A method and system for intelligent matching of machining parameters in slow wire EDM.

Abstract

This invention relates to the field of machining technology, and particularly to an intelligent matching method and system for slow wire EDM machining parameters. The method involves loading electrode wire parameters and workpiece parameters to configure machining parameters and obtain first machining parameters. When the monitored environmental parameters are inconsistent with the standard environmental parameters, the method uses the electrode wire parameters, workpiece parameters, first machining parameters, and monitored environmental parameters as constraints to search the slow wire EDM machining sample set, performs lumped value evaluation, and obtains the fitted dimensions of the workpiece after cutting. When the dimensional deviation between the fitted dimensions and the standard dimensions of the workpiece after cutting is less than or equal to the dimensional tolerance, metal cutting control is performed based on the first machining parameters; otherwise, the first machining parameters are updated. This invention improves the accuracy and stability of metal cutting using slow wire EDM, reduces trial cutting costs and scrap rates, increases machining efficiency, and reduces production burden.

Description

Technical Field

[0001] This invention relates to the field of metal cutting technology, and in particular to an intelligent matching method and system for slow wire EDM machining parameters. Background Technology

[0002] Wire EDM machines are high-precision conductive metal cutting devices, commonly used for cutting precision parts made of materials such as mold steel and aluminum alloys. Traditional wire EDM machine parameter configuration relies on operator experience and does not consider environmental fluctuations. The control method is simplistic and lacks a dynamic adjustment mechanism, requiring exhaustive trial and error for parameter adjustments, resulting in low efficiency. Existing technologies suffer from high trial-cutting costs, easy deviations in metal cutting dimensions, and high scrap rates. Summary of the Invention

[0003] This invention addresses the technical problems of high trial cutting costs, easy deviations in metal cutting dimensions, and high scrap rates in existing technologies by providing an intelligent matching method and system for slow wire EDM machining parameters.

[0004] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0005] In a first aspect, the present invention provides an intelligent matching method for slow wire EDM machining parameters, comprising: loading electrode wire parameters and workpiece parameters to perform machining parameter configuration and obtain a first machining parameter, wherein the first machining parameter has a label identifying a standard environmental parameter; when the monitored environmental parameter is inconsistent with the standard environmental parameter, using the electrode wire parameter, the workpiece parameter, the first machining parameter, and the monitored environmental parameter as constraints, retrieving a slow wire EDM machining sample set, performing lumped value evaluation, and obtaining a fitted size of the workpiece after cutting; when the dimensional deviation between the fitted size of the workpiece after cutting and the standard size of the workpiece after cutting is less than or equal to the dimensional tolerance, performing metal cutting control based on the first machining parameter; otherwise, updating the first machining parameter.

[0006] Optionally, loading electrode wire parameters and workpiece parameters to execute machining parameter configuration and obtain the first machining parameter includes: configuring the standard environmental parameters through a user terminal; collecting multiple sets of data under the constraints of the standard environmental parameters and machine tool model, wherein any set of the multiple sets of data includes electrode wire parameter recording data, workpiece parameter recording data, and a label identifying the selected machining parameter; using the label identifying the selected machining parameter as supervision, and the electrode wire parameter recording data and workpiece parameter recording data as input, retrieving the multiple sets of data and training a machining parameter configuration model through machine learning; processing the electrode wire parameters and workpiece parameters through the machining parameter configuration model, outputting the first machining parameter; and constructing a label identifying the standard environmental parameters based on the standard environmental parameters.

[0007] The process involves using the labels indicating the selected processing parameters as supervision, and the electrode wire parameter recording data and the workpiece parameter recording data as input. Multiple sets of data are retrieved and trained using machine learning to create a processing parameter configuration model. This includes: constructing a first branch, a second branch, and a backbone network based on a fully connected neural network; connecting the first branch and the second branch in parallel to the backbone network to obtain the processing parameter configuration model architecture; and training the processing parameter configuration model architecture using the electrode wire parameter recording data as input to the first branch, the workpiece parameter recording data as input to the second branch, and the labels indicating the selected processing parameters as supervision to obtain the processing parameter configuration model.

[0008] Optionally, when the monitored environmental parameters are inconsistent with the standard environmental parameters, the process includes: using the electrode wire parameters and the workpiece parameters as constraints, performing a machining dimension correlation analysis on the environmental parameter attributes to obtain a correlation degree set; traversing the correlation degree set and comparing it with the sum of the correlation degree sets to obtain the environmental parameter attribute weight distribution; calculating the normalized deviation of the environmental parameter attributes between the monitored environmental parameters and the standard environmental parameters; based on the environmental parameter attribute weight distribution and combined with the normalized deviation of the environmental parameter attributes, performing a comprehensive deviation analysis to obtain an environmental inconsistency coefficient; when the environmental inconsistency coefficient is greater than or equal to the inconsistency coefficient threshold, the monitored environmental parameters are considered inconsistent with the standard environmental parameters; otherwise, the monitored environmental parameters are considered consistent with the standard environmental parameters.

[0009] Optionally, it also includes: when the monitored environmental parameters are consistent with the standard environmental parameters, performing metal cutting control based on the first processing parameters.

