CN115951634A - Numerical control machining energy consumption prediction method and device, electronic equipment and storage medium - Google Patents
Numerical control machining energy consumption prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115951634A CN115951634A CN202310240482.7A CN202310240482A CN115951634A CN 115951634 A CN115951634 A CN 115951634A CN 202310240482 A CN202310240482 A CN 202310240482A CN 115951634 A CN115951634 A CN 115951634A
- Authority
- CN
- China
- Prior art keywords
- energy consumption
- machining
- consumption prediction
- characteristic
- network
- 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
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P70/00—Climate change mitigation technologies in the production process for final industrial or consumer products
- Y02P70/10—Greenhouse gas [GHG] capture, material saving, heat recovery or other energy efficient measures, e.g. motor control, characterised by manufacturing processes, e.g. for rolling metal or metal working
Landscapes
- Numerical Control (AREA)
Abstract
The invention provides a numerical control machining energy consumption prediction method, a device, electronic equipment and a storage medium, and relates to the technical field of numerical control combined machining, wherein the method comprises the following steps: acquiring running rotating speed information of the processing machine tool, inputting the running rotating speed information into a first characteristic extraction network to extract a main shaft rotating speed change characteristic of the processing machine tool, and acquiring a first characteristic output by the first characteristic extraction network; rasterizing the three-dimensional model of the workpiece to obtain each three-dimensional grid of the three-dimensional model, and adding corresponding cutting information into the three-dimensional grids according to the processing technological parameters of the workpiece to obtain a voxel model; inputting the voxel model into a trained second feature extraction network to extract the features of the voxel model and obtain second features output by the voxel model feature extraction network; and inputting the first characteristic and the second characteristic into the trained energy consumption prediction network, and acquiring a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network. The method can realize the processing energy consumption prediction facing to the curved surface.
Description
Technical Field
The invention relates to the technical field of numerical control combined machining, in particular to a method and a device for predicting numerical control machining energy consumption, electronic equipment and a storage medium.
Background
A complex numerical control machine tool such as a five-axis machine tool can realize motion cutting in the five-axis direction and complex curved surface machining, and in the process of carrying out multi-axis motion cutting to carry out curved surface machining, part characteristics, tool tracks and the running state environment of the machine tool are very complex, while the existing machine tool energy consumption prediction method is only oriented to the machining process with simple motion forms such as plane machining, shaft machining and the like, and is not oriented to the curved surface machining energy consumption prediction method.
Disclosure of Invention
The invention provides a numerical control machining energy consumption prediction method, a numerical control machining energy consumption prediction device, electronic equipment and a storage medium, which are used for overcoming the defect that no curved surface-oriented machining energy consumption prediction method exists in the prior art and realizing the energy consumption prediction of curved surface machining.
The invention provides a numerical control machining energy consumption prediction method, which comprises the following steps:
acquiring operation rotating speed information of a processing machine tool, wherein the operation rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, inputting the operation rotating speed information into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and acquiring a first characteristic output by the first characteristic extraction network;
acquiring a three-dimensional model of the workpiece, rasterizing the three-dimensional model to obtain each three-dimensional grid of the three-dimensional model, acquiring machining process parameters of the workpiece, and adding corresponding cutting information in the three-dimensional grids according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves into the three-dimensional grids for cutting;
inputting the voxel model into a trained second feature extraction network to extract features of the voxel model, and acquiring second features output by the voxel model feature extraction network;
inputting the first characteristic and the second characteristic into a trained energy consumption prediction network, and obtaining a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, wherein the energy consumption prediction network is obtained by training based on a plurality of groups of data including machining energy consumption labels.
According to the numerical control machining energy consumption prediction method provided by the invention, the energy consumption prediction network comprises a fusion module, a correction module and an output module; the inputting the first characteristic and the second characteristic into a trained energy consumption prediction network to obtain a processing energy consumption prediction result of the workpiece output by the energy consumption prediction network includes:
inputting the first feature and the second feature into the fusion module to fuse the first feature and the second feature, and acquiring a first predicted feature output by the fusion module;
acquiring processing resource state information of the processing machine tool when the processing machine tool processes the workpiece, wherein the processing resource state information reflects the state of processing resources of the processing machine tool when the processing machine tool actually processes the workpiece, inputting the processing resource state information into the correction module to extract a feature reflecting the influence of the processing resource state on energy consumption, and acquiring a correction feature output by the correction module;
correcting the first prediction characteristic based on the correction characteristic to obtain a second prediction characteristic;
and inputting the second prediction characteristic into the output module, and acquiring the processing energy consumption prediction result output by the output module.
