CN115809523A - Cutting parameter prediction processing method and device, electronic equipment and storage medium - Google Patents
Cutting parameter prediction processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115809523A CN115809523A CN202211563596.7A CN202211563596A CN115809523A CN 115809523 A CN115809523 A CN 115809523A CN 202211563596 A CN202211563596 A CN 202211563596A CN 115809523 A CN115809523 A CN 115809523A
- Authority
- CN
- China
- Prior art keywords
- sample data
- prediction model
- verticality
- cut
- cutting
- 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.)
- Pending
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
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application provides a cutting parameter prediction processing method, a device, an electronic device and a storage medium, wherein the method comprises the following steps: the method comprises the steps of obtaining technological parameters to be cut, inputting the technological parameters to be cut into a target prediction model obtained through pre-training, obtaining the prediction non-verticality corresponding to the technological parameters to be cut, and verifying whether the technological parameters to be cut meet the requirement of the target non-verticality or not by utilizing the prediction non-verticality. The non-perpendicularity of the process parameter to be cut can be predicted through the target prediction model obtained through pre-training, and the non-perpendicularity of the process parameter to be cut can be accurately and effectively predicted, so that the optimal cutting parameter can be determined according to the non-perpendicularity obtained through prediction, the optimal non-perpendicularity can be cut, and the efficiency and the accuracy of determining the cutting process parameter are improved.
Description
Technical Field
The present application relates to the field of prediction models, and in particular, to a method and an apparatus for processing a cutting parameter prediction, an electronic device, and a storage medium.
Background
For the steel plate structure industry, the cutting quality of the steel plate cutting workpiece determines the workload and quality of the subsequent assembly and welding of other workpieces, so that the process parameters of cutting are analyzed and predicted, the production workload can be effectively reduced, and the quality of a steel structure during the subsequent welding can be controlled. The air plasma cutting machine has wide application in steel plate cutting, when the steel plate is cut, the non-verticality of the cut is influenced by the current intensity, the cutting speed, the gas effect and the distance between a plasma gun and a metal thick plate, and the non-verticality of the cut of the thick plate is predicted through the cutting parameters.
In the prior art, a large number of experimental databases are used to find out the corresponding non-verticality from the experimental databases according to the cutting parameters.
However, such a method requires a lot of experiments, which results in waste of material cost and time cost, and in addition, with such a method, the relationship between data is complicated, so that it is not easy to find out the optimal cutting parameter, which may result in failure to find out the optimal non-verticality.
Disclosure of Invention
An object of the present application is to provide a method and an apparatus for processing cutting parameter prediction, an electronic device, and a storage medium, which improve accuracy of cutting process parameter prediction.
In order to achieve the above purpose, the embodiments of the present application adopt the following technical solutions:
in a first aspect, an embodiment of the present application provides a cutting parameter prediction processing method, where the method includes:
acquiring technological parameters to be cut, wherein the technological parameters to be cut comprise: current intensity, cutting speed, and gas pressure;
inputting the technological parameters to be cut into a target prediction model obtained by pre-training to obtain the predicted non-verticality corresponding to the technological parameters to be cut, and verifying whether the technological parameters to be cut meet the requirement of the target non-verticality or not by using the predicted non-verticality.
Optionally, before the step of inputting the to-be-cut process parameter into a target prediction model obtained through pre-training to obtain the predicted non-perpendicularity corresponding to the to-be-cut process parameter, the method further includes:
obtaining a plurality of sample data groups and labeling non-verticality corresponding to each sample data group, wherein each sample data group respectively comprises: current intensity, cutting speed and gas pressure;
inputting each sample data set into an initial prediction model to obtain the prediction non-verticality corresponding to each sample data set;
and performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain the target prediction model.
Optionally, obtaining a plurality of sample data sets and the non-verticality of the label corresponding to each sample data set includes:
determining different cutting process parameters according to different current intensities, cutting speeds and gas pressures, and forming the different cutting process parameters into the plurality of sample data sets;
and obtaining the marked non-verticality corresponding to each sample data group according to the actual non-verticality obtained by cutting according to each sample data group.
Optionally, the initial prediction model includes a hidden layer and an output layer, where the hidden layer includes a plurality of hidden nodes;
inputting each sample data set into an initial prediction model to obtain the predicted non-verticality corresponding to each sample data set, including:
respectively inputting the current intensity, the cutting speed and the gas pressure in the sample data set into each input node of the hidden layer;
determining an output value of the hidden layer according to the input function and the output function of the hidden layer;
and inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to the sample data set.
Optionally, the determining an output value of the hidden layer according to the input function and the output function of the hidden layer includes:
determining a function value of an input function of the hidden layer according to the sample data set, the number of neurons in an input layer, the weight between the input neuron and the hidden neuron, the weight between the hidden neuron and the neuron in an output layer, the number of hidden neurons and the bias of hidden nodes;
determining a function value of an output function of the hidden layer according to the function value of the input function of the hidden layer and a linear function;
and determining the output value of the hidden layer according to the function value of the output function of the hidden layer.
