CN117226599B - Numerical control machine tool thermal error prediction method, device, equipment and medium - Google Patents
Numerical control machine tool thermal error prediction method, device, equipment and medium Download PDFInfo
- Publication number
- CN117226599B CN117226599B CN202311493495.1A CN202311493495A CN117226599B CN 117226599 B CN117226599 B CN 117226599B CN 202311493495 A CN202311493495 A CN 202311493495A CN 117226599 B CN117226599 B CN 117226599B
- Authority
- CN
- China
- Prior art keywords
- thermal
- sample
- error prediction
- information
- target
- 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
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000012549 training Methods 0.000 claims abstract description 40
- 238000003062 neural network model Methods 0.000 claims description 12
- 230000015654 memory Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 230000004927 fusion Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 abstract description 5
- 230000007547 defect Effects 0.000 abstract description 3
- 238000003754 machining Methods 0.000 abstract description 3
- 238000013136 deep learning model Methods 0.000 abstract description 2
- 238000012546 transfer Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000020169 heat generation Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Numerical Control (AREA)
Abstract
The invention provides a method, a device, equipment and a medium for predicting thermal errors of a numerical control machine tool, which relate to the technical field of numerical control machining data processing, and the method comprises the following steps: acquiring thermal information of a target machine tool, wherein the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool; inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool; inputting the thermal information into a trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model; and fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool. The method overcomes the defect of inaccurate prediction results of the deep learning model caused by less training data, and improves the accuracy of the thermal error prediction results of the numerical control machine.
Description
Technical Field
The invention relates to the technical field of numerical control processing data processing, in particular to a numerical control machine tool thermal error prediction method, a device, equipment and a medium.
Background
In the process of machining, each part of the numerical control machine tool can generate heat, the heat is mutually transferred to form a relatively complex heat transfer relationship, and the heat generation of the part can lead to thermal deformation and machining errors. In the prior art, there is a method for predicting thermal errors by using a data-driven model, that is, a machine learning or deep learning-based method, in which the thermal errors are predicted by using state data of a machine tool, but this method needs to rely on a large amount of labeled training data, but the amount of labeled thermal error training data in the field of numerically controlled machine tools is small, which results in low accuracy of thermal error data predicted by this method.
Disclosure of Invention
The invention provides a thermal error prediction method, a device, equipment and a medium for a numerical control machine tool, which are used for solving the defect of low accuracy of a thermal error prediction result in the prior art and realizing the effect of improving the accuracy of the thermal error prediction result.
The invention provides a thermal error prediction method of a numerical control machine tool, which comprises the following steps:
acquiring thermal information of a target machine tool, wherein the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool;
inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
inputting the thermal information into a trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model;
and fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool.
According to the method for predicting the thermal error of the numerical control machine tool provided by the invention, the fusion of the first thermal error prediction result and the second thermal error prediction result to obtain the target thermal error prediction result of the target machine tool comprises the following steps:
determining a first confidence level of the first thermal error prediction result and a second confidence level of the second thermal error prediction result based on the thermal information;
and based on the first confidence coefficient and the second confidence coefficient, fusing the first thermal error prediction result and the second thermal error prediction result to obtain the target thermal error prediction result.
According to the method for predicting the thermal error of the numerical control machine tool provided by the invention, the determining of the first confidence coefficient of the first thermal error prediction result and the second confidence coefficient of the second thermal error prediction result based on the thermal information comprises the following steps:
determining first target sample thermal information based on the thermal information, each sample thermal information, and the thermal error physical equation;
acquiring a first target sample thermal error prediction result, wherein the first target sample thermal error prediction result is obtained by inputting the first target sample thermal information into the thermal error physical equation;
determining a target first sample confidence coefficient based on a thermal error label corresponding to the first target sample thermal information and the first target sample thermal error prediction result;
the first confidence is determined based on the target first sample confidence.