According to the numerical control machining energy consumption prediction method provided by the invention, the first feature extraction network, the second feature extraction network and the energy consumption prediction network are obtained by training based on a plurality of groups of training data, and each group of training data comprises sample running rotating speed information, a sample voxel model, sample machining resource state information and a machining energy consumption label corresponding to a sample workpiece, wherein the sample running rotating speed information, the sample voxel model, the sample machining resource state information and the machining energy consumption label correspond to the sample workpiece; the training process of the first feature extraction network, the second feature extraction network and the energy consumption prediction network comprises the following steps:
acquiring a plurality of groups of training data to form a first training batch, and updating parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network respectively based on each group of training data in the first training batch;
after parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are updated based on the last group of training data in the first training batch, selecting a plurality of training data in the first training batch to form a second training batch, and updating parameters of the correction module in the energy consumption prediction network based on each group of training data in the second training batch, wherein processing process files of the sample workpieces corresponding to each group of training data in the second training batch are the same;
and re-executing the step of obtaining the plurality of groups of training data to form the first training batch until the parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are converged.
According to the numerical control machining energy consumption prediction method provided by the invention, the machining resources comprise the machining tool and the machining tool, and the machining resource state information reflects the thermal deformation degree of the machining tool and the machining tool in the machining process and the wear degree of the machining tool.
According to the numerical control machining energy consumption prediction method provided by the invention, the acquiring of the machining resource state information of the machining tool when machining the workpiece comprises the following steps:
acquiring a temperature field signal of the processing machine tool when the workpiece is processed;
acquiring a vibration signal of the machining tool when the workpiece is machined;
and combining the temperature field signal and the vibration signal to obtain the processing resource state information.
According to the numerical control machining energy consumption prediction method provided by the invention, the step of acquiring the running rotating speed information of the machining tool comprises the following steps:
and acquiring a processing technology file of the workpiece, and acquiring the running rotating speed information of the processing machine tool based on the processing technology file.
According to the numerical control machining energy consumption prediction method provided by the invention, the cutting information comprises the feed amount and the motion trail of the machining tool.
The invention also provides a device for predicting the energy consumption of numerical control machining, which comprises:
the first characteristic extraction module is used for acquiring running rotating speed information of a processing machine tool, the running rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, the running rotating speed information is input into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and first characteristics output by the first characteristic extraction network are acquired;
the voxel model generating module is used for acquiring a three-dimensional model of the workpiece, rasterizing the three-dimensional model to obtain each three-dimensional lattice of the three-dimensional model, acquiring machining process parameters of the workpiece, and adding corresponding cutting information into the three-dimensional lattices according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves into the three-dimensional lattices to cut;
the second feature extraction module is used for inputting the voxel model to a trained second feature extraction network to extract the features of the voxel model and acquiring second features output by the voxel model feature extraction network;
the energy consumption prediction module is used for inputting the first characteristic and the second characteristic into a trained energy consumption prediction network to obtain a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, and the energy consumption prediction network is obtained based on a plurality of groups of data including machining energy consumption labels through training.
The invention also provides electronic equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the processor executes the program, the numerical control machining energy consumption prediction method is realized.
The present invention also provides a non-transitory computer readable storage medium, on which a computer program is stored, the computer program, when executed by a processor, implementing the method for predicting energy consumption in nc machining as described in any one of the above.
The numerical control machining energy consumption prediction method, the numerical control machining energy consumption prediction device, the electronic equipment and the storage medium provided by the invention have the advantages that the three-dimensional model of the machined workpiece is processed into the voxel model, cutting information reflecting the cutting amount of the cutter during cutting of the three-dimensional grid is associated in each three-dimensional grid of the voxel model, so that the complex motion condition of the cutter in the curved surface machining process can be divided into fine-grained information, a second characteristic reflecting the motion information of the cutter can be extracted by adopting a second characteristic extraction network, the running rotating speed information of the machine tool related to energy consumption is input into a first characteristic extraction network to provide a first characteristic reflecting the rotating speed change of a main shaft, the first characteristic and the second characteristic are input into a trained energy consumption prediction network to predict the machining energy consumption, and the curved surface-oriented machining energy consumption prediction is realized.
Drawings
In order to more clearly illustrate the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic flow chart of a numerical control machining energy consumption prediction method provided by the invention;
FIG. 2 is a schematic structural diagram of a numerical control machining energy consumption prediction device provided by the invention;
fig. 3 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The inventor finds that the existing machine tool machining energy consumption prediction is directed to simple plane machining or shaft machining, and the moving track of a cutter is simple in the machining process with a simple motion form. However, for machine tool machining facing a curved surface (for example, a five-axis machine tool can realize motion cutting in the five-axis direction to machine a complex curved surface), the motion form of a tool in the machining process is complex, and the existing machine tool machining energy consumption prediction method is not suitable for machine tool machining facing a curved surface. Aiming at the defect, the invention provides a numerical control machining energy consumption prediction method, which comprises the steps of processing a three-dimensional model of a machined workpiece into a voxel model, associating cutting information reflecting the cutting amount of a cutter when the cutter cuts the three-dimensional grid in each three-dimensional grid of the voxel model, dividing the complex motion condition of the cutter in the curved surface machining process into fine-grained information, extracting a second characteristic reflecting the motion information of the cutter by adopting a second characteristic extraction network, inputting the running rotating speed information of a machine tool related to energy consumption into a first characteristic extraction network to provide a first characteristic reflecting the rotating speed change of a main shaft, and inputting the first characteristic and the second characteristic into a trained energy consumption prediction network to predict the machining energy consumption, thereby realizing the prediction of the machining energy consumption facing the curved surface.