[0010] Optionally, using the electrode wire parameters, workpiece parameters, first processing parameters, and monitoring environment parameters as constraints, a slow wire EDM processing sample set is retrieved, and mode analysis is performed to obtain the fitted dimensions of the workpiece after cutting. This includes: obtaining a slow wire EDM processing sample to be analyzed, wherein the slow wire EDM processing sample to be analyzed includes electrode wire recording parameters, workpiece recording parameters, environment recording parameters, processing recording parameters, and measured dimensions of the workpiece after cutting; when the electrode wire recording parameters are the same as the electrode wire parameters, and the workpiece recording parameters are the same as the workpiece parameters, and the environment recording parameters are... If the recorded parameters are consistent with the monitoring environment parameters, and the first processing parameter is the same as the processing record parameter, the slow wire EDM processing sample to be analyzed is considered to satisfy the constraints. The measured dimensions of the cut workpiece are added to the slow wire EDM processing sample set. In this case, any type of parameter representing the same two sets of data is the same, and the deviation of any numerical parameter is less than or equal to the corresponding attribute deviation threshold. When the amount of data in the slow wire EDM processing sample set is greater than or equal to the fitting data amount threshold, mode analysis is performed on the slow wire EDM processing sample set to obtain the fitted dimensions of the cut workpiece.

[0011] Optionally, updating the first processing parameter includes: obtaining the updated processing parameter; when the data volume of the updated processing parameter is less than the optimized data volume threshold, using the updated processing parameter as a forbidden processing parameter, performing random configuration of processing parameters, and updating the first processing parameter, wherein the updated processing parameter is different from the forbidden processing parameter; when the data volume of the updated processing parameter is greater than or equal to the optimized data volume threshold, using the size deviation as a fitness index, optimizing the processing parameters based on the updated processing parameter, and updating the first processing parameter.

[0012] Secondly, the present invention provides an intelligent matching system for slow wire EDM machining parameters, comprising:

[0013] The first processing parameter acquisition module is used to load electrode wire parameters and workpiece parameters to execute processing parameter configuration and obtain the first processing parameter, wherein the first processing parameter has a label identifying standard environmental parameters;

[0014] The workpiece fitting size acquisition module after cutting is used to retrieve the slow wire EDM processing sample set, perform lumped value evaluation, and obtain the workpiece fitting size after cutting when the monitored environmental parameters are inconsistent with the standard environmental parameters, using the electrode wire parameters, the workpiece parameters, the first processing parameters, and the monitored environmental parameters as constraints;

[0015] The first processing parameter evaluation module is used to perform metal cutting control based on the first processing parameters when the dimensional deviation between the fitted size of the cut workpiece and the standard size of the cut workpiece is less than or equal to the dimensional tolerance; otherwise, it updates the first processing parameters.

[0016] By implementing this invention, it is possible to load electrode wire parameters and workpiece parameters to execute processing parameter configuration and obtain first processing parameters. The first processing parameters have labels that identify standard environmental parameters, which can ensure the correlation between processing parameters and environmental conditions and processing object characteristics, improve the accuracy of initial configuration, avoid the potential deviation caused by parameters being out of sync with the environment, and provide a clear reference for subsequent environmental comparison and parameter adjustment, laying a coherent foundation for processing parameter control.

[0017] By implementing this invention, when the monitored environmental parameters are inconsistent with the standard environmental parameters, the electrode wire parameters, the workpiece parameters, the first processing parameters, and the monitored environmental parameters are used as constraints to retrieve a slow wire EDM processing sample set, perform lumped value evaluation, and obtain the fitted size of the workpiece after cutting. This invention can retrieve matching samples and evaluate the fitted size with multi-dimensional parameter constraints when the monitored environment is inconsistent with the standard environment, thus overcoming the shortcomings of traditional processes that ignore environmental factors. It can predict the processing size without trial cutting, reducing material and time costs. Moreover, the high matching degree of the samples ensures the reliability of the fitted size, provides accurate data support for parameter adjustment, and improves the adaptability to complex processing environments.

[0018] By implementing this invention, when the dimensional deviation between the fitted size of the cut workpiece and the standard size of the cut workpiece is less than or equal to the dimensional tolerance, metal cutting control is executed based on the first processing parameters; otherwise, the first processing parameters are updated. This avoids defective parameters leading to scrap, breaks through the limitations of traditional single-modification of offset, flexibly selects the update method according to the amount of updated parameter data, takes into account both processing accuracy and efficiency, realizes dynamic parameter optimization, ensures that the final workpiece meets the accuracy requirements, and balances processing quality and production progress.

[0019] In summary, by implementing this invention, the dimensional accuracy and stability of slow wire EDM can be significantly improved, the metal cutting accuracy and stability can be improved, the scrap rate can be reduced, the trial cutting cost can be reduced, and the overall processing efficiency can be improved. Attached Figure Description

[0020] Figure 1 A flowchart illustrating an intelligent matching method for slow wire EDM machining parameters provided by the present invention;

[0021] Figure 2 This is a schematic diagram of the structure of an intelligent matching system for slow wire EDM machining parameters provided by the present invention.

[0022] In the attached diagram, the components represented by each number are as follows:

[0023] Module 11 for acquiring first processing parameters, module 12 for acquiring fitted dimensions of cut workpiece, and module 13 for evaluating first processing parameters. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0026] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0027] Example 1, as Figure 1 As shown, this embodiment of the invention provides a method and system for intelligent matching of slow wire EDM machining parameters, including:

[0028] S100 loads electrode wire parameters and workpiece parameters to execute processing parameter configuration and obtains first processing parameters, wherein the first processing parameters have a label identifying standard environmental parameters;

[0029] S200 When the monitored environmental parameters are inconsistent with the standard environmental parameters, the electrode wire parameters, the workpiece parameters, the first processing parameters and the monitored environmental parameters are used as constraints to retrieve the slow wire EDM processing sample set, perform lumped value evaluation, and obtain the fitted size of the workpiece after cutting.