According to the method for predicting the thermal error of the numerical control machine tool provided by the invention, the method for determining the thermal information of the first target sample based on the thermal information, the thermal information of each sample and the thermal error physical equation comprises the following steps:
respectively inputting the sample thermal information into the thermal error physical equation to obtain first sample thermal error prediction results corresponding to the sample thermal information;
obtaining first sample confidence degrees corresponding to the sample thermal information respectively according to the thermal error labels corresponding to the sample thermal information and the first sample thermal error prediction results;
clustering each sample thermal information based on a first sample confidence coefficient corresponding to each sample thermal information respectively to obtain each first class;
acquiring an average value of each first class, and determining a target first class in each first class based on the thermal information and the average value of each first class;
and selecting the sample thermal information closest to the thermal information in the first target class as the first target sample thermal information.
According to the method for predicting the thermal error of the numerical control machine tool provided by the invention, the determining of the first confidence coefficient of the first thermal error prediction result and the second confidence coefficient of the second thermal error prediction result based on the thermal information comprises the following steps:
determining second target sample thermal information based on the thermal information, each of the sample thermal information, and the thermal error prediction depth model;
acquiring a second sample thermal error prediction result corresponding to the second target sample thermal information, wherein the second sample thermal error prediction result is obtained by inputting the second target sample thermal information into the trained thermal error prediction depth model;
determining a target second sample confidence level based on a thermal error label corresponding to the second target sample thermal information and the second sample thermal error prediction result;
determining the second confidence based on the target second sample confidence.
According to the method for predicting the thermal error of the numerical control machine tool provided by the invention, the method for determining the thermal information of the second target sample based on the thermal information, the thermal information of each sample and the thermal error prediction depth model comprises the following steps:
respectively inputting the sample thermal information into the trained thermal error prediction depth model, and obtaining second sample thermal error prediction results respectively corresponding to the sample thermal information output by the thermal error prediction depth model;
obtaining second sample confidence degrees corresponding to the sample thermal information respectively according to the thermal error labels corresponding to the sample thermal information and the second sample thermal error prediction results;
clustering each sample thermal information based on second sample confidence degrees respectively corresponding to each sample thermal information to obtain each second class;
acquiring an average value of each second class, and determining a target second class in each second class based on the thermal information and the average value of each second class;
selecting the sample thermal information closest to the thermal information in the second target class as the second target sample thermal information;
in the training process of the thermal error prediction depth model, the training weight of first sample thermal information is greater than the training weight of second sample thermal information, the first sample thermal information is the corresponding sample thermal information with the first sample confidence coefficient lower than a preset value, and the second sample thermal information is the corresponding sample thermal information with the first sample confidence coefficient higher than the preset value.
According to the method for predicting the thermal error of the numerical control machine tool provided by the invention, the fusion of the first thermal error prediction result and the second thermal error prediction result based on the first confidence coefficient and the second confidence coefficient is carried out to obtain the target thermal error prediction result, and the method comprises the following steps:
and inputting the first thermal error prediction result, the first confidence coefficient, the second thermal error prediction result and the second confidence coefficient into a trained two-channel neural network model for fusion, and obtaining the target thermal error prediction result output by the two-channel neural network model.
The invention also provides a device for predicting the thermal error of the numerical control machine tool, which comprises:
the system comprises a thermal information acquisition module, a control module and a control module, wherein the thermal information acquisition module is used for acquiring thermal information of a target machine tool, and the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool;
the first prediction module is used for inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
the second prediction module is used for inputting the thermal information into the trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model;
and the fusion module is used for fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the method for predicting the thermal error of the numerical control machine tool when executing the computer program.
The invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of predicting thermal errors of a numerically controlled machine tool as described in any of the preceding.
According to the thermal error prediction method, the device, the equipment and the medium of the numerical control machine, the thermal error prediction is carried out by establishing the thermal error physical equation, the thermal error prediction result obtained based on the thermal error physical equation and the thermal error prediction result obtained based on the thermal error prediction depth model are fused to obtain the final target thermal error prediction result, the defect of inaccurate deep learning model prediction result caused by less training data is overcome, and the accuracy of the thermal error prediction result of the numerical control machine is improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a numerical control machine thermal error prediction method provided by the invention;
FIG. 2 is a schematic diagram of a thermal error prediction device of a numerical control machine tool;
fig. 3 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The method for predicting the thermal error of the numerical control machine tool provided by the invention is described below with reference to fig. 1, and as shown in fig. 1, the method provided by the invention comprises the following steps:
s110, acquiring thermal information of the target machine tool, wherein the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool.