As shown in fig. 1, the numerical control machining energy consumption prediction method provided by the invention comprises the following steps:
s100, obtaining operation rotating speed information of a processing machine tool, wherein the operation rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, inputting the operation rotating speed information into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and obtaining a first characteristic output by the first characteristic extraction network;
s200, acquiring a three-dimensional model of the workpiece, rasterizing the three-dimensional model to obtain each three-dimensional grid of the three-dimensional model, acquiring machining workpiece parameters of the workpiece, and adding corresponding cutting information into the three-dimensional grids according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves into the three-dimensional grids for cutting.
The inventor researches and discovers that in the process of processing complex curved surfaces, the energy consumption of the machine tool is related to a plurality of factors, mainly including the rotating speed of a main shaft of the machine tool and the cutting amount of a processing cutter. However, in the process of machining a complex curved surface, because the running track is complex, the direction and the feed amount of the running track at each stage of the running track are difficult to analyze the influence of the cutting amount on the energy consumption. In the method provided by the embodiment, in addition to extracting the characteristics of the operating speed information of the processing machine tool for energy consumption prediction, the three-dimensional model of the workpiece is divided into a plurality of three-dimensional grids, the cutting amount of the processing tool when moving into the three-dimensional grids for cutting is associated in each three-dimensional grid, and after a voxel model is formed, the characteristics of the voxel model are extracted. Thus, the cutting amount which is changed in a complex manner in the machining process is divided into fine-grained information, and a feature extraction network can be adopted for feature extraction.
In actual production, when a numerical control machine tool processes a workpiece, a processing file of the workpiece is firstly compiled, the processing file is input into the machine tool, and the machine tool processes the workpiece based on the processing file. The operation rotating speed of the main shaft of the processing machine tool in the processing process of the workpiece can be read from the processing file, and the operation rotating speed information is obtained. Namely, the acquiring of the operating speed information of the processing machine tool includes:
and acquiring a processing technology file of the workpiece, and acquiring the running rotating speed information of the processing machine tool based on the processing technology file.
The operation rotating speed information of the processing machine tool is obtained from the pre-programmed processing technology file, the automatic extraction of the operation rotating speed information can be realized, and for workpieces sharing the same processing drawing, the operation rotating speed information can be extracted and processed only once, so that the calculated amount in the process of obtaining the processing energy consumption prediction result is reduced, and the real-time performance of energy consumption prediction is improved.
And after the running rotating speed information is obtained, inputting the running rotating speed information into a first feature extraction network, extracting the main shaft rotating speed change feature when the machining machine tool machines the workpiece by the first feature extraction network, and taking the feature extracted by the first feature extraction network as the first feature. The first feature extraction network may adopt the structure of an existing feature extraction network.
And rasterizing the three-dimensional model of the workpiece, wherein the rasterization processing can be realized by adopting the conventional three-dimensional rasterization processing mode. The machining process parameters include a machining track of the machining tool and the feed amount of the machining tool in each section of the machining track, and the machining process parameters can be extracted from a machining process file of the workpiece which is written in advance. According to the machining track and the feed amount, the accurate cutting amount can be calculated. In one possible implementation, the cutting information may be a cutting amount of the machining tool moving in the three-dimensional grid to perform cutting. However, since the trajectory may be a curve, calculating an accurate cut amount may result in a large calculation amount. In the method provided by the invention, the cutting information comprises the feed amount and the motion trail of the processing cutter. That is to say, in the method provided by the present invention, the accurate cutting amount is not directly calculated, but the motion trajectory and the feed amount capable of reflecting the cutting amount are used to be associated with each three-dimensional grid obtained by rasterizing the three-dimensional model of the workpiece, so as to obtain the voxel model corresponding to the workpiece. Therefore, accurate cutting amount does not need to be calculated, the calculation amount is reduced, and the real-time performance of machine tool energy consumption prediction is improved.
After the voxel model of the workpiece is obtained, the numerical control machining energy consumption prediction method further comprises the following steps:
s300, inputting the voxel model into a trained second feature extraction network to extract the features of the voxel model, and acquiring the second features output by the voxel model feature extraction network.