[0030] S300 When the dimensional deviation between the fitted size of the cut workpiece and the standard size of the cut workpiece is less than or equal to the dimensional tolerance, metal cutting control is executed based on the first processing parameters; otherwise, the first processing parameters are updated.

[0031] In step S100 of this embodiment, loading electrode wire parameters and workpiece parameters to perform machining parameter configuration and obtain first machining parameters include:

[0032] Configure the standard environment parameters through the user terminal;

[0033] Constrained by the standard environmental parameters and machine tool model, multiple sets of data are collected, wherein each set of multiple sets of data includes electrode wire parameter recording data, workpiece parameter recording data, and a label identifying the selected processing parameters;

[0034] Using the labels of the selected processing parameters as supervision, and the electrode wire parameter recording data and the workpiece parameter recording data as input, the processing parameter configuration model is trained by retrieving the multiple sets of data and using machine learning.

[0035] The electrode wire parameters and the workpiece parameters are processed through the processing parameter configuration model, and the first processing parameters are output.

[0036] And based on the standard environmental parameters, labels that identify the standard environmental parameters are constructed.

[0037] In this embodiment, the purpose of step S100 is to construct an intelligent machining parameter configuration mechanism based on standard environmental parameters and actual machining data to generate first machining parameters that are adapted to a specific standard environment, machine tool model, and specific electrode wire and workpiece, and to clarify their association with the standard environment. By replacing traditional empirical parameter configuration with a machine learning model, the accuracy and adaptability of the initial machining parameters are improved, laying a standardized foundation for subsequent responses to environmental fluctuations and ensuring machining accuracy.

[0038] Specifically, the first step is to set the standard environmental parameters required for processing through the user terminal, which will serve as the benchmark environmental conditions for subsequent parameter configurations. For example, the standard environmental parameters could be an indoor temperature of 21±2℃, humidity of 40%~80%, and no direct sunlight.

[0039] Furthermore, multiple sets of historical machining data need to be collected under the constraints of set standard environmental parameters and specific machine tool models. Each set of data needs to include electrode wire parameter records, workpiece parameter records, and tags identifying the selected machining parameters. The electrode wire parameters include the electrode wire material and diameter, such as brass material and 0.18mm diameter. The workpiece parameters include the workpiece material, thickness, and surface roughness, such as mold steel material, 50mm thickness, and Ra0.8μm surface roughness. The tags identifying the selected machining parameters are used to uniquely correspond to and clearly identify the machining parameters used in a specific machining scenario. These machining parameters include discharge voltage, current, pulse interval, and electrode wire feed speed, etc.

[0040] In step S100 of this application embodiment, using the label identifying the selected processing parameters as supervision, and using the electrode wire parameter recording data and the workpiece parameter recording data as input, the processing parameter configuration model is trained through machine learning by retrieving the multiple sets of data, including:

[0041] Based on a fully connected neural network, a first branch, a second branch, and a backbone network are constructed respectively. The first branch and the second branch are connected in parallel to the backbone network to obtain the processing parameter configuration model architecture.

[0042] Using the electrode wire parameter recording data as the first branch input, the workpiece parameter recording data as the second branch input, and the label identifying the selected processing parameter as supervision, the processing parameter configuration model architecture is trained to obtain the processing parameter configuration model.

[0043] In this embodiment of the application, the training of the processing parameter configuration model aims to build a machine learning model that can accurately map the relationship between electrode wire parameters, workpiece parameters and processing parameters. The core purpose is to extract high-dimensional abstract features by processing different types of parameters through branching, and then to improve the learning ability and prediction accuracy of the processing parameter configuration model for complex parameter relationships through backbone network fusion analysis, thereby providing reliable model support for the subsequent generation of the first suitable processing parameters.

[0044] First, based on a fully connected neural network, a first branch, a second branch, and a backbone network need to be constructed. The first branch and the second branch are then connected in parallel to the backbone network to obtain the processing parameter configuration model architecture. The first branch processes electrode wire parameter recording data, the second branch processes workpiece parameter recording data, and the backbone network fuses the features output from the two branches to complete the processing parameter configuration task.

[0045] In the construction of the three branches, the first branch contains three fully connected layers: the first layer has 64 neurons, the second layer has 32 neurons, and the third layer has 16 neurons, each using the ReLU activation function. The second branch also contains three fully connected layers: the first layer has 64 neurons, the second layer has 32 neurons, and the third layer has 16 neurons, each using the ReLU activation function. The backbone network contains two fully connected layers: the first layer has 32 neurons using the ReLU activation function; the second layer has the same number of neurons as the dimension of the processing parameters and uses a linear activation function.

[0046] In the parameter configuration of the processing parameter configuration model, the learning rate is set to 0.001; the batch size is set to 32; the regularization coefficient is set to 0.0001; the optimizer is Adam; and the loss function is the mean squared error loss function.

[0047] The training data for the machining parameter configuration model consists of multiple sets of historical machining data collected under the constraints of standard environmental parameters and machine tool model. Each set of historical machining data includes electrode wire parameter records, workpiece parameter records, and labels identifying the selected machining parameters. Prepare at least 10,000 sets of data that meet the above requirements.

[0048] In training the processing parameter configuration model, the initial number of training rounds is set to 200. If the convergence criterion is not met, the number of rounds can be increased appropriately. The processing parameter configuration model is considered to have converged when the change in validation loss is less than 0.0001 for 10 consecutive rounds.

[0049] Furthermore, the current electrode wire parameters and workpiece parameters need to be input into the trained model, and the model will output the first suitable processing parameters.