Optionally, the inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to the sample data includes:
determining a function value of an input function of the output layer according to the weight between the hidden layer and the output layer, the output value of the hidden layer and the bias of the node of the output layer;
determining an output function of the output layer according to the function value of the input function of the output layer and the activation function;
and obtaining the predicted non-verticality corresponding to the sample data set according to the output function.
Optionally, the performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain the target prediction model includes:
determining a relative error between the predicted non-verticality and the marked non-verticality according to the predicted non-verticality and the marked non-verticality corresponding to the sample data set;
determining whether the initial prediction model is an optimal model or not according to the relative error and a preset threshold range, and if so, taking the initial prediction model as a target prediction model;
if not, updating parameters in the initial prediction model to obtain a new initial prediction model, inputting each sample data set into the new initial prediction model again, re-determining the relative error, re-judging whether the new initial prediction model is the optimal model, performing iteration until an iteration stop condition is met, and taking the new initial prediction model meeting the iteration stop condition as a target prediction model.
In a second aspect, an embodiment of the present application further provides a cutting parameter prediction processing apparatus, where the apparatus includes:
the acquisition module is used for acquiring the technological parameters to be cut, and the technological parameters to be cut comprise: current intensity, cutting speed, and gas pressure;
and the input module is used for inputting the process parameters to be cut into a target prediction model obtained by pre-training to obtain the predicted non-verticality corresponding to the process parameters to be cut so as to verify whether the process parameters to be cut meet the requirement of the target non-verticality by using the predicted non-verticality.
Optionally, the input module is specifically configured to:
obtaining a plurality of sample data groups and labeling non-verticality corresponding to each sample data group, wherein each sample data group respectively comprises: current intensity, cutting speed, and gas pressure;
inputting each sample data set into an initial prediction model to obtain the prediction non-verticality corresponding to each sample data set;
and performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain the target prediction model.
Optionally, the obtaining module is specifically configured to:
determining different cutting process parameters according to different current intensities, cutting speeds and gas pressures, and forming the different cutting process parameters into the plurality of sample data groups;
and obtaining the marked non-verticality corresponding to each sample data group according to the actual non-verticality obtained by cutting according to each sample data group.
Optionally, the initial prediction model includes a hidden layer and an output layer, where the hidden layer includes a plurality of hidden nodes;
optionally, the input module is specifically configured to:
respectively inputting the current intensity, the cutting speed and the gas pressure in the sample data set into each input node of the hidden layer;
determining an output value of the hidden layer according to the input function and the output function of the hidden layer;
and inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to the sample data set.
Optionally, the input module is specifically configured to:
determining a function value of an input function of the hidden layer according to the sample data set, the number of neurons in an input layer, the weight between the input neuron and the hidden neuron, the weight between the hidden neuron and the neuron in an output layer, the number of hidden neurons and the bias of hidden nodes;
determining a function value of an output function of the hidden layer according to the function value of the input function of the hidden layer and a linear function;
and determining the output value of the hidden layer according to the function value of the output function of the hidden layer.
Optionally, the input module is specifically configured to:
determining a function value of an input function of the output layer according to the weight between the hidden layer and the output layer, the output value of the hidden layer and the bias of the node of the output layer;
determining an output function of the output layer according to the function value of the input function of the output layer and the activation function;
and obtaining the predicted non-verticality corresponding to the sample data set according to the output function.
Optionally, the input module is specifically configured to:
determining a relative error between the predicted non-verticality and the marked non-verticality according to the predicted non-verticality and the marked non-verticality corresponding to the sample data set;
determining whether the initial prediction model is an optimal model or not according to the relative error and a preset threshold range, and if so, taking the initial prediction model as a target prediction model;
if not, updating parameters in the initial prediction model to obtain a new initial prediction model, inputting each sample data set into the new initial prediction model again, re-determining the relative error, re-judging whether the new initial prediction model is the optimal model, performing iteration until an iteration stop condition is met, and taking the new initial prediction model meeting the iteration stop condition as a target prediction model.
In a third aspect, an embodiment of the present application further provides an electronic device, including: the cutting parameter prediction processing method comprises a processor, a storage medium and a bus, wherein the storage medium stores program instructions executable by the processor, when an application program runs, the processor and the storage medium communicate through the bus, and the processor executes the program instructions to execute the steps of the cutting parameter prediction processing method according to the first aspect.
In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and the computer program is read and executes the steps of the cutting parameter prediction processing method according to the first aspect.