And after rasterizing the three-dimensional model of the workpiece and associating the cutting information, performing feature extraction by using the second feature extraction network to obtain the second feature. The second feature extraction network may be obtained by modifying the structure of an existing network for extracting the voxel model features, which is basically a capture point for outputting the model. And in the method provided by the invention, the second feature extraction network is different in feature used for outputting subsequent continuous processing. In a possible implementation manner, the second feature extraction Network is obtained by modifying a structure of a VGN (virtual instrumentation Network, a capture detection point generation Network), specifically, an output unit of the VGN is a 3-enhanced unit, that is, the VGN has three output ends.
S400, inputting the first characteristic and the second characteristic into a trained energy consumption prediction network, and obtaining a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, wherein the energy consumption prediction network is obtained by training based on multiple groups of data including machining energy consumption labels.
As explained above, the first characteristic reflects a change in the spindle speed of the machine tool, and the second characteristic reflects the amount of cutting by the machine tool when machining the workpiece. The method provided by the invention uses the first characteristic and the second characteristic together for predicting the energy consumption of the processing machine tool.
Specifically, the energy consumption prediction network comprises a fusion module, a correction module and an output module. The step of inputting the first characteristic and the enemy characteristic into a trained energy consumption prediction network to obtain a processing energy consumption prediction result of the workpiece output by the energy consumption prediction network comprises the following steps:
inputting the first feature and the second feature into the fusion module to fuse the first feature and the second feature, and obtaining a first predicted feature output by the fusion module;
acquiring processing resource state information of the processing machine tool when the processing machine tool processes the workpiece, wherein the processing resource state information reflects the state of processing resources of the processing machine tool when the processing machine tool actually processes the workpiece, inputting the processing resource state information into the correction module to extract a feature reflecting the influence of the processing resource state on energy consumption, and acquiring a correction feature output by the correction module;
correcting the first prediction characteristic based on the correction characteristic to obtain a second prediction characteristic;
and inputting the second prediction characteristic into the output module, and acquiring the processing energy consumption prediction result output by the output module.
As can be seen from the foregoing description, the first characteristic and the second characteristic are both obtained based on a machining process file of the workpiece written in advance, that is, the first characteristic reflects a theoretical spindle speed of the machining tool when machining the workpiece, and the second characteristic reflects a theoretical cut amount of the machining tool when machining the workpiece. In actual production, the spindle rotation speed and the cutting amount are affected by the state of machining resources. Therefore, in order to make the machining energy consumption prediction result more accurate, in the numerical control machining energy consumption prediction method provided by the invention, after the first characteristic and the second characteristic are obtained, the first characteristic and the second characteristic are fused to obtain the first prediction characteristic, then the first prediction characteristic is corrected according to the machining resource state information in the actual machining process of the workpiece to obtain the second prediction characteristic, and the machining energy consumption prediction result is output based on the second prediction characteristic, so that the accuracy of energy consumption prediction can be improved.
Further, the first feature and the second feature are obtained based on a pre-written machining process file of the workpiece, and for workpieces with the same size and the same machining requirements, the machining process file may actually be shared, that is, for workpieces with the same machining process file, the first feature and the second feature may be calculated only once, and only the corresponding correction feature needs to be calculated when each workpiece is actually machined, and then the second prediction feature is obtained for prediction. After the first prediction feature is obtained, the first prediction feature and the processing technology workpiece corresponding to the first prediction feature are stored in a preset storage space, a processing technology file corresponding to a next workpiece is read when the processing machine performs processing of the next workpiece, if the processing technology file same as the processing technology file corresponding to the next workpiece exists in the preset storage space, the first prediction feature corresponding to the processing technology file is read from the preset storage space, and the first prediction feature stored in the preset storage space is reused for processing energy consumption prediction of the next workpiece. Therefore, the calculation amount can be greatly reduced, and the real-time energy consumption prediction in the processing process is realized.
Specifically, the machining resource includes the machine tool and the machining tool, and the machining resource state information reflects a degree of thermal deformation of the machine tool and the machining tool in a machining process and a degree of wear of the machining tool. The acquiring of the processing resource state information of the processing machine tool when processing the workpiece includes:
acquiring a temperature field signal of the processing machine tool when the workpiece is processed;
acquiring a vibration signal of the machining tool when the workpiece is machined;
and combining the temperature field signal and the vibration signal to obtain the processing resource state information.