[0050] The electrode wire parameters in the current machining scenario are input into the first branch of the machining parameter configuration model, and the workpiece parameters are input into the second branch. The first and second branches respectively extract and process features from the input parameters, obtaining high-dimensional abstract features which are then transmitted to the backbone network. The backbone network fuses and analyzes the features from the two branches, calculates based on the mapping relationships learned by the model, and finally outputs the first machining parameters that match the current electrode wire and workpiece parameters, including discharge voltage, current, pulse interval, wire feed speed, and other machining parameters.

[0051] Based on the initially configured standard environmental parameters, a label is constructed to identify the standard environmental parameters for the first processing parameter output, such as indoor temperature 21±2℃, humidity 40%~80%, and no direct sunlight, to clearly define the baseline environmental conditions corresponding to the parameter. This label is bound to the first processing parameter, explicitly recording the baseline environmental conditions corresponding to the first processing parameter, thus establishing a correspondence between the first processing parameter and the standard environmental parameters.

[0052] In step S200 of this application embodiment, when the monitored environmental parameters are inconsistent with the standard environmental parameters, the following steps are taken:

[0053] Using the electrode wire parameters and the workpiece parameters as constraints, a machining dimension correlation analysis is performed on the environmental parameter attributes to obtain a correlation degree set;

[0054] Traverse the set of correlation degrees and compare it with the sum of the correlation degree sets to obtain the weight distribution of environmental parameter attributes;

[0055] Calculate the normalized deviation of the environmental parameter attributes between the monitored environmental parameters and the standard environmental parameters;

[0056] Based on the weight distribution of the environmental parameter attributes, and combined with the normalized deviation of the environmental parameter attributes, a comprehensive deviation analysis is performed to obtain the environmental inconsistency coefficient.

[0057] When the environmental inconsistency coefficient is greater than or equal to the inconsistency coefficient threshold, the monitored environmental parameters are considered to be inconsistent with the standard environmental parameters.

[0058] Otherwise, the monitoring environment parameters are considered to be consistent with the standard environment parameters.

[0059] In step S200 of this application embodiment, the method further includes: when the monitored environmental parameters are consistent with the standard environmental parameters, performing metal cutting control based on the first processing parameters.

[0060] In this embodiment of the application, the purpose of step S200 is to obtain a scientific environmental consistency judgment result by quantitatively analyzing the influence weight of environmental parameters on the processing dimensions and the actual deviation, so as to avoid ignoring the differences in the degree of influence of different environmental factors on processing accuracy due to simple comparison of environmental parameter values, and to provide a reliable basis for whether to start the processing parameter adjustment process in the future.

[0061] First, the machining dimension correlation analysis needs to be performed on the environmental parameter attributes, using the electrode wire parameters and the workpiece parameters as constraints, to obtain the correlation degree set.

[0062] Specifically, a correlation analysis is performed using the normalized parameters of dimensional deviation as the baseline sequence and the normalized parameters of environmental parameters as the comparison sequence. The dimensional deviation refers to the difference between the actual measured size of the workpiece after wire EDM machining and the standard size of the cut workpiece required by the drawing; it is a core indicator for measuring whether the machining accuracy meets the standards. Specifically, dimensional deviation = actual measured size of workpiece - standard size of cut workpiece. A positive result indicates that the actual size of the workpiece is too large; a negative result indicates that the actual size of the workpiece is too small; a result of 0 indicates that the actual size is completely consistent with the standard size.

[0063] In constructing the reference sequence, it is necessary to normalize the dimensional deviation data of the processed workpiece to obtain the normalized parameters of the dimensional deviation, which are used as the reference sequence, denoted as X0. This sequence reflects the degree of deviation of the processed dimensions.

[0064] Constructing the alignment sequence requires normalizing the measured data of environmental parameters such as temperature and humidity to obtain normalized parameters. Each environmental parameter corresponds to an alignment sequence, denoted as X. i , i = 1, 2, ..., n, where n is the number of environmental parameters, and these sequences reflect the actual state of different environmental parameters.

[0065] Furthermore, the correlation coefficient needs to be calculated using the formula for grey relational analysis. The formula is: ξ i (k)=(Δmin+ρΔmax) / (Δ i (k)+ρΔmax), where ρ is the resolution coefficient, ranging from 0 to 1, usually taken as 0.5, ξ i (k) is the correlation coefficient between the i-th alignment sequence and the reference sequence at the k-th sample.

[0066] For each alignment sequence, calculate its correlation coefficient with the baseline sequence across all samples to obtain the correlation coefficient sequence.

[0067] Next, the correlation coefficient sequence of each aligned sequence is averaged to obtain the correlation degree r between the aligned sequence and the reference sequence. i That is, r i =(1 / m)Σξ i (k), where m is the number of samples.

[0068] A correlation set is formed by the correlation degrees corresponding to all environmental parameters. Each value in this set reflects the degree of correlation between the corresponding environmental parameter and the dimensional deviation. For example, {temperature correlation degree: 0.5, humidity correlation degree: 0.3, light correlation degree: 0.2}.

[0069] Furthermore, it is necessary to iterate through the aforementioned set of correlation degrees and compare it with the sum of the correlation degrees in the set to obtain the weight distribution of the environmental parameter attributes. That is, the correlation degree values ​​of all environmental parameter attributes in the correlation degree set are added together to obtain the sum. Taking the above set as an example, the sum is 0.5 + 0.3 + 0.2 = 1.0. Iterating through each correlation degree value in the correlation degree set, each individual correlation degree value is divided by the sum of the correlation degrees to obtain the weight of that environmental parameter attribute. For example, temperature weight = 0.5 ÷ 1.0 = 0.5, humidity weight = 0.3 ÷ 1.0 = 0.3, and light weight = 0.2 ÷ 1.0 = 0.2, ultimately forming the environmental parameter attribute weight distribution {temperature: 0.5, humidity: 0.3, light: 0.2}.