The heat source component refers to a component which can generate heat in the operation process of the target machine tool, and the heat source component comprises a plurality of components which can generate heat actively and components which are connected with the components generating heat actively and generate heat passively due to heat conduction. The operating state information of the target machine tool includes operating parameters of the target machine tool such as spindle rotation speed, hardness of the processed material, and the like.
S120, inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool.
In the method provided by the invention, a thermal conduction physical model of the target machine tool is constructed in advance, namely, the heat transfer phenomenon in the machine tool is described based on thermodynamic and thermal conduction principles. Different components may have different rates of heat generation due to differences in their structure, materials and principles of operation, e.g., a spindle rotating at high speed may generate much more heat than a rail moving at low speed. The relative positioning and structural relationship between the components may result in different paths for heat transfer and dissipation, e.g., a click closely coupled to the spindle may be affected by the heat generated by the spindle. The thermal error physical equation may include a heat transfer equation for a plurality of components, and the high resolution heat transfer equation may enable independent, detailed modeling of each heat source component of the target machine tool, taking into account all heat transfer key factors. Each heat transfer equation considers component characteristics and interactions between components. Substituting the thermal information into a thermal error physical equation, and obtaining a first thermal error prediction result through analysis. The thermal error physical equation considers an ideal physical heat transfer process, and in actual operation, there are a plurality of factors affecting an ideal state, so that a first thermal error prediction result obtained based on the thermal error physical equation has a limitation.
And S130, inputting the thermal information into the trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model.
The thermal error prediction depth model is a model obtained based on training of multiple sets of training data, and can adopt a Bi-LSTM model, namely a bidirectional long and short term memory network model, which is a special network structure in deep learning, and can capture time sequence dependency in data, wherein the Bi-LSTM comprises LSTM in two directions, one from front to back and one from back to front, which means that the network can utilize past information and also can utilize future information, so that the network can better capture the most critical and representative characteristics in the time sequence data.
Because the effect of deep learning is greatly affected by the data volume, and the existing general database has no numerical control machine tool thermal error data with labels temporarily, the automatic acquisition and labeling of a large amount of data also requires high cost, the training data for training the thermal error prediction depth model is less, and various situations cannot be covered, so that the accuracy of the second thermal error prediction result output by the thermal error prediction depth model is poor.
And S140, fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool.
According to the method provided by the invention, the first thermal error prediction result obtained based on the thermal error physical equation and the second thermal error prediction result obtained based on the thermal error prediction depth model are fused to obtain the target thermal error prediction result of the target machine tool, so that the accuracy of the thermal error prediction result is improved. Specifically, fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool, including:
determining a first confidence level of the first thermal error prediction result and a second confidence level of the second thermal error prediction result based on the thermal information;
and based on the first confidence coefficient and the second confidence coefficient, fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result.
The method provided by the invention is used for presetting the confidence coefficient of the thermal error prediction result obtained by two modes (a mode based on a thermal error physical equation and a mode based on a thermal error prediction depth model), fusing the two modes based on the confidence coefficient, and reflecting the confidence coefficient of the thermal error prediction result obtained by processing thermal information in the two modes.
Determining a first confidence of the first thermal error prediction result based on the thermal information, comprising:
determining first target sample thermal information based on the thermal information, each sample thermal information, and a thermal error physical equation;
acquiring a first target sample thermal error prediction result, wherein the first target sample thermal error prediction result is obtained by inputting first target sample thermal information into a thermal error physical equation;
determining a target first sample confidence level based on a thermal error label corresponding to the first target sample thermal information and a first target sample thermal error prediction result;
a first confidence is determined based on the target first sample confidence.
The sample thermal information is thermal information in training data for training the thermal error prediction depth model, the training data for training the thermal error prediction depth model comprises a plurality of groups, each group of training data comprises the sample thermal information and a thermal error label corresponding to the sample thermal information, and the thermal error label is a true value. The sample thermal information can be acquired when the target machine tool runs, and the real thermal error of the target machine tool is acquired as a thermal error label corresponding to the sample thermal information.