The temperature field signal can be acquired based on a preset temperature sensor, and the machining tool is installed on the machining machine tool when numerical control machining is carried out, namely the temperature field signal of the machining machine tool can reflect the heating condition of the machining machine tool and the heating condition of the machining tool. The vibration signal can be acquired by a vibration sensor arranged on the cutter or a mounting device of the cutter. When the machining tool is worn, relative to a state without wear, large vibration is generated in the machining process, so that the vibration signal can reflect the wear degree of the machining tool. The heat generated by the processing machine tool in the processing process can cause the deformation of the main shaft and other transmission devices, the transmission efficiency is reduced, the energy consumption is increased, the heat generated and the abrasion of the processing tool in the processing process can cause the reduction of the cutting capacity, the main shaft is required to provide larger torque in the cutting process, and the energy consumption is increased. According to the method provided by the invention, the temperature field signal and the vibration signal are collected, the processing resource state information obtained by combination can reflect the influence of processing resources on energy consumption in the actual processing process, the correction characteristic is obtained based on the processing resource state information, and then the fusion result of the first characteristic and the second characteristic obtained based on the theoretical file is corrected according to the correction characteristic, so that the influence of heating of the processing resources and prop abrasion in the actual processing process can be considered on the basis of reducing the calculated amount in the energy consumption prediction process and realizing real-time prediction of energy consumption, and the accuracy of the energy consumption prediction result is ensured.
The first feature extraction network, the second feature extraction network and the energy consumption prediction network are obtained by training based on multiple groups of training data, wherein each group of training data comprises sample running rotating speed information, a sample voxel model, sample processing resource state information and a processing energy consumption label corresponding to a sample workpiece; the training process of the first feature extraction network, the second feature extraction network and the energy consumption prediction network comprises the following steps:
acquiring a plurality of groups of training data to form a first training batch, and updating parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network respectively based on each group of training data in the first training batch;
after parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are updated based on the last group of training data in the first training batch, selecting a plurality of training data in the first training batch to form a second training batch, and updating parameters of the correction module in the energy consumption prediction network based on each group of training data in the second training batch, wherein processing process files of the sample workpieces corresponding to each group of training data in the second training batch are the same;
and re-executing the step of obtaining the plurality of groups of training data to form the first training batch until the parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are converged.
In the method provided by the invention, the first feature extraction network, the second feature extraction network and the energy consumption prediction network are trained together. As described above, for a plurality of workpieces having the same process file, since the first feature and the second feature are extracted features input into the same network by using the parameters extracted from the same process file, the difference of the energy consumption prediction results for a plurality of workpieces having the same process file actually depends only on the modification process in the energy consumption prediction network. In order to improve the training efficiency of the first feature extraction network, the second feature extraction network and the energy consumption prediction network, based on the characteristics of the correction process in the energy consumption prediction network, in the method provided by the invention, when network training is carried out, after parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are updated by adopting a batch of training data, other network parameters are fixed, and only the parameter of the correction module in the energy consumption prediction network is updated based on the training data corresponding to a batch of sample workpieces of the same processing process file, so that the efficient training of the correction model can be realized, and the training efficiency of all network parameters is further improved.
The updating parameters of the first feature extraction network, the second feature extraction network, and the energy consumption prediction network based on each set of the training data in the first training batch respectively comprises:
obtaining a sample energy consumption prediction result output by the energy consumption prediction network based on sample running speed information, a sample voxel model and sample processing resource state information in a group of training data in the first training batch, obtaining a loss function based on a difference between the sample energy consumption prediction result and a processing energy consumption label in the group of training data, and updating parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network based on the loss function;
and re-executing the step of obtaining the energy consumption prediction result output by the energy consumption prediction network based on the sample running speed information, the sample voxel model and the sample processing resource state information in the training data in the training batch until each training data in the training batch is used for updating the parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network, or the parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network converge.
Updating the parameters of the correction module in the energy consumption prediction network based on each set of the training data in the second training batch includes:
acquiring a sample energy consumption prediction result output by the energy consumption prediction network based on sample running speed information, a sample voxel model and sample processing resource state information in a group of training data in the second training batch, acquiring a loss function based on a difference between the sample energy consumption prediction result and a processing energy consumption label in the group of training data, and updating parameters of the correction module based on the loss function;
and re-executing the step of obtaining the energy consumption prediction result of the sample output by the energy consumption prediction network based on the sample operation rotation speed information, the sample voxel model and the sample processing resource state information in the training data group in the second training batch until each training data group in the second training batch is used for updating the parameter of the correction module or the parameter convergence of the correction module.
After the training of the first feature extraction network, the second feature extraction network and the energy consumption prediction network is completed, the steps S100-S400 are executed based on the trained first feature extraction network, the trained second feature extraction network and the trained energy consumption prediction network, and numerical control machining energy consumption prediction is realized.
The numerical control machining energy consumption prediction device provided by the invention is described below, and the numerical control machining energy consumption prediction device described below and the numerical control machining energy consumption prediction method described above can be referred to correspondingly. As shown in fig. 2, the present invention provides a numerical control machining energy consumption prediction apparatus, including: a first feature extraction module 201, a voxel model generation module 202, a second feature extraction module 203, and an energy consumption prediction module 204.