[0070] Furthermore, it is necessary to calculate the normalized deviation of the monitored environmental parameters from the standard environmental parameters. That is, for each environmental parameter attribute, calculate the difference between the measured value of the monitored environmental parameter and the set value of the standard environmental parameter. For example, if the standard temperature is 21℃ and the monitored temperature is 23℃, then the temperature deviation = 23 - 21 = 2℃; if the standard humidity is 60% and the monitored humidity is 66%, then the humidity deviation = 66 - 60 = 6%.

[0071] Next, the allowable standard fluctuation range for each environmental parameter needs to be defined. This range is determined based on processing accuracy requirements and historical data. For example, the standard fluctuation range for temperature is ±2℃, meaning the allowable deviation range is 4℃; the standard fluctuation range for humidity is ±10%, meaning the allowable deviation range is 20%. Then, the deviation value of each environmental parameter is divided by its corresponding standard fluctuation range to obtain the normalized deviation. For example, the normalized deviation for temperature = 2 ÷ 4 = 0.5, and the normalized deviation for humidity = 6 ÷ 20 = 0.3.

[0072] Furthermore, based on the weight distribution of the environmental parameter attributes and combined with the normalized deviation of the environmental parameter attributes, a comprehensive deviation analysis is needed to obtain the environmental inconsistency coefficient.

[0073] The product of the normalized deviation and the weight is used as the coordinate distance for each dimension, and the environmental inconsistency coefficient is obtained by calculating the Euclidean distance. Each environmental parameter attribute is considered as an independent dimension, and the number of dimensions is the same as the number of environmental parameter attributes. For example, temperature, humidity, and vibration are three-dimensional.

[0074] For each dimension, the normalized deviation of that environmental parameter attribute multiplied by its corresponding weight is calculated as the coordinate distance for that dimension. Taking the data above as an example, the coordinate distance for temperature is 0.5 × 0.5 = 0.25, and for humidity it is 0.3 × 0.3 = 0.09. If the normalized deviation for illumination is 0.4, then the coordinate distance for illumination is 0.4 × 0.2 = 0.08. Therefore, the coordinate distances for temperature, humidity, and illumination are 0.25, 0.09, and 0.08, respectively. The Euclidean distance is calculated by squaring the coordinate distances for all dimensions, summing the squares, and then taking the square root. The sum of the squares of the three coordinates is 0.0625 + 0.0081 + 0.0064 = 0.077, so the Euclidean distance is √0.077 ≈ 0.277, which is the environmental inconsistency coefficient.

[0075] Next, it is necessary to set the non-consistency coefficient threshold in advance. This non-consistency coefficient threshold is preset based on historical processing data and accuracy requirements, such as 0.3. This non-consistency coefficient threshold represents the critical value at which environmental differences have a substantial impact on processing accuracy.

[0076] If the environmental inconsistency coefficient is greater than or equal to the inconsistency coefficient threshold, it is determined that the monitored environmental parameters are inconsistent with the standard environmental parameters, and the subsequent parameter adjustment process needs to be initiated.

[0077] If the environmental inconsistency coefficient is less than the inconsistency coefficient threshold, then the two are determined to be consistent, and processing is directly performed based on the first processing parameters obtained in step S100 without the need to adjust the processing parameters.

[0078] In step S200 of this application embodiment, using the electrode wire parameters, the workpiece parameters, the first processing parameters, and the monitoring environment parameters as constraints, a slow wire EDM processing sample set is retrieved, and mode analysis is performed to obtain the fitted dimensions of the workpiece after cutting, including:

[0079] Obtain a slow wire EDM machining sample to be analyzed, wherein the slow wire EDM machining sample to be analyzed includes electrode wire recording parameters, workpiece recording parameters, environmental recording parameters, machining recording parameters, and workpiece measurement recording dimensions after cutting;

[0080] When the electrode wire recording parameters are the same as the electrode wire parameters, the workpiece recording parameters are the same as the workpiece parameters, the environmental recording parameters are consistent with the monitoring environment parameters, and the first processing parameters are the same as the processing recording parameters, it is considered that the slow wire EDM processing sample to be analyzed satisfies the constraints. The measured and recorded dimensions of the workpiece after cutting are added to the slow wire EDM processing sample set. In this case, any type of parameter that represents the same two sets of data is the same, and any numerical parameter deviation is less than or equal to the corresponding attribute deviation threshold.

[0081] When the amount of data in the slow wire EDM machining sample set is greater than or equal to the threshold of the amount of fitted data, a mode analysis is performed on the slow wire EDM machining sample set to obtain the fitted dimensions of the workpiece after cutting.

[0082] In step S200 of this embodiment, the purpose of the above step is to obtain the fitted size of the cut workpiece that matches the current processing conditions by accurately searching and analyzing the sample set of slow wire EDM machining, thus providing a reliable basis for subsequent judgment on whether the processing parameters need to be adjusted. By strictly constraining the sample selection conditions and using mode analysis, it is ensured that the fitted size can truly reflect the workpiece size trend under the current processing conditions, thereby improving the accuracy of parameter adjustment decisions.