The sample thermal information has a corresponding thermal error label, and the corresponding thermal error label can be used for evaluating the confidence level of a thermal error prediction result obtained by processing the sample thermal information based on a thermal error physical equation. And selecting the confidence corresponding to the sample thermal information closest to the thermal information as the first confidence. Specifically, determining first target sample thermal information based on the thermal information, the respective sample thermal information, and the thermal error physical equation includes:
respectively inputting the thermal information of each sample into a thermal error physical equation, and obtaining a first sample thermal error prediction result corresponding to the thermal information of each sample;
obtaining first sample confidence coefficients corresponding to the sample thermal information respectively according to the thermal error labels corresponding to the sample thermal information and the first sample thermal error prediction results;
clustering the thermal information of each sample based on the first sample confidence coefficient corresponding to the thermal information of each sample respectively to obtain each first class;
acquiring an average value of each first class, and determining a target first class in each first class based on the thermal information and the average value of each first class;
and selecting the sample thermal information closest to the thermal information in the first target class as the first target sample thermal information.
The first sample confidence corresponding to the sample thermal information reflects the difference between the first sample thermal error prediction result corresponding to the sample thermal information and the thermal error label, and if the first sample thermal error prediction result corresponding to the sample thermal information is closer to the thermal error label corresponding to the first sample thermal error prediction result, the first sample confidence corresponding to the sample thermal information is higher, and otherwise, the first sample confidence corresponding to the sample thermal information is lower.
And clustering the thermal information of each sample according to the first sample confidence corresponding to each thermal information of the sample, namely, the thermal information of the samples with similar corresponding first sample confidence is gathered into a class, and the clustering result reflects which class of thermal information has more accurate prediction capability for a thermal error physical equation. The average value of each first class is the average value of the sample thermal information included in the first class, the thermal information is matched with the average value of each first class, the first class corresponding to the nearest average value is selected as a target first class, then the sample thermal information nearest to the thermal information is selected as first target sample thermal information in the target first class, and the first sample confidence corresponding to the first target sample thermal information is used as the first confidence of the first thermal error prediction result.
Determining a second confidence level for the second thermal error prediction result based on the thermal information, comprising:
determining second target sample thermal information based on the thermal information, each sample thermal information, and the thermal error prediction depth model;
acquiring a second sample thermal error prediction result corresponding to the second target sample thermal information, wherein the second sample thermal error prediction result is obtained by inputting the second target sample thermal information into a trained thermal error prediction depth model;
determining a target second sample confidence level based on a thermal error label corresponding to the second target sample thermal information and a second sample thermal error prediction result;
a second confidence is determined based on the target second sample confidence.
Specifically, determining second target sample thermal information based on the thermal information, the sample thermal information, and the thermal error prediction depth model, includes:
respectively inputting the thermal information of each sample into a trained thermal error prediction depth model, and obtaining second sample thermal error prediction results respectively corresponding to the thermal information of each sample output by the thermal error prediction depth model;
obtaining second sample confidence degrees respectively corresponding to the sample thermal information according to the thermal error labels corresponding to the sample thermal information and the second sample thermal error prediction results;
clustering the thermal information of each sample based on the second sample confidence coefficient corresponding to the thermal information of each sample respectively to obtain each second class;
acquiring an average value of each second class, and determining a target second class in each second class based on the thermal information and the average value of each second class;
and selecting the sample thermal information closest to the thermal information in the second target class as second target sample thermal information.
The second sample confidence corresponding to the sample thermal information reflects the difference between the second sample thermal error prediction result corresponding to the sample thermal information and the thermal error label, and if the second sample thermal error prediction result corresponding to the sample thermal information is closer to the thermal error label corresponding to the second sample thermal error prediction result, the second sample confidence corresponding to the sample thermal information is higher, otherwise the second sample confidence corresponding to the sample thermal information is lower.
And clustering the thermal information of each sample according to the second sample confidence coefficient corresponding to each sample, namely, the thermal information of samples with similar corresponding second sample confidence coefficient is gathered into one class, and the clustering result reflects the thermal error prediction depth model has more accurate prediction capability on the thermal information of which class. The average value of each second class is the average value of the sample thermal information included in the second class, the thermal information is matched with the average value of each second class, the second class corresponding to the nearest average value is selected as a target second class, then the sample thermal information nearest to the thermal information is selected as second target sample thermal information in the target second class, and the second sample confidence corresponding to the second target sample thermal information is used as the second confidence of the second thermal error prediction result.