The beneficial effect of this application is:
according to the cutting parameter prediction processing method, the device, the electronic equipment and the storage medium, the process parameter to be cut is obtained, the process parameter to be cut is input into the target prediction model obtained through pre-training, the predicted non-verticality corresponding to the process parameter to be cut is obtained, and whether the process parameter to be cut meets the requirement of the target non-verticality or not is verified through the predicted non-verticality. The non-perpendicularity of the process parameter to be cut can be predicted through the target prediction model obtained through pre-training, and the non-perpendicularity of the process parameter to be cut can be accurately and effectively predicted, so that the optimal cutting parameter can be determined according to the non-perpendicularity obtained through prediction, the optimal non-perpendicularity can be cut, and the efficiency and the accuracy of determining the cutting process parameter are improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a schematic flowchart of a cutting parameter prediction processing method according to an embodiment of the present disclosure;
fig. 2 is a complete flowchart of a method for training a target prediction model according to an embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a prediction model provided in an embodiment of the present application;
FIG. 4 is a schematic flow chart illustrating another method for training an initial prediction model according to an embodiment of the present disclosure;
FIG. 5 is a schematic diagram illustrating a complete flow of a method for training an initial prediction model according to an embodiment of the present application;
fig. 6 is a schematic diagram of an apparatus for predicting a cutting parameter according to an embodiment of the present disclosure;
fig. 7 is a block diagram of an electronic device according to an embodiment of the present disclosure.
Detailed Description
In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it should be understood that the drawings in the present application are for illustrative and descriptive purposes only and are not used to limit the scope of protection of the present application. Additionally, it should be understood that the schematic drawings are not necessarily drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of the present application. It should be understood that the operations of the flow diagrams may be performed out of order, and steps without logical context may be performed in reverse order or simultaneously. One skilled in the art, under the guidance of this application, may add one or more other operations to, or remove one or more operations from, the flowchart.
In addition, the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that in the embodiments of the present application, the term "comprising" is used to indicate the presence of the features stated hereinafter, but does not exclude the addition of further features.
For the steel structure industry, the cutting instruction of the steel plate cutting workpiece determines the workload and the quality of the assembly and the welding of other subsequent workpieces, so that the processing parameters of cutting are analyzed and predicted, the production workload can be effectively reduced, and the instruction of a steel structure during subsequent welding can be effectively controlled. The air plasma cutting machine has low economic cost and wide application prospect, and can realize excellent measurement among cutting instructions, cutting speed and cost, so the air plasma cutting machine has wide application.
When the air plasma cutting machine cuts the steel plate, the non-verticality of the notch is influenced by the current intensity, the cutting speed, the gas pressure and the distance between the plasma gun and the metal thick plate, and when any cutting process parameter is changed, the non-verticality of the notch is changed. Therefore, further determination of the process parameters of air plasma cutting is required to optimize the non-perpendicularity of the cut of the thick plate, and thus the assembly time and welding quality of the subsequent other workpieces.
The method is based on the self-developed three-axis numerical control air plasma cutting equipment, and analysis and prediction are carried out on the steel plate cutting process parameters. Because the cutting process parameters are more and the optional parameter range is larger, a large amount of experiments are needed to find out the corresponding non-perpendicularity according to the cutting process parameters, analysis is carried out according to the obtained experimental data, so that waste of material cost and time cost can be caused, the relation among various groups of data is complex, and the optimal cutting parameters are not easy to find out.
The method in the embodiment of the present application may be applied to an electronic device, which may be a terminal device having a computing processing capability and a display function, such as a desktop computer and a notebook computer, or may also be a server. When the electronic equipment receives the cutting parameters, the method in the embodiment of the application is used for predicting and obtaining the non-verticality corresponding to the cutting parameters.
Fig. 1 is a cutting parameter prediction processing method provided in an embodiment of the present application, where the method is applied to the foregoing electronic device, and the method may include:
s101, acquiring technological parameters to be cut.
The parameters of the process to be cut may include current intensity, cutting speed and gas pressure.
Optionally, the current intensity refers to the current intensity when the plasma cutting device cuts, and for example, the current intensity may be 50A, 55A, 60A, and other current intensity values; the cutting speed refers to the speed of the plasma cutting equipment during cutting, and can be, for example, cutting speed values of different values such as 20cm/min,30cm/min,40cm/min and the like; the other gas refers to the gas pressure when the plasma cutting equipment cuts, and for example, the gas pressure can be 6Bar, 7Bar, 8Bar and other gas pressure values with different values.
Optionally, when the plasma cutting apparatus performs cutting, the required process parameters to be cut refer to all data consisting of the current intensity, the cutting speed and the gas pressure, specifically, different current intensities, cutting speeds and gas pressures can be used as one process parameter to be cut, and the corresponding non-verticality is predicted according to the process parameter to be cut.