[0083] Specifically, it is necessary to obtain samples of the slow wire EDM machining to be analyzed. This involves collecting historical sample data of slow wire EDM machining. Each sample to be analyzed must contain complete five-dimensional information: electrode wire recording parameters, workpiece recording parameters, environmental recording parameters, machining recording parameters, and the measured dimensions of the workpiece after cutting. The measured dimensions of the workpiece after cutting are the actual dimensions of the sample workpiece after machining.

[0084] Further analysis is needed on the slow wire EDM machining sample and the current machining conditions to determine whether the four constraints are met: the electrode wire recorded parameters are the same as the electrode wire parameters, the workpiece recorded parameters are the same as the workpiece parameters, the environment recorded parameters are consistent with the monitoring environment parameters, and the first machining parameter is the same as the machining recorded parameters.

[0085] The method for determining whether the recorded parameters of the electrode wire are the same as the current electrode wire parameters is as follows: if it is a type parameter, such as material type, it must be completely consistent; if it is a numerical parameter, such as wire diameter, the deviation between the two must be less than or equal to the corresponding attribute deviation threshold, such as wire diameter deviation ≤ 0.001 mm.

[0086] The method for determining whether the workpiece recorded parameters are the same as the current workpiece parameters is as follows: type parameters, such as workpiece material, must be completely consistent; the deviation of numerical parameters such as workpiece thickness must be ≤ the corresponding attribute deviation threshold, such as workpiece thickness deviation ≤ 0.01mm.

[0087] The method for determining whether environmental recording parameters are consistent with monitoring environmental parameters is as follows: using the aforementioned environmental inconsistency coefficient determination logic, if the environmental inconsistency coefficient between the environmental recording parameters of the slow wire EDM sample to be analyzed and the current monitoring environmental parameters is less than the inconsistency coefficient threshold, then they are considered consistent.

[0088] The method for determining whether the first processing parameter is the same as the processing record parameter of the sample is as follows: the type parameter is completely consistent, and the deviation of numerical parameters such as discharge voltage is ≤ the corresponding attribute deviation threshold, such as voltage deviation ≤ 1V.

[0089] The measured dimensions of the workpiece after cutting, which meet all constraints, will be included in the slow wire EDM sample set.

[0090] When the amount of data in the slow wire EDM machining sample set is greater than or equal to the preset fitting data threshold, such as 50 sets, a mode analysis is performed on the dimensions of all workpiece measurement records after cutting in the sample set—the dimension value with the highest frequency is counted, and this value is the fitted dimension of the workpiece after cutting, representing the workpiece dimension most likely to occur under the current processing conditions.

[0091] In step S300 of this application embodiment, updating the first processing parameters includes:

[0092] Obtain the updated processing parameters;

[0093] When the amount of data for the updated processing parameters is less than the optimized data amount threshold, the updated processing parameters are used as forbidden processing parameters, and random configuration of processing parameters is performed to update the first processing parameters. The updated processing parameters are different from the forbidden processing parameters.

[0094] When the amount of updated processing parameters is greater than or equal to the optimization data threshold, the processing parameters are optimized based on the updated processing parameters, using size deviation as the fitness index, and the first processing parameters are updated.

[0095] In this embodiment of the application, the purpose of step S300 is to flexibly select a suitable parameter update strategy based on the amount of data of the updated processing parameters, so as to achieve effective optimization of the first processing parameters, and ensure that when the monitoring environment is inconsistent with the standard environment and the deviation of the fitted size exceeds the tolerance, the processing parameters that are suitable for the current conditions can be obtained quickly, thereby ensuring that the size of the workpiece after processing meets the standard requirements and improving the stability of processing accuracy.

[0096] First, it is necessary to obtain the updated processing parameters, that is, to collect similar processing scenarios with similar electrode wire parameters, workpiece parameters, and environmental parameters in the past, and to form an updated processing parameter dataset.

[0097] When the amount of data for updated processing parameters is less than the optimization data threshold, such as less than 30 sets, the updated processing parameters are set as prohibited processing parameters, meaning that these parameters are prohibited from being used again.

[0098] Then, the processing parameters are randomly configured. Within the preset range of processing parameter values, such as the reasonable range of discharge voltage, current, pulse interval, etc., new processing parameters are randomly generated, ensuring that the new parameters are different from the prohibited processing parameters. The difference is that the type parameters are completely different, and the numerical parameter deviation is greater than the corresponding attribute deviation threshold. The first processing parameters are updated with the generated new parameters.

[0099] When the amount of data with updated processing parameters is greater than or equal to the optimization data amount threshold:

[0100] Using dimensional deviation as a fitness index, the smaller the deviation, the higher the fitness.

[0101] Based on the updated processing parameter dataset, processing parameters can be optimized using methods such as genetic algorithms and particle swarm optimization. The optimal parameters are selected by searching the processing parameter space for the combination with the highest fitness, i.e., the smallest dimensional deviation. The first processing parameters are then updated using these optimal parameters.

[0102] This case-by-case update strategy quickly generates usable parameters to ensure processing continuity when data is insufficient, and improves processing accuracy by finding better parameters when data is sufficient, thus balancing efficiency and optimization effect.