In the method, in order to make the finally obtained target thermal error prediction result more accurate, according to the confidence coefficient of the thermal error prediction result corresponding to the thermal error physical equation, the training process of the thermal error prediction depth model is adjusted so that the thermal error prediction results output by the thermal error prediction depth model and the thermal error prediction depth model can compensate each other, specifically, in the training process of the thermal error prediction depth model, the training weight of the first sample thermal information is greater than the training weight of the second sample thermal information, the first sample thermal information is the sample thermal information with the corresponding first sample confidence coefficient lower than the preset value, and the second sample thermal information is the sample thermal information with the corresponding first sample confidence coefficient higher than the preset value.
That is, in the training process of the thermal error prediction depth model, the training weight of the sample thermal information with poor prediction effect of the thermal error physical equation is improved, so that the thermal error prediction depth model has higher prediction capability for the thermal information with poor prediction effect of the thermal error physical equation, thus the shortages of the thermal error physical equation can be compensated, and the accuracy of the obtained target thermal error prediction result is improved.
The training weight of the first sample thermal information is greater than that of the second sample thermal information, and the training weight of the first sample thermal information can be realized by setting the training gradient of the training data of the first sample thermal information to be greater than that of the training data of the second sample thermal information, and when model parameters are updated based on training loss, the model parameters are updated by multiplying the training loss corresponding to the training data of the first sample thermal information by a coefficient greater than 1 and then serving as new training loss, and the model parameters are updated by directly adopting the training loss corresponding to the training data of the second sample thermal information.
Based on the first confidence coefficient and the second confidence coefficient, fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result, wherein the method comprises the following steps:
and inputting the first thermal error prediction result, the first confidence coefficient, the second thermal error prediction result and the second confidence coefficient into the trained two-channel neural network model for fusion, and obtaining the target thermal error prediction result of the two-channel neural network model.
The two-channel neural network model is a small model, has small parameters, can obtain better results by adopting less data, and comprises the following training processes:
taking a first sample thermal error prediction result and a first sample confidence coefficient corresponding to the sample thermal information as input of one channel of the two-channel neural network model, and taking a second sample thermal error prediction result and a second sample confidence coefficient corresponding to the sample thermal information as input of the other channel of the two-channel neural network model to obtain a sample target thermal error prediction result output by the two-channel neural network model;
and obtaining training loss based on the thermal error label corresponding to the thermal error prediction result of the sample target and the thermal information of the sample, and updating the two-channel neural network model based on the training loss.
The first sample confidence coefficient corresponding to the first sample thermal error prediction result and the second sample confidence coefficient corresponding to the second sample thermal error prediction result are real values which can be obtained based on the thermal error label corresponding to the sample thermal information, so that the two-channel neural network model can obtain higher training quality, and the model parameters which can fuse the two input thermal error prediction results based on the confidence coefficients can be approximated more quickly, and the more accurate thermal error prediction result can be obtained.
The numerical control machine thermal error prediction device provided by the invention is described below, and the numerical control machine thermal error prediction device described below and the numerical control machine thermal error prediction method described above can be correspondingly referred to each other. As shown in fig. 2, the thermal error prediction apparatus for a numerically-controlled machine tool provided by the present invention includes:
a thermal information obtaining module 210, configured to obtain thermal information of the target machine tool, where the thermal information includes temperature information of a heat source component of the target machine tool and operation state information of the target machine tool;
the first prediction module 220 is configured to input thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
the second prediction module 230 is configured to input thermal information into the trained thermal error prediction depth model, and obtain a second thermal error prediction result output by the thermal error prediction depth model;
and the fusion module 240 is configured to fuse the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool.
Fig. 3 illustrates a physical schematic diagram of an electronic device, as shown in fig. 3, where the electronic device may include: processor 310, communication interface (Communications Interface) 320, memory 330 and communication bus 340, wherein processor 310, communication interface 320, memory 330 accomplish communication with each other through communication bus 340. The processor 310 may invoke logic instructions in the memory 330 to perform a numerical control machine thermal error prediction method comprising: acquiring thermal information of a target machine tool, wherein the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool;
inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
inputting the thermal information into the trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model;
and fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool.