For example, a set of process parameters to be cut may be: the current intensity is 50A, the cutting speed is 20cm/min, and the gas pressure is 6Bar; another set of process parameters to be cut may be, for example: the current intensity is 50A, the cutting speed is 30cm/min, and the gas pressure is 7Bar; another set of process parameters to be cut may be, for example: the current intensity is 55A, the cutting speed is 20cm/min, and the gas pressure is 5Bar.
S102, inputting the technological parameters to be cut into a target prediction model obtained through pre-training, and obtaining the predicted non-verticality corresponding to the technological parameters to be cut.
Optionally, when the predicted non-perpendicularity corresponding to the process parameter to be cut is obtained through the target prediction model, whether the process parameter to be cut meets the requirement of the target non-perpendicularity can be verified by using the predicted non-perpendicularity, wherein the target non-perpendicularity refers to the optimal non-perpendicularity for cutting by the plasma cutting equipment. If the target non-perpendicularity is met, the plasma cutting equipment can be controlled to cut the steel plate according to the to-be-cut technological parameters corresponding to the predicted non-perpendicularity; if the target non-perpendicularity is not met, the predicted non-perpendicularity is not the optimal non-perpendicularity, cutting cannot be carried out according to the to-be-cut technological parameters corresponding to the predicted non-perpendicularity, and other to-be-cut technological parameters can be reselected for prediction.
In the embodiment, the to-be-cut technological parameters are acquired and input into the pre-trained target prediction model to acquire the predicted non-perpendicularity corresponding to the to-be-cut technological parameters, so that whether the to-be-cut technological parameters meet the target non-perpendicularity requirement is verified by utilizing the predicted non-perpendicularity. The non-perpendicularity of the process parameter to be cut can be predicted through the target prediction model obtained through pre-training, and the non-perpendicularity of the process parameter to be cut can be accurately and effectively predicted, so that the optimal cutting parameter can be determined according to the non-perpendicularity obtained through prediction, the optimal non-perpendicularity can be cut, and the efficiency and the accuracy of determining the cutting process parameter are improved.
Fig. 2 is a schematic diagram of a training method of a target prediction model according to an embodiment of the present application, and as shown in fig. 2, before inputting a to-be-cut process parameter into a target prediction model obtained by pre-training in step S102 to obtain a predicted non-verticality corresponding to the to-be-cut process parameter, the method may include:
s201, obtaining a plurality of sample data groups and labeling non-verticality corresponding to each sample data group.
Wherein each sample data group respectively comprises current intensity, cutting speed and gas pressure; the marked non-verticality corresponding to each sample data set refers to the actual non-verticality obtained by measurement after each sample data set is actually cut.
Optionally, the value range of the current intensity in the sample data set may be set to be 43A to 60A; the value range of the cutting speed can be 20 cm/min-40 cm/min; the value range of the gas pressure can be 3.0 Bar-6.5 Bar.
Optionally, 50 groups of sample data groups may be obtained, and the plurality of sample data groups are divided into three parts, where a part of the sample data groups may be used for training the initial prediction model, for example, seventy percent of sample data may be selected as a training sample data group; a portion of the sample data set may be used to validate the trained initial prediction model, for example, fifteen percent of the sample data may be selected as the validation training sample data set; a portion of the sample data set may be used to test the trained initial predictive model, for example fifteen percent of the sample data set may be selected as the test sample data set. Other ratios can be selected for the data ratio to be distributed, and the embodiment of the application is not limited to this.
S202, inputting each sample data set into the initial prediction model to obtain the prediction non-verticality corresponding to each sample data set.
The sample data set may be the training sample data set in S201, the training sample data set is input to the initial prediction model, and an algorithm in the initial prediction model is used to perform calculation according to the input sample data set, so as to obtain the predicted non-verticality corresponding to each sample data set.
Illustratively, a predicted non-perpendicularity 1 may be obtained for the sample data set 1; the predicted non-verticality 2 can be obtained for the sample data set 2; the predicted non-perpendicularity 3 can be obtained for the set of sample data 3.
And S203, carrying out iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain a target prediction model.
Optionally, according to the predicted non-verticality corresponding to each sample array in S202 and the labeled non-verticality of each sample data array obtained in S101, a preset method is adopted to perform iterative optimization on the initial prediction model, and the initial prediction model meeting the iteration stop condition is used as the target prediction model.
In this embodiment, the optimized target prediction model is obtained by training the initial training model through the obtained sample data, so that the non-verticality obtained based on the prediction of the trained target prediction model is more accurate, and the corresponding non-verticality is not found from the test database according to the cutting process parameters through a large number of test databases.
Optionally, the obtaining of the plurality of sample data sets and the non-verticality of the label corresponding to each sample data set in step S201 may include:
optionally, different cutting process parameters are determined according to different current intensities, cutting speeds and gas pressures, and the different cutting process parameters are combined into a plurality of sample data sets.