The first feature extraction module 201 is used for obtaining operation rotating speed information of a processing machine tool, the operation rotating speed information reflects the change situation of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, the operation rotating speed information is input into a first feature extraction network to extract the rotating speed change feature of the main shaft of the processing machine tool, and the first feature output by the first feature extraction network is obtained.
The voxel model generating module 202 is configured to obtain a three-dimensional model of the workpiece, perform rasterization on the three-dimensional model to obtain three-dimensional lattices of the three-dimensional model, obtain machining process parameters of the workpiece, and add corresponding cutting information to the three-dimensional lattices according to the machining process parameters to obtain a voxel model, where the cutting information reflects a cutting amount when a machining tool moves into the three-dimensional lattices to perform cutting;
the second feature extraction module 203 is configured to input the voxel model to a trained second feature extraction network to extract features of the voxel model, and obtain second features output by the voxel model feature extraction network;
the energy consumption prediction module 204 is configured to input the first characteristic and the second characteristic into a trained energy consumption prediction network, and obtain a processing energy consumption prediction result of the workpiece output by the energy consumption prediction network, where the energy consumption prediction network is obtained by training based on multiple sets of data including processing energy consumption labels.
Fig. 3 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 3: a processor (processor) 310, a communication Interface (communication Interface) 320, a memory (memory) 330 and a communication bus 340, wherein the processor 310, the communication Interface 320 and the memory 330 communicate with each other via the communication bus 340. The processor 310 may call logic instructions in the memory 330 to perform a method of numerical control machining energy consumption prediction, the method comprising: acquiring operation rotating speed information of a processing machine tool, wherein the operation rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, inputting the operation rotating speed information into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and acquiring a first characteristic output by the first characteristic extraction network;
acquiring a three-dimensional model of the workpiece, rasterizing the three-dimensional model to obtain each three-dimensional grid of the three-dimensional model, acquiring machining process parameters of the workpiece, and adding corresponding cutting information in the three-dimensional grids according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves into the three-dimensional grids for cutting;
inputting the voxel model into a trained second feature extraction network to extract features of the voxel model, and acquiring second features output by the voxel model feature extraction network;
inputting the first characteristic and the second characteristic into a trained energy consumption prediction network, and obtaining a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, wherein the energy consumption prediction network is obtained by training based on a plurality of groups of data including machining energy consumption labels.
In addition, the logic instructions in the memory 330 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute 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), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, the computer program being implemented by a processor to execute the numerical control machining energy consumption prediction method provided by the above methods, the method including: acquiring operation rotating speed information of a processing machine tool, wherein the operation rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, inputting the operation rotating speed information into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and acquiring a first characteristic output by the first characteristic extraction network;
acquiring a three-dimensional model of the workpiece, rasterizing the three-dimensional model to obtain each three-dimensional grid of the three-dimensional model, acquiring machining process parameters of the workpiece, and adding corresponding cutting information in the three-dimensional grids according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves into the three-dimensional grids for cutting;
inputting the voxel model into a trained second feature extraction network to extract features of the voxel model, and acquiring second features output by the voxel model feature extraction network;
inputting the first characteristic and the second characteristic into a trained energy consumption prediction network, and obtaining a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, wherein the energy consumption prediction network is obtained by training based on a plurality of groups of data including machining energy consumption labels.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present 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 solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A numerical control machining energy consumption prediction method is characterized by comprising the following steps:
acquiring operation rotating speed information of a processing machine tool, wherein the operation rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, inputting the operation rotating speed information into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and acquiring a first characteristic output by the first characteristic extraction network;
acquiring a three-dimensional model of the workpiece, performing rasterization processing on the three-dimensional model to obtain each three-dimensional grid of the three-dimensional model, acquiring machining process parameters of the workpiece, and adding corresponding cutting information in the three-dimensional grids according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves to the three-dimensional grids for cutting;
inputting the voxel model into a trained second feature extraction network to extract features of the voxel model, and acquiring second features output by the voxel model feature extraction network;
inputting the first characteristic and the second characteristic into a trained energy consumption prediction network, and obtaining a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, wherein the energy consumption prediction network is obtained by training based on a plurality of groups of data including machining energy consumption labels.