Further, the logic instructions in the memory 330 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the method for predicting thermal errors of a numerical control machine provided by the above methods, the method comprising: acquiring thermal information of a target machine tool, wherein the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool;
inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
inputting the thermal information into the trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model;
and fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. The thermal error prediction method of the numerical control machine tool is characterized by comprising the following steps of:
acquiring thermal information of a target machine tool, wherein the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool;
inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
inputting the thermal information into a trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model;
fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool;
the fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool comprises the following steps:
determining a first confidence level of the first thermal error prediction result and a second confidence level of the second thermal error prediction result based on the thermal information;
based on the first confidence coefficient and the second confidence coefficient, fusing the first thermal error prediction result and the second thermal error prediction result to obtain the target thermal error prediction result;
the determining, based on the thermal information, a first confidence of the first thermal error prediction result and a second confidence of the second thermal error prediction result includes:
determining first target sample thermal information based on the thermal information, each sample thermal information, and the thermal error physical equation;
acquiring a first target sample thermal error prediction result, wherein the first target sample thermal error prediction result is obtained by inputting the first target sample thermal information into the thermal error physical equation;
determining a target first sample confidence coefficient based on a thermal error label corresponding to the first target sample thermal information and the first target sample thermal error prediction result;
determining the first confidence based on the target first sample confidence;
the determining the first target sample thermal information based on the thermal information, the respective sample thermal information, and the thermal error physical equation includes:
respectively inputting the sample thermal information into the thermal error physical equation to obtain first sample thermal error prediction results corresponding to the sample thermal information;
obtaining first sample confidence degrees corresponding to the sample thermal information respectively according to the thermal error labels corresponding to the sample thermal information and the first sample thermal error prediction results;
clustering each sample thermal information based on a first sample confidence coefficient corresponding to each sample thermal information respectively to obtain each first class;
acquiring an average value of each first class, and determining a target first class in each first class based on the thermal information and the average value of each first class;
and selecting the sample thermal information closest to the thermal information in the first target class as the first target sample thermal information.
2. The method of claim 1, wherein determining a first confidence level of the first thermal error prediction result and a second confidence level of the second thermal error prediction result based on the thermal information comprises:
determining second target sample thermal information based on the thermal information, each of the sample thermal information, and the thermal error prediction depth model;
acquiring a second sample thermal error prediction result corresponding to the second target sample thermal information, wherein the second sample thermal error prediction result is obtained by inputting the second target sample thermal information into the trained thermal error prediction depth model;
determining a target second sample confidence level based on a thermal error label corresponding to the second target sample thermal information and the second sample thermal error prediction result;
determining the second confidence based on the target second sample confidence.
3. The method of claim 2, wherein determining the second target sample thermal information based on the thermal information, each of the sample thermal information, and the thermal error prediction depth model comprises:
respectively inputting the sample thermal information into the trained thermal error prediction depth model, and obtaining second sample thermal error prediction results respectively corresponding to the sample thermal information output by the thermal error prediction depth model;
obtaining second sample confidence degrees corresponding to the sample thermal information respectively according to the thermal error labels corresponding to the sample thermal information and the second sample thermal error prediction results;
clustering each sample thermal information based on second sample confidence degrees respectively corresponding to each sample thermal information to obtain each second class;
acquiring an average value of each second class, and determining a target second class in each second class based on the thermal information and the average value of each second class;
selecting the sample thermal information closest to the thermal information in the second target class as the second target sample thermal information;
in the training process of the thermal error prediction depth model, the training weight of first sample thermal information is greater than the training weight of second sample thermal information, the first sample thermal information is the corresponding sample thermal information with the first sample confidence coefficient lower than a preset value, and the second sample thermal information is the corresponding sample thermal information with the first sample confidence coefficient higher than the preset value.
4. The method for predicting thermal errors of a numerical control machine according to claim 1, wherein the fusing the first thermal error prediction result and the second thermal error prediction result based on the first confidence level and the second confidence level to obtain the target thermal error prediction result includes:
and inputting the first thermal error prediction result, the first confidence coefficient, the second thermal error prediction result and the second confidence coefficient into a trained two-channel neural network model for fusion, and obtaining the target thermal error prediction result output by the two-channel neural network model.