For example, the current intensity 43A, the cutting speed 20cm/min, and the gas pressure 6.5Bar may be a set of sample data; the current intensity 43A, the cutting speed 30cm/min and the gas pressure 5Bar can be a group of sample data; the amperage 43A, cutting speed 40cm/min, gas pressure 5Bar may be a set of sample data set amperage 46A, cutting speed 25cm/min, gas pressure 4.7Bar may be a set of sample data set.
Optionally, obtaining an actual non-perpendicularity obtained by cutting each sample data group to obtain a labeled non-perpendicularity corresponding to each sample data group, specifically, for different sample data groups, cutting a 10mm steel plate by using a plasma cutting device at a height of 5.1mm between a cutting gun and the steel plate, measuring the non-perpendicularity of a cut obtained by cutting each sample data group after cutting, and taking the measured non-perpendicularity as the actual non-perpendicularity of each sample data group, that is, the labeled non-perpendicularity corresponding to each sample data group; specifically, the plasma cutting equipment can be used for cutting the steel plate according to 50 groups of different cutting process parameters, so as to obtain 50 groups of sample data sets.
Optionally, the initial prediction model in the above description may include a hidden layer and an output layer, where the hidden layer may include a plurality of hidden nodes, and a schematic structural diagram of the initial prediction model is shown in fig. 3, and the schematic structural diagram of the initial prediction model is composed of 3 input nodes, a hidden layer including 10 hidden nodes, and an output layer including an output node.
Optionally, the step S202 of inputting each sample data set into the initial prediction model to obtain the predicted non-perpendicularity corresponding to each sample data set may include:
optionally, the current intensity, the cutting speed, and the gas pressure in the sample data set are respectively input into each input node in the hidden layer, the hidden layer may include 3 input nodes, such as input node 1, input node 2, and input node 3, the current intensity may be input from the input node 1, the cutting speed may be input from the input node 2, and the gas pressure may be input from the input node 3.
Optionally, the output value of the hidden layer is determined according to the input function and the output function of the hidden layer, the output value of the hidden layer is input into the output layer, and the predicted non-verticality corresponding to each sample data set is obtained.
Optionally, the determining an output value of the hidden layer according to the input function and the output function of the hidden layer may include:
optionally, determining a function value of an input function of the hidden layer according to the sample data set, the number of neurons in the input layer, the weight between the neurons in the input layer and the hidden neurons, the weight between the neurons in the hidden layer and the neurons in the output layer, the number of the hidden neurons, and the bias of hidden nodes, where the input layer refers to 3 input nodes; specifically, the expression (one) may be used.
Wherein J is the number of neurons in the input layer, C i，k Is the weight between input layer neurons and hidden neurons, is the inputCutting process parameters including current intensity, cutting speed and gas pressure, b k For the bias of hidden layer nodes, k is the number of hidden layer neurons, and the pointer _ layer is the function value of the hidden layer's input function.
Alternatively, the function value of the output function of the hidden layer is determined according to the function value of the input function of the hidden layer and the linear function, and the output value of the hidden layer is determined according to the function value of the output function of the hidden layer, and then the output value is the function value of the output layer of the hidden layer, and specifically, the output value can be expressed by using the formula (two).
h k ＝f 1 (ladder _ layer) equation (two)
Wherein f is 1 Is a sigmoid function of the activation function, h k Is the output value of the hidden layer node.
Optionally, the inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to each sample data set may include:
alternatively, the function value of the input function of the output layer is determined according to the weight between the hidden layer and the output layer, the output value of the hidden layer, and the offset of the output layer node, and specifically, may be represented by formula (three).
Wherein D is k，z Is the weight between hidden layer neurons and output layer neurons, h k Is the output value of the hidden layer, g z Is the offset of the output layer node, and z refers to the number of output layer neurons.
Alternatively, the output function of the output layer is determined from the function value of the input function of the output layer and the activation function, and may be specifically expressed using formula (iv).
o z ＝f 2 (output _ layer) formula (IV)
Wherein, f 2 Is a linear function, O z Is the output value of the output function.
Optionally, the predicted non-perpendicularity corresponding to the sample data set is obtained according to the output function, where the predicted non-perpendicularity refers to an output value of the output function.
Fig. 4 is a schematic flowchart of another method for training an initial prediction model according to an embodiment of the present application, and as shown in fig. 4, in step S203, the performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain a target prediction model may include:
s301, determining the relative error between the prediction non-perpendicularity and the labeling non-perpendicularity according to the prediction non-perpendicularity and the labeling non-perpendicularity corresponding to each sample data set.
The relative error may be specifically expressed by using formula (five).
S302, determining whether the initial prediction model is the optimal model or not according to the relative error and a preset threshold range, if so, executing the step S303, and if not, executing the step S304.
And S303, taking the initial prediction model as a target prediction model.
Specifically, if the calculated relative error is smaller than a preset threshold range, the initial prediction model is determined to be the optimal model, and the initial prediction model is used as the target prediction model.