2. The numerical control machining energy consumption prediction method according to claim 1, wherein the energy consumption prediction network comprises a fusion module, a correction module and an output module; the inputting the first characteristic and the second characteristic into a trained energy consumption prediction network to obtain a processing energy consumption prediction result of the workpiece output by the energy consumption prediction network includes:
inputting the first feature and the second feature into the fusion module to fuse the first feature and the second feature, and acquiring a first predicted feature output by the fusion module;
acquiring processing resource state information of the processing machine tool when the processing machine tool processes the workpiece, wherein the processing resource state information reflects the state of processing resources of the processing machine tool when the processing machine tool actually processes the workpiece, inputting the processing resource state information into the correction module to extract a feature reflecting the influence of the processing resource state on energy consumption, and acquiring a correction feature output by the correction module;
correcting the first prediction characteristic based on the correction characteristic to obtain a second prediction characteristic;
and inputting the second prediction characteristic into the output module, and acquiring the processing energy consumption prediction result output by the output module.
3. The numerical control machining energy consumption prediction method according to claim 2, wherein the first feature extraction network, the second feature extraction network and the energy consumption prediction network are obtained by training based on a plurality of groups of training data, and each group of training data comprises sample running speed information, a sample voxel model, sample machining resource state information and a machining energy consumption label corresponding to a sample workpiece; the training process of the first feature extraction network, the second feature extraction network and the energy consumption prediction network comprises the following steps:
acquiring a plurality of groups of training data to form a first training batch, and updating parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network respectively based on each group of training data in the first training batch;
after parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are updated based on the last group of training data in the first training batch, selecting a plurality of training data in the first training batch to form a second training batch, and updating parameters of the correction module in the energy consumption prediction network based on each group of training data in the second training batch, wherein processing process files of the sample workpieces corresponding to each group of training data in the second training batch are the same;
and re-executing the step of obtaining the plurality of groups of training data to form the first training batch until the parameters of the first feature extraction network, the second feature extraction network and the energy consumption prediction network are converged.
4. The numerical control machining energy consumption prediction method according to claim 2, wherein the machining resource includes the machine tool and the machining tool, and the machining resource state information reflects a degree of thermal deformation of the machine tool and the machining tool during machining and a degree of wear of the machining tool.
5. The numerical control machining energy consumption prediction method according to claim 3, wherein the acquiring of the machining resource state information of the machining tool when machining the workpiece comprises:
acquiring a temperature field signal of the processing machine tool when the workpiece is processed;
acquiring a vibration signal of the machining tool when the workpiece is machined;
and combining the temperature field signal and the vibration signal to obtain the processing resource state information.
6. The numerical control machining energy consumption prediction method according to claim 1, wherein the obtaining of the operating rotation speed information of the machining tool comprises:
and acquiring a processing technology file of the workpiece, and acquiring the running rotating speed information of the processing machine tool based on the processing technology file.
7. The numerical control machining energy consumption prediction method according to claim 1, wherein the cutting information includes a feed amount and a motion trajectory of the machining tool.
8. A numerical control machining energy consumption prediction apparatus, comprising:
the first characteristic extraction module is used for acquiring running rotating speed information of a processing machine tool, the running rotating speed information reflects the change condition of the rotating speed of a main shaft in the process of processing a workpiece by the processing machine tool, the running rotating speed information is input into a first characteristic extraction network to extract the rotating speed change characteristic of the main shaft of the processing machine tool, and first characteristics output by the first characteristic extraction network are acquired;
the voxel model generating module is used for acquiring a three-dimensional model of the workpiece, rasterizing the three-dimensional model to obtain each three-dimensional lattice of the three-dimensional model, acquiring machining process parameters of the workpiece, and adding corresponding cutting information into the three-dimensional lattices according to the machining process parameters to obtain a voxel model, wherein the cutting information reflects the cutting amount when a machining tool moves into the three-dimensional lattices to cut;
the second feature extraction module is used for inputting the voxel model to a trained second feature extraction network to extract the features of the voxel model and acquiring second features output by the voxel model feature extraction network;
the energy consumption prediction module is used for inputting the first characteristic and the second characteristic into a trained energy consumption prediction network to obtain a machining energy consumption prediction result of the workpiece output by the energy consumption prediction network, and the energy consumption prediction network is obtained based on a plurality of groups of data including machining energy consumption labels through training.
9. 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 nc machining energy consumption prediction method according to any one of claims 1 to 7 when executing the program.