5. The utility model provides a digit control machine tool thermal error prediction unit which characterized in that includes:
the system comprises a thermal information acquisition module, a control module and a control module, wherein the thermal information acquisition module is used for acquiring thermal information of a target machine tool, and the thermal information comprises temperature information of a heat source component of the target machine tool and working state information of the target machine tool;
the first prediction module is used for inputting the thermal information into a preset thermal error physical equation to obtain a first thermal error prediction result of the target machine tool;
the second prediction module is used for inputting the thermal information into the trained thermal error prediction depth model to obtain a second thermal error prediction result output by the thermal error prediction depth model;
the fusion module is used for fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool;
the fusing the first thermal error prediction result and the second thermal error prediction result to obtain a target thermal error prediction result of the target machine tool comprises the following steps:
determining a first confidence level of the first thermal error prediction result and a second confidence level of the second thermal error prediction result based on the thermal information;
based on the first confidence coefficient and the second confidence coefficient, fusing the first thermal error prediction result and the second thermal error prediction result to obtain the target thermal error prediction result;
the determining, based on the thermal information, a first confidence of the first thermal error prediction result and a second confidence of the second thermal error prediction result includes:
determining first target sample thermal information based on the thermal information, each sample thermal information, and the thermal error physical equation;
acquiring a first target sample thermal error prediction result, wherein the first target sample thermal error prediction result is obtained by inputting the first target sample thermal information into the thermal error physical equation;
determining a target first sample confidence coefficient based on a thermal error label corresponding to the first target sample thermal information and the first target sample thermal error prediction result;
determining the first confidence based on the target first sample confidence;
the determining the first target sample thermal information based on the thermal information, the respective sample thermal information, and the thermal error physical equation includes:
respectively inputting the sample thermal information into the thermal error physical equation to obtain first sample thermal error prediction results corresponding to the sample thermal information;
obtaining first sample confidence degrees corresponding to the sample thermal information respectively according to the thermal error labels corresponding to the sample thermal information and the first sample thermal error prediction results;
clustering each sample thermal information based on a first sample confidence coefficient corresponding to each sample thermal information respectively to obtain each first class;
acquiring an average value of each first class, and determining a target first class in each first class based on the thermal information and the average value of each first class;
and selecting the sample thermal information closest to the thermal information in the first target class as the first target sample thermal information.
6. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of predicting thermal errors of a numerically controlled machine tool according to any one of claims 1 to 4 when executing the computer program.
7. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the numerical control machine thermal error prediction method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311493495.1A CN117226599B (en) | 2023-11-10 | 2023-11-10 | Numerical control machine tool thermal error prediction method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311493495.1A CN117226599B (en) | 2023-11-10 | 2023-11-10 | Numerical control machine tool thermal error prediction method, device, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117226599A CN117226599A (en) | 2023-12-15 |
| CN117226599B true CN117226599B (en) | 2024-01-30 |
Family
ID=89093091
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311493495.1A Active CN117226599B (en) | 2023-11-10 | 2023-11-10 | Numerical control machine tool thermal error prediction method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117226599B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108188821A (en) * | 2018-01-08 | 2018-06-22 | 东北大学 | A kind of Ball-screw in NC Machine Tools feed system Thermal Error Forecasting Methodology |
| CN108803486A (en) * | 2018-08-16 | 2018-11-13 | 重庆理工大学 | Numerical control machining tool heat error prediction based on deep learning network in parallel and compensation method |
| CN111414977A (en) * | 2020-03-09 | 2020-07-14 | 西南交通大学 | Weighted integration temperature sensitive point combination selection method for machine tool spindle thermal error modeling |