And S304, if not, updating parameters in the initial prediction model to obtain a new initial prediction model, inputting each sample data set into the new initial prediction model again, determining relative errors again, judging whether the new initial prediction model is the optimal model again, performing iteration until an iteration stop condition is met, and taking the new initial prediction model meeting the iteration stop condition as a target prediction model.
Specifically, if the calculated relative error is larger than the preset threshold range, updating parameters in the initial prediction model, for example, updating weight parameters of a hidden layer and an output layer in the initial prediction model, taking the updated initial prediction model as a new initial prediction model, inputting each sample data set into the new initial prediction model again, re-determining the relative error, determining whether the new initial prediction model is optimal, performing iteration until an iteration stop condition is met, and taking the new initial prediction model when the iteration stop condition is met as a target prediction model. The iteration stopping condition can be that the iteration number reaches a preset iteration number, and at the moment, the iteration is stopped, and a new initial prediction model when the iteration is stopped is used as a target prediction model; the new initial prediction model can be an optimal prediction model, at the moment, iteration is stopped, and the optimal prediction model when iteration is stopped is used as a target prediction model.
Fig. 5 is a schematic diagram of a complete process of a method for training an initial prediction model according to an embodiment of the present application, and as shown in fig. 5, the complete training process for the initial prediction model includes:
s401, obtaining a plurality of sample data groups and labeling non-verticality corresponding to each sample data group.
S402, inputting each sample data group into an initial prediction model for training to obtain the prediction non-verticality of each sample data group.
And S403, judging whether the relative error between the prediction non-verticality and the labeling non-verticality is larger than a preset threshold range.
And S404, if so, taking the initial prediction model as a target prediction model.
And S405, if not, updating the parameters of the initial prediction model to obtain a new initial prediction model, and re-executing the steps S402-S403.
The above steps S401 to S405 are specifically described in the foregoing embodiments, and are not described herein again.
Optionally, the target prediction model finished by training may be used for predicting, and the obtained result is the data in table 1, for example, the relative error between the predicted non-perpendicularity and the labeled non-perpendicularity of the 44 sets of sample data sets shown in table 1, it can be seen from the table that the predicted non-perpendicularity of each sample data set obtained by predicting using the target prediction model and the actual non-perpendicularity obtained by measuring each sample data set, that is, the labeled non-perpendicularity have a certain similarity, and part of the predicted non-perpendicularity is consistent with the labeled non-perpendicularity data, and the relative error of more than half of the data is less than 1%, so that the target prediction model finished by training has a better prediction ability, and can be used for predicting the cutting parameters of the plasma cutting equipment.
TABLE 1
Fig. 6 is a schematic diagram of an apparatus of a cutting parameter prediction processing method according to an embodiment of the present application, as shown in fig. 6, the apparatus includes:
an obtaining module 401, configured to obtain parameters of a process to be cut, where the parameters of the process to be cut include: current intensity, cutting speed and gas pressure;
an input module 402, configured to input the to-be-cut process parameter into a target prediction model obtained through pre-training, to obtain a predicted non-verticality corresponding to the to-be-cut process parameter, so as to verify whether the to-be-cut process parameter meets a target non-verticality requirement by using the predicted non-verticality.
Optionally, the input module 402 is specifically configured to:
obtaining a plurality of sample data groups and labeling non-verticality corresponding to each sample data group, wherein each sample data group respectively comprises: current intensity, cutting speed and gas pressure;
inputting each sample data set into an initial prediction model to obtain the prediction non-verticality corresponding to each sample data set;
and performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain the target prediction model.
Optionally, the obtaining module 401 is specifically configured to:
determining different cutting process parameters according to different current intensities, cutting speeds and gas pressures, and forming the different cutting process parameters into the plurality of sample data sets;
and obtaining the marked non-verticality corresponding to each sample data set according to the actual non-verticality obtained by cutting according to each sample data set.
Optionally, the initial prediction model includes a hidden layer and an output layer, where the hidden layer includes multiple hidden nodes;
optionally, the input module 402 is specifically configured to:
respectively inputting the current intensity, the cutting speed and the gas pressure in the sample data set into each input node of the hidden layer;
determining an output value of the hidden layer according to the input function and the output function of the hidden layer;
and inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to the sample data set.
Optionally, the input module 402 is specifically configured to:
determining a function value of an input function of the hidden layer according to the sample data set, the number of neurons in an input layer, the weight between the input neuron and the hidden neuron, the weight between the hidden neuron and the neuron in an output layer, the number of hidden neurons and the bias of hidden nodes;
determining a function value of an output function of the hidden layer according to the function value of the input function of the hidden layer and a linear function;
and determining the output value of the hidden layer according to the function value of the output function of the hidden layer.