[0103] Example 2, as Figure 2As shown, based on the same inventive concept as the intelligent matching method for slow wire EDM machining parameters provided in Embodiment 1, this embodiment of the invention also provides an intelligent matching system for slow wire EDM machining parameters, comprising:

[0104] The first processing parameter acquisition module 11 is used to load electrode wire parameters and workpiece parameters to execute processing parameter configuration and obtain the first processing parameter, wherein the first processing parameter has a label identifying standard environmental parameters;

[0105] The workpiece fitting size acquisition module 12 after cutting is used to retrieve the slow wire EDM processing sample set, perform lumped value evaluation, and obtain the workpiece fitting size after cutting when the monitored environmental parameters are inconsistent with the standard environmental parameters, using the electrode wire parameters, the workpiece parameters, the first processing parameters and the monitored environmental parameters as constraints;

[0106] The first processing parameter evaluation module 13 is used to perform metal cutting control based on the first processing parameters when the dimensional deviation between the fitted size of the cut workpiece and the standard size of the cut workpiece is less than or equal to the dimensional tolerance; otherwise, it updates the first processing parameters.

[0107] Furthermore, the first processing parameter acquisition module 11 includes the following execution steps:

[0108] Configure the standard environment parameters through the user terminal;

[0109] Constrained by the standard environmental parameters and machine tool model, multiple sets of data are collected, wherein each set of multiple sets of data includes electrode wire parameter recording data, workpiece parameter recording data, and a label identifying the selected processing parameters;

[0110] Using the labels of the selected processing parameters as supervision, and the electrode wire parameter recording data and the workpiece parameter recording data as input, the processing parameter configuration model is trained by retrieving the multiple sets of data and using machine learning.

[0111] The electrode wire parameters and the workpiece parameters are processed through the processing parameter configuration model, and the first processing parameters are output.

[0112] And based on the standard environmental parameters, labels that identify the standard environmental parameters are constructed.

[0113] Using the labels of the selected processing parameters as supervision, and taking the electrode wire parameter recording data and the workpiece parameter recording data as input, the processing parameter configuration model is trained through machine learning by retrieving the multiple sets of data, including:

[0114] Based on a fully connected neural network, a first branch, a second branch, and a backbone network are constructed respectively. The first branch and the second branch are connected in parallel to the backbone network to obtain the processing parameter configuration model architecture.

[0115] Using the electrode wire parameter recording data as the first branch input, the workpiece parameter recording data as the second branch input, and the label identifying the selected processing parameter as supervision, the processing parameter configuration model architecture is trained to obtain the processing parameter configuration model.

[0116] Furthermore, the workpiece fitting size acquisition module 12 after cutting includes the following execution steps:

[0117] Using the electrode wire parameters and the workpiece parameters as constraints, a machining dimension correlation analysis is performed on the environmental parameter attributes to obtain a correlation degree set;

[0118] Traverse the set of correlation degrees and compare it with the sum of the correlation degree sets to obtain the weight distribution of environmental parameter attributes;

[0119] Calculate the normalized deviation of the environmental parameter attributes between the monitored environmental parameters and the standard environmental parameters;

[0120] Based on the weight distribution of the environmental parameter attributes, and combined with the normalized deviation of the environmental parameter attributes, a comprehensive deviation analysis is performed to obtain the environmental inconsistency coefficient.

[0121] When the environmental inconsistency coefficient is greater than or equal to the inconsistency coefficient threshold, the monitored environmental parameters are considered to be inconsistent with the standard environmental parameters.

[0122] Otherwise, the monitoring environment parameters are considered to be consistent with the standard environment parameters.

[0123] When the monitored environmental parameters are consistent with the standard environmental parameters, metal cutting control is executed based on the first processing parameters.

[0124] Obtain a slow wire EDM machining sample to be analyzed, wherein the slow wire EDM machining sample to be analyzed includes electrode wire recording parameters, workpiece recording parameters, environmental recording parameters, machining recording parameters, and workpiece measurement recording dimensions after cutting;

[0125] When the electrode wire recording parameters are the same as the electrode wire parameters, the workpiece recording parameters are the same as the workpiece parameters, the environmental recording parameters are consistent with the monitoring environment parameters, and the first processing parameters are the same as the processing recording parameters, it is considered that the slow wire EDM processing sample to be analyzed satisfies the constraints. The measured and recorded dimensions of the workpiece after cutting are added to the slow wire EDM processing sample set. In this case, any type of parameter that represents the same two sets of data is the same, and any numerical parameter deviation is less than or equal to the corresponding attribute deviation threshold.

[0126] When the amount of data in the slow wire EDM machining sample set is greater than or equal to the threshold of the amount of fitted data, a mode analysis is performed on the slow wire EDM machining sample set to obtain the fitted dimensions of the workpiece after cutting.

[0127] Furthermore, the first processing parameter evaluation module 13 includes the following execution steps:

[0128] Obtain the updated processing parameters;

[0129] When the amount of data for the updated processing parameters is less than the optimized data amount threshold, the updated processing parameters are used as forbidden processing parameters, and random configuration of processing parameters is performed to update the first processing parameters. The updated processing parameters are different from the forbidden processing parameters.

[0130] When the amount of updated processing parameters is greater than or equal to the optimization data threshold, the processing parameters are optimized based on the updated processing parameters, using size deviation as the fitness index, and the first processing parameters are updated.