| CN114332984A (en) * | 2021-12-06 | 2022-04-12 | 腾讯科技（深圳）有限公司 | Training data processing method, device and storage medium |
| CN115470842A (en) * | 2022-08-24 | 2022-12-13 | 襄阳华中科技大学先进制造工程研究院 | Machine tool thermal error prediction method, device, equipment and storage medium |
| CN115758208A (en) * | 2022-11-09 | 2023-03-07 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Traction converter fault diagnosis method and device, computer equipment and storage medium |
| CN116415501A (en) * | 2023-04-11 | 2023-07-11 | 重庆大学 | MGU-A thermal error prediction model creation method and thermal error control system based on digital twin |
| CN116454872A (en) * | 2023-04-06 | 2023-07-18 | 中国华能集团清洁能源技术研究院有限公司 | Wind power prediction method and device, electronic equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
-
2023
- 2023-11-10 CN CN202311493495.1A patent/CN117226599B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108188821A (en) * | 2018-01-08 | 2018-06-22 | 东北大学 | A kind of Ball-screw in NC Machine Tools feed system Thermal Error Forecasting Methodology |
| CN108803486A (en) * | 2018-08-16 | 2018-11-13 | 重庆理工大学 | Numerical control machining tool heat error prediction based on deep learning network in parallel and compensation method |
| CN111414977A (en) * | 2020-03-09 | 2020-07-14 | 西南交通大学 | Weighted integration temperature sensitive point combination selection method for machine tool spindle thermal error modeling |
| CN114332984A (en) * | 2021-12-06 | 2022-04-12 | 腾讯科技（深圳）有限公司 | Training data processing method, device and storage medium |
| CN115470842A (en) * | 2022-08-24 | 2022-12-13 | 襄阳华中科技大学先进制造工程研究院 | Machine tool thermal error prediction method, device, equipment and storage medium |
| CN115758208A (en) * | 2022-11-09 | 2023-03-07 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Traction converter fault diagnosis method and device, computer equipment and storage medium |
| CN116454872A (en) * | 2023-04-06 | 2023-07-18 | 中国华能集团清洁能源技术研究院有限公司 | Wind power prediction method and device, electronic equipment and storage medium |
| CN116415501A (en) * | 2023-04-11 | 2023-07-11 | 重庆大学 | MGU-A thermal error prediction model creation method and thermal error control system based on digital twin |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117226599A (en) | 2023-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6871842B2 (en) | Machining simulation condition optimization method, machining simulation equipment, machining simulation system and program | |
| CN114237155A (en) | Error prediction and compensation method, system and medium for multi-axis numerical control machining | |
| CN116009480B (en) | Fault monitoring method, device and equipment of numerical control machine tool and storage medium | |
| CN109582588A (en) | Method for generating test case, device and electronic equipment | |
| CN110889091A (en) | Machine tool thermal error prediction method and system based on temperature sensitive interval segmentation | |
| JP2019135638A (en) | Data driving method for automatically detecting abnormal workpiece during production process | |
| CN115993804B (en) | Cutter parameter adjustment method based on numerical control machine tool and related equipment | |
| CN114281950B (en) | Data retrieval method and system based on multi-graph weighted fusion | |
| CN117226599B (en) | Numerical control machine tool thermal error prediction method, device, equipment and medium | |
| CN116926241B (en) | Immune cell culture control method and system | |
| JP2019040423A (en) | Detection device, detection method, and detection program | |
| CN117033053A (en) | Data analysis method and system applied to automobile marketing management system software | |
| US20040267508A1 (en) | Computer numerical control code-simulating system and method | |
| CN117953977B (en) | Recombinant mesenchymal stem cell culture control method and system | |
| CN116680961B (en) | Measurement compensation method, device, equipment and storage medium considering clamping force deformation | |
| CN116400905B (en) | Code automatic generation method for regulating and controlling multiple devices and related devices | |
| CN107817761B (en) | Part processing method and system based on error iterative learning | |
| CN117908464B (en) | Machine tool contour error prediction method, system, device and readable storage medium | |
| CN117033052A (en) | Object abnormality diagnosis method and system based on model identification | |
| CN115951634B (en) | Numerical control processing energy consumption prediction method and device, electronic equipment and storage medium | |
| CN117401179A (en) | Unmanned aerial vehicle-based flight performance testing method and system | |
| CN117273846A (en) | Visual interaction processing method and system for digital mall based on artificial intelligence | |
| CN116010903A (en) | Main shaft thermal error prediction method and device based on multi-source heterogeneous data | |
| US20050027378A1 (en) | Machine tool turret optimizing apparatus and method | |
| CN117078345A (en) | Tourism activity content pushing method and system based on artificial intelligence |
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 |