10. A non-transitory computer readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the nc machining energy consumption prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310240482.7A CN115951634B (en) | 2023-03-14 | 2023-03-14 | Numerical control processing energy consumption prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310240482.7A CN115951634B (en) | 2023-03-14 | 2023-03-14 | Numerical control processing energy consumption prediction method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115951634A true CN115951634A (en) | 2023-04-11 |
| CN115951634B CN115951634B (en) | 2023-05-23 |
Family
ID=85893044
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310240482.7A Active CN115951634B (en) | 2023-03-14 | 2023-03-14 | Numerical control processing energy consumption prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115951634B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016101624A1 (en) * | 2014-12-26 | 2016-06-30 | 重庆大学 | Machining workshop workpiece energy consumption quota formulation method |
| US20200201282A1 (en) * | 2017-06-26 | 2020-06-25 | Jiangnan University | Energy consumption prediction system and method based on the decision tree for CNC lathe turning |
| CN113110355A (en) * | 2021-04-29 | 2021-07-13 | 武汉科技大学 | Method for predicting and optimizing machining energy consumption of workpiece driven by digital twin |
| CN114611379A (en) * | 2022-02-10 | 2022-06-10 | 北京工业大学 | Machining process energy-saving planning method based on data driving |
| CN115167279A (en) * | 2022-09-07 | 2022-10-11 | 中科航迈数控软件(深圳)有限公司 | Energy consumption prediction method and system for numerical control machine tool and related equipment |
| CN115755758A (en) * | 2022-12-19 | 2023-03-07 | 重庆忽米网络科技有限公司 | Machine tool machining control method based on neural network model |
-
2023
- 2023-03-14 CN CN202310240482.7A patent/CN115951634B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016101624A1 (en) * | 2014-12-26 | 2016-06-30 | 重庆大学 | Machining workshop workpiece energy consumption quota formulation method |
| US20200201282A1 (en) * | 2017-06-26 | 2020-06-25 | Jiangnan University | Energy consumption prediction system and method based on the decision tree for CNC lathe turning |
| CN113110355A (en) * | 2021-04-29 | 2021-07-13 | 武汉科技大学 | Method for predicting and optimizing machining energy consumption of workpiece driven by digital twin |
| CN114611379A (en) * | 2022-02-10 | 2022-06-10 | 北京工业大学 | Machining process energy-saving planning method based on data driving |
| CN115167279A (en) * | 2022-09-07 | 2022-10-11 | 中科航迈数控软件(深圳)有限公司 | Energy consumption prediction method and system for numerical control machine tool and related equipment |
| CN115755758A (en) * | 2022-12-19 | 2023-03-07 | 重庆忽米网络科技有限公司 | Machine tool machining control method based on neural network model |
Non-Patent Citations (3)
| Title |
|---|
| QINGE XIAO等: "An industrial data based investigation into effects of process parameters on cutting power and energy efficiency", 《2017 13TH IEEE CONFERENCE ON AUTOMATION SCIENCE AND ENGINEERING (CASE)》 * |
| QINGE XIAO等: "Deep Learning Based Modeling for Cutting Energy Consumed in CNC Turning Process", 《2018 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS (SMC)》 * |
| 李聪波 等, 《机械工程学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115951634B (en) | 2023-05-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Application framework of digital twin-driven product smart manufacturing system: A case study of aeroengine blade manufacturing | |
| US20220379380A1 (en) | Hybrid additive and subtractive manufacturing | |
| KR102229859B1 (en) | Operation management method, apparatus and system using machine learning and transfer learning | |
| CN104252153A (en) | CNC (computer numerical control) processing program generating system and method | |
| KR20210062440A (en) | Manufacturing apparatus of machine tool using digital twin and the method thereof | |
| CN108581384A (en) | A kind of four axis turn-milling cutting method of monoblock type impeller based on UG and Vericut | |
| CN116021339B (en) | Method and related device for monitoring cutting force of main shaft of numerical control machine tool | |
| US20200324413A1 (en) | Method and system for predicting collision of machining path | |
| Adjoul et al. | Algorithmic strategy for optimizing product design considering the production costs | |
| Ganser et al. | Knowledge-based adaptation of product and process design in blisk manufacturing | |
| CN110402188A (en) | Drive system and its assessment | |
| CN115951634A (en) | Numerical control machining energy consumption prediction method and device, electronic equipment and storage medium | |
| CN116339242B (en) | Free-form surface cutter path generation method and related equipment | |
| US10762699B2 (en) | Machining parameter automatic generation system | |
| CN112916883A (en) | Cutter relieving deformation prediction method for turning machining of navigation shaft | |
| CN105005210A (en) | Mechanical-electrical integration simulation system and use method thereof | |
| CN112666895B (en) | Numerical control machining speed planning method and system based on double-code combined action | |
| Yusof et al. | Computer aided process planning: a comprehensive survey | |
| EP3832415A1 (en) | Automation system, tracking device of said automation system, and method for controlling the same | |
| CN102819237A (en) | Method for generating simulation target blank in solid milling simulation process | |
| JP5763352B2 (en) | NC program creation device | |
| Nagata et al. | Development of post-processor approach for an industrial robot FANUC R2000 i C | |
| CN110990998B (en) | Intelligent manufacturing process system for gearbox body | |
| CN116400905B (en) | Code automatic generation method for regulating and controlling multiple devices and related devices | |
| CN114160847B (en) | Variable-rotation-speed processing method, system, equipment and medium |
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 |