[0131] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0132] Those skilled in the art will understand that embodiments of the present invention can provide methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0134] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0135] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0136] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0137] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for intelligent matching of machining parameters in slow wire EDM, characterized in that, Applications in slow wire EDM machines include: Load electrode wire parameters and workpiece parameters to execute machining parameter configuration and obtain first machining parameters, wherein the first machining parameters have labels identifying standard environmental parameters; When the monitored environmental parameters are inconsistent with the standard environmental parameters, the electrode wire parameters, the workpiece parameters, the first processing parameters and the monitored environmental parameters are used as constraints to retrieve the slow wire EDM processing sample set, perform lumped value evaluation, and obtain the fitted size of the workpiece after cutting. When the dimensional deviation between the fitted size of the cut workpiece and the standard size of the cut workpiece is less than or equal to the dimensional tolerance, metal cutting control is executed based on the first processing parameters; otherwise, the first processing parameters are updated. When the monitored environmental parameters are inconsistent with the standard environmental parameters, including: Using the electrode wire parameters and the workpiece parameters as constraints, a machining dimension correlation analysis is performed on the environmental parameter attributes to obtain a correlation degree set; Traverse the set of correlation degrees and compare it with the sum of the correlation degree sets to obtain the weight distribution of environmental parameter attributes; Calculate the normalized deviation of the environmental parameter attributes between the monitored environmental parameters and the standard environmental parameters; Based on the weight distribution of the environmental parameter attributes, and combined with the normalized deviation of the environmental parameter attributes, a comprehensive deviation analysis is performed to obtain the environmental inconsistency coefficient. When the environmental inconsistency coefficient is greater than or equal to the inconsistency coefficient threshold, the monitored environmental parameters are considered to be inconsistent with the standard environmental parameters. Otherwise, the monitoring environment parameters are considered to be consistent with the standard environment parameters; Updating the first processing parameter includes: Obtain the updated processing parameters; When the amount of data for the updated processing parameters is less than the optimized data amount threshold, the updated processing parameters are used as forbidden processing parameters, and random configuration of processing parameters is performed to update the first processing parameters. The updated processing parameters are different from the forbidden processing parameters. When the amount of updated processing parameters is greater than or equal to the optimization data threshold, the processing parameters are optimized based on the updated processing parameters, using size deviation as the fitness index, and the first processing parameters are updated.

2. The method as described in claim 1, characterized in that, Load electrode wire parameters and workpiece parameters, execute machining parameter configuration, and obtain the first machining parameters, including: Configure the standard environment parameters through the user terminal; Constrained by the standard environmental parameters and machine tool model, multiple sets of data are collected, wherein each set of multiple sets of data includes electrode wire parameter recording data, workpiece parameter recording data, and a label identifying the selected processing parameters; Using the labels of the selected processing parameters as supervision, and the electrode wire parameter recording data and the workpiece parameter recording data as input, the processing parameter configuration model is trained by retrieving the multiple sets of data and using machine learning. The electrode wire parameters and the workpiece parameters are processed through the processing parameter configuration model, and the first processing parameters are output. And based on the standard environmental parameters, labels that identify the standard environmental parameters are constructed.

3. The method as described in claim 2, characterized in that, Using the labels of the selected processing parameters as supervision, and taking the electrode wire parameter recording data and the workpiece parameter recording data as input, the processing parameter configuration model is trained through machine learning by retrieving the multiple sets of data, including: Based on a fully connected neural network, a first branch, a second branch, and a backbone network are constructed respectively. The first branch and the second branch are connected in parallel to the backbone network to obtain the processing parameter configuration model architecture. Using the electrode wire parameter recording data as the first branch input, the workpiece parameter recording data as the second branch input, and the label identifying the selected processing parameter as supervision, the processing parameter configuration model architecture is trained to obtain the processing parameter configuration model.

4. The method as described in claim 1, characterized in that, Also includes: When the monitored environmental parameters are consistent with the standard environmental parameters, metal cutting control is executed based on the first processing parameters.

5. The method as described in claim 1, characterized in that, Using the electrode wire parameters, workpiece parameters, first processing parameters, and monitoring environment parameters as constraints, a slow wire EDM processing sample set is retrieved, and mode analysis is performed to obtain the fitted dimensions of the workpiece after cutting, including: Obtain a slow wire EDM machining sample to be analyzed, wherein the slow wire EDM machining sample to be analyzed includes electrode wire recording parameters, workpiece recording parameters, environmental recording parameters, machining recording parameters, and workpiece measurement recording dimensions after cutting; When the electrode wire recording parameters are the same as the electrode wire parameters, the workpiece recording parameters are the same as the workpiece parameters, the environmental recording parameters are consistent with the monitoring environment parameters, and the first processing parameters are the same as the processing recording parameters, it is considered that the slow wire EDM processing sample to be analyzed satisfies the constraints. The measured and recorded dimensions of the workpiece after cutting are added to the slow wire EDM processing sample set. In this case, any type of parameter that represents the same two sets of data is the same, and any numerical parameter deviation is less than or equal to the corresponding attribute deviation threshold. When the amount of data in the slow wire EDM machining sample set is greater than or equal to the threshold of the amount of fitted data, a mode analysis is performed on the slow wire EDM machining sample set to obtain the fitted dimensions of the workpiece after cutting.

6. A smart matching system for slow wire EDM machining parameters, characterized in that, The system is used to implement the intelligent matching method for slow wire EDM machining parameters as described in any one of claims 1-5, the system comprising: The first processing parameter acquisition module is used to load electrode wire parameters and workpiece parameters to execute processing parameter configuration and obtain the first processing parameter, wherein the first processing parameter has a label identifying standard environmental parameters; The workpiece fitting size acquisition module after cutting is used to retrieve the slow wire EDM processing sample set, perform lumped value evaluation, and obtain the workpiece fitting size after cutting when the monitored environmental parameters are inconsistent with the standard environmental parameters, using the electrode wire parameters, the workpiece parameters, the first processing parameters, and the monitored environmental parameters as constraints; The first processing parameter evaluation module is used to perform metal cutting control based on the first processing parameters when the dimensional deviation between the fitted size of the cut workpiece and the standard size of the cut workpiece is less than or equal to the dimensional tolerance; otherwise, it updates the first processing parameters.

Images