Optionally, the input module 402 is specifically configured to:
determining a function value of an input function of the output layer according to the weight between the hidden layer and the output layer, the output value of the hidden layer and the bias of the node of the output layer;
determining an output function of the output layer according to the function value of the input function of the output layer and the activation function;
and obtaining the predicted non-verticality corresponding to the sample data set according to the output function.
Optionally, the input module 402 is specifically configured to:
determining a relative error between the predicted non-perpendicularity and the labeled non-perpendicularity according to the predicted non-perpendicularity and the labeled non-perpendicularity corresponding to the sample data set;
determining whether the initial prediction model is an optimal model or not according to the relative error and a preset threshold range, and if so, taking the initial prediction model as a target prediction model;
if not, updating parameters in the initial prediction model to obtain a new initial prediction model, inputting each sample data set into the new initial prediction model again, re-determining the relative error, re-judging whether the new initial prediction model is the optimal model, performing iteration until an iteration stop condition is met, and taking the new initial prediction model meeting the iteration stop condition as a target prediction model.
Fig. 7 is a block diagram of an electronic device 500 according to an embodiment of the present disclosure. As shown in fig. 7, the electronic device may include: a processor 501 and a memory 502.
Optionally, a bus 503 may be further included, wherein the memory 502 is configured to store machine-readable instructions executable by the processor 501, when the electronic device 500 runs, the processor 501 and the memory 502 store communication via the bus 503, and the machine-readable instructions are executed by the processor 501 to perform the method steps in the above method embodiments.
The embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the computer program executes the method steps in the foregoing cutting parameter prediction processing method embodiment.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and the apparatus described above may refer to corresponding processes in the method embodiments, and are not described in detail in this application. In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. The above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is merely a logical division, and there may be other divisions in actual implementation, and for example, a plurality of modules or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or modules through some communication interfaces, and may be in an electrical, mechanical or other form.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including 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 application. 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 various media capable of storing program codes.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application.
Claims (10)
1. A method for processing prediction of cutting parameters, the method comprising:
acquiring technological parameters to be cut, wherein the technological parameters to be cut comprise: current intensity, cutting speed, and gas pressure;
inputting the technological parameters to be cut into a target prediction model obtained by pre-training to obtain the predicted non-verticality corresponding to the technological parameters to be cut, and verifying whether the technological parameters to be cut meet the requirement of the target non-verticality or not by using the predicted non-verticality.
2. The cutting parameter prediction processing method according to claim 1, wherein before the step of inputting the to-be-cut process parameter into a pre-trained target prediction model to obtain the predicted non-perpendicularity corresponding to the to-be-cut process parameter, the method further comprises:
obtaining a plurality of sample data groups and labeling non-verticality corresponding to each sample data group, wherein each sample data group comprises: current intensity, cutting speed, and gas pressure;
inputting each sample data set into an initial prediction model to obtain the prediction non-verticality corresponding to each sample data set;
and performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain the target prediction model.
3. The method of claim 2, wherein obtaining a plurality of sample data sets and a labeled non-verticality corresponding to each sample data set comprises:
determining different cutting process parameters according to different current intensities, cutting speeds and gas pressures, and forming the different cutting process parameters into the plurality of sample data sets;
and obtaining the marked non-verticality corresponding to each sample data group according to the actual non-verticality obtained by cutting according to each sample data group.
4. The cutting parameter prediction processing method according to claim 2, wherein the initial prediction model includes a hidden layer and an output layer, and the hidden layer includes a plurality of hidden nodes;
inputting each sample data set into an initial prediction model to obtain the predicted non-verticality corresponding to each sample data set, including:
respectively inputting the current intensity, the cutting speed and the gas pressure in the sample data set into each input node of the hidden layer;
determining an output value of the hidden layer according to the input function and the output function of the hidden layer;
and inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to the sample data set.
5. The method according to claim 4, wherein the determining the output value of the hidden layer according to the input function and the output function of the hidden layer comprises:
determining a function value of an input function of the hidden layer according to the sample data set, the number of neurons in an input layer, the weight between the input neuron and the hidden neuron, the weight between the hidden neuron and the neuron in an output layer, the number of hidden neurons and the bias of hidden nodes;
determining a function value of an output function of the hidden layer according to the function value of the input function of the hidden layer and a linear function;
and determining the output value of the hidden layer according to the function value of the output function of the hidden layer.
6. The method according to claim 4, wherein the inputting the output value of the hidden layer into the output layer to obtain the predicted non-verticality corresponding to the sample data comprises:
determining a function value of an input function of the output layer according to the weight between the hidden layer and the output layer, the output value of the hidden layer and the bias of the node of the output layer;
determining an output function of the output layer according to the function value of the input function of the output layer and the activation function;
and obtaining the predicted non-verticality corresponding to the sample data set according to the output function.
