deep learning |
| CN110058569B (en) * | 2019-05-19 | 2021-05-11 | 重庆理工大学 | Numerical control machine tool thermal error modeling method based on optimized fuzzy neural network |
| CN110568817A (en) * | 2019-09-13 | 2019-12-13 | 深圳市烨嘉为技术有限公司 | machine tool motion temperature difference compensation method based on big data analysis and prejudgment |
| CN111240268A (en) * | 2020-01-14 | 2020-06-05 | 重庆大学 | Axle system thermal error modeling method and thermal error compensation system based on SLSTM neural network |
| CN111240268B (en) * | 2020-01-14 | 2021-03-16 | 重庆大学 | Axle system thermal error modeling method and thermal error compensation system based on SLSTM neural network |
| CN114609967A (en) * | 2020-12-04 | 2022-06-10 | 迈鑫机械工业股份有限公司 | Real-time space precision compensation intelligent module of numerical control machine |
| CN113021078A (en) * | 2021-03-01 | 2021-06-25 | 珠海艾诚精密模具有限公司 | Method for detecting dynamic elongation of NC spindle |
| CN113341878A (en) * | 2021-06-23 | 2021-09-03 | 重庆理工大学 | Thermal error measuring method of five-axis numerical control machine tool |
| CN114265365A (en) * | 2021-12-21 | 2022-04-01 | 重庆大学 | Gear grinding machine thermal error dynamic modeling and compensation method based on online measurement |
| CN114265365B (en) * | 2021-12-21 | 2024-05-28 | 重庆大学 | Online measurement-based dynamic modeling and compensation method for thermal error of gear grinding machine |
| CN116224905A (en) * | 2023-03-02 | 2023-06-06 | 广东工业大学 | Machine tool thermal error prediction method and system based on joint distribution self-adaption |
| CN116224905B (en) * | 2023-03-02 | 2023-10-20 | 广东工业大学 | Machine tool thermal error prediction method and system based on joint distribution self-adaption |
| CN117226599A (en) * | 2023-11-10 | 2023-12-15 | 上海诺倬力机电科技有限公司 | Numerical control machine tool thermal error prediction method, device, equipment and medium |
| CN117226599B (en) * | 2023-11-10 | 2024-01-30 | 上海诺倬力机电科技有限公司 | Numerical control machine tool thermal error prediction method, device, equipment and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108803486B (en) | 2021-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108803486A (en) | Numerical control machining tool heat error prediction based on deep learning network in parallel and compensation method | |
| WO2021174525A1 (en) | Parts surface roughness and cutting tool wear prediction method based on multi-task learning | |
| CN111366123B (en) | Part surface roughness and cutter wear prediction method based on multi-task learning | |
| WO2020041956A1 (en) | Bayes- and fault tree-based reliability evaluation method for computer numerical control machine tool | |
| WO2021128576A1 (en) | Tool condition monitoring dataset enhancement method based on generative adversarial network | |
| CN110561191B (en) | Numerical control machine tool cutter abrasion data processing method based on PCA and self-encoder | |
| Wu et al. | Modeling and analysis of tool wear prediction based on SVD and BiLSTM | |
| CN106842922B (en) | Numerical control machining error optimization method | |
| CN113807570B (en) | XGBoost-based reservoir dam risk level assessment method and system | |
| CN114619292B (en) | Milling cutter wear monitoring method based on fusion of wavelet denoising and attention mechanism with GRU network | |
| Zhang et al. | Temperature variable optimization for precision machine tool thermal error compensation on optimal threshold | |
| CN111881574B (en) | Wind turbine generator set key component reliability modeling method based on distribution function optimization | |
| CN113554621B (en) | Cutter wear state identification system and method based on wavelet scale map and deep migration learning | |
| CN113496312A (en) | Cutter residual life prediction method and equipment based on multi-dimensional feature extraction fusion and long-term and short-term memory network and storage medium | |
| CN110096810A (en) | A kind of industrial process flexible measurement method based on layer-by-layer data extending deep learning | |
| Qin et al. | Tool wear identification and prediction method based on stack sparse self-coding network | |
| CN111126255A (en) | Numerical control machine tool cutter wear value prediction method based on deep learning regression algorithm | |
| Liu et al. | Thermal error modeling for machine tools: Mechanistic analysis and solution for the pseudocorrelation of temperature-sensitive points | |
| CN111047476A (en) | Dam structure safety monitoring accurate prediction method and system based on RBF neural network | |
| Gao et al. | New tool wear estimation method of the milling process based on multisensor blind source separation | |
| Wang et al. | Research on tool remaining life prediction method based on CNN-LSTM-PSO | |
| CN116643020A (en) | Method for improving surface quality of milling workpiece based on high-density water-based foam cutting fluid | |
| You et al. | Normalized Variational Auto-Encoder With the Adaptive Activation Function for Tool Setting in Ultraprecision Turning | |
| CN114357851A (en) | Numerical control milling surface roughness prediction method based on DAE-RNN | |
| CN112116003B (en) | Self-adaptive dynamic fault diagnosis method and system for complex sensor 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 |