7. The cutting parameter prediction processing method according to claim 2, wherein the performing iterative optimization on the initial prediction model according to the prediction non-verticality and the labeling non-verticality corresponding to each sample data set to obtain the target prediction model comprises:
determining a relative error between the predicted non-perpendicularity and the labeled non-perpendicularity according to the predicted non-perpendicularity and the labeled non-perpendicularity corresponding to the sample data set;
determining whether the initial prediction model is an optimal model or not according to the relative error and a preset threshold range, and if so, taking the initial prediction model as a target prediction model;
if not, updating parameters in the initial prediction model to obtain a new initial prediction model, inputting each sample data set into the new initial prediction model again, re-determining the relative error, re-judging whether the new initial prediction model is the optimal model, performing iteration until an iteration stop condition is met, and taking the new initial prediction model meeting the iteration stop condition as a target prediction model.
8. A cutting parameter prediction processing apparatus, comprising:
the acquisition module is used for acquiring the technological parameters to be cut, and the technological parameters to be cut comprise: current intensity, cutting speed, and gas pressure;
and the input module is used for inputting the process parameters to be cut into a target prediction model obtained by pre-training to obtain the predicted non-verticality corresponding to the process parameters to be cut so as to verify whether the process parameters to be cut meet the requirement of the target non-verticality by using the predicted non-verticality.
9. An electronic device, comprising a memory and a processor, wherein the memory stores a computer program executable by the processor, and the processor implements the steps of the cutting parameter prediction processing method according to any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the cutting parameter prediction processing method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211563596.7A CN115809523A (en) | 2022-12-07 | 2022-12-07 | Cutting parameter prediction processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211563596.7A CN115809523A (en) | 2022-12-07 | 2022-12-07 | Cutting parameter prediction processing method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115809523A true CN115809523A (en) | 2023-03-17 |
Family
ID=85485219
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211563596.7A Pending CN115809523A (en) | 2022-12-07 | 2022-12-07 | Cutting parameter prediction processing method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115809523A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116213962A (en) * | 2023-05-10 | 2023-06-06 | 杭州乾瑭云科技有限公司 | Metal plate cutting control method and system based on state prediction |
-
2022
- 2022-12-07 CN CN202211563596.7A patent/CN115809523A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116213962A (en) * | 2023-05-10 | 2023-06-06 | 杭州乾瑭云科技有限公司 | Metal plate cutting control method and system based on state prediction |
| CN116213962B (en) * | 2023-05-10 | 2023-08-11 | 杭州乾瑭云科技有限公司 | Metal plate cutting control method and system based on state prediction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11267059B2 (en) | Control device of wire electric discharge machine and machine learning device | |
| CN111814385B (en) | Method, device and computer equipment for predicting quality of machined part | |
| CN110561191B (en) | Numerical control machine tool cutter abrasion data processing method based on PCA and self-encoder | |
| Guh | Robustness of the neural network based control chart pattern recognition system to non‐normality | |
| US11914339B2 (en) | Datum selection methods and systems for minimizing hole position errors in group hole machining of large components | |
| CN115809523A (en) | Cutting parameter prediction processing method and device, electronic equipment and storage medium | |
| CN110928237B (en) | Vibration signal-based numerical control machining center flutter online identification method | |
| JP2019135638A (en) | Data driving method for automatically detecting abnormal workpiece during production process | |
| CN108388745B (en) | Least square support vector machine soft measuring method based on distributed parallel local optimization parameters | |
| CN114282376A (en) | Anti-collision and over-cutting-based NC code detection method and device and intelligent terminal | |
| CN110532629A (en) | A kind of method for predicting residual useful life of multi-modal degenerative process | |
| CN112884081B (en) | On-line intelligent predicting method for quality of workpieces produced in small batches of multiple varieties | |
| CN108073464A (en) | A kind of time series data abnormal point detecting method and device based on speed and acceleration | |
| CN112016598A (en) | SDAE-based partial discharge pattern recognition method, computer device and storage medium | |
| CN111597499B (en) | Industrial equipment fault maintenance scheme generation method and system | |
| CN112363465B (en) | Expert rule set training method, trainer and industrial equipment early warning system | |
| CN114139636B (en) | Abnormal operation processing method and device | |
| CN113836826A (en) | Key parameter determination method and device, electronic device and storage medium | |
| CN113159419A (en) | Group feature portrait analysis method, device and equipment and readable storage medium | |
| CN112758106A (en) | Vehicle running track prediction method and device | |
| CN112132340A (en) | Improved particle swarm-based gas concentration prediction method and device | |
| CN111142026A (en) | Data processing method and device and electronic equipment | |
| CN115438452B (en) | Reliable transmission detection method for time sequence network signals | |
| Haq et al. | TransNAS-TSAD: Harnessing Transformers for Multi-Objective Neural Architecture Search in Time Series Anomaly Detection | |
| CN114386196B (en) | Method for evaluating mechanical property prediction accuracy of plate strip |
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 |