TWI778460B - Online prediction method of tool-electrode consumption and prediction method of machining accuracy - Google Patents

Abstract

An online prediction method of tool-electrode consumption suited for an electrical discharge machining （EDM） apparatus including design of experiment, electrode consumption variables are extracted from machining parameters of the EDM apparatus, and a correlation between the machining parameters and the electrode consumption variables through correlation analysis, and a prediction model is obtained to predict effective contacting area of the tool-electrode and a workpiece-electrode. A prediction method of machining accuracy is also provided.

Description

On-line prediction method of machining electrode consumption and prediction method of machining accuracy

The present invention relates to a method for predicting machining accuracy, and in particular, to an online predicting method for machining electrode consumption and a method for predicting machining accuracy.

When electrical discharge machining is performed, the workpiece size is not as expected, and the surface roughness of the workpiece is often poor due to factors such as poor slag discharge, abnormal short circuit or electrode consumption. Furthermore, the general electrical discharge machining equipment cannot effectively predict the machining accuracy of the workpiece during the machining process, and can only measure the machining accuracy of the machined workpiece off-line through the measuring equipment.

In order to solve the above deficiencies, a prediction model has been used to monitor relevant machining information during the process of electrical discharge machining to predict the machining accuracy of subsequent workpieces. However, the above model cannot reflect the consumption of the discharge electrode in real time, and the consumption of the discharge electrode can only be obtained by off-line measurement, which greatly affects the utilization rate of the electrical discharge machining equipment.

In view of this, there is a need to provide a method for predicting the accuracy of electrical discharge machining to solve the aforementioned problems.

The present invention provides an on-line prediction method of machining electrode consumption and a prediction method of machining accuracy, so as to predict the consumption of electric discharge electrodes in real time in the process of electric discharge machining, and adjust the machining parameters of electric discharge machining equipment accordingly, so as to obtain better electric discharge machining precision.

The online prediction method of machining electrode consumption of the present invention is suitable for electric discharge machining equipment, including: experimental design, extraction of variables affecting electrode consumption from machining parameters of electric discharge machining equipment, and obtaining machining parameters and influencing electrode consumption through correlation analysis. The correlation of variables to obtain a model that can predict the effective contact area between the machining electrode and the workpiece.

The method for predicting machining accuracy of the present invention includes: extracting a plurality of key machining characteristic values of electrical discharge machining equipment by a modeling substrate extraction method, wherein the key machining characteristic values include those obtained by the online prediction method of the aforementioned machining electrode consumption; The machining characteristic value establishes a prediction model of electric discharge machining accuracy; and real-time sensing of multiple sets of discharge voltage signals and multiple sets of discharge current signals of the electric discharge machining equipment are used as input values, and sent to the machining accuracy prediction model to output at least one electric discharge machining accuracy prediction in real time. value as the output value.

Based on the above, through the online prediction method of machining electrode consumption, through its experimental design, the variables affecting the electrode consumption are extracted from the machining parameters of the electrical discharge machining equipment, and the correlation between the machining parameters and the variables affecting the electrode consumption is obtained by correlation analysis. to obtain a model that can predict the effective contact area between the machining electrode and the workpiece. In this way, the effective contact area obtained from the model can represent the discharge machining capability that the current discharge electrode can perform on the workpiece, and it is also equivalent to reflect the consumption level of the discharge electrode after the previous processing, so the relevant control system is The relevant machining parameters of the electric discharge machining equipment can be adjusted adaptively according to the current electric discharge machining capability of the electric discharge electrode, so as to be adapted to the machining process required by the workpiece and gain the machining accuracy.

FIG. 1 is a schematic diagram of a prediction system of electric discharge machining accuracy according to an embodiment of the present invention. Referring to FIG. 1 , the electric discharge machining accuracy prediction system 200 of the present embodiment includes an electric discharge machining equipment 210 and a data processing module 220 , and the data processing module 220 is connected to the electric discharge machining equipment 210 via a message. The electrical discharge machining apparatus 210 has a main shaft 211 and a tool-electrode 211 a provided on the main shaft 211 , so as to perform electrical discharge machining on the workpiece P1 through the machining electrode 211 a. The data processing module 220 includes a control feedback unit 221 , a sensing unit 222 and a processing unit 223 . The control feedback unit 221 is used for sensing the distance signal, and is used for judging whether the sensing unit 222 needs to capture the process data (including the discharge voltage signal and the discharge current signal) during the online processing of the electrical discharge machining equipment 210 . The processing unit 223 can create a plurality of processing features according to the process data captured by the sensing unit 222 , and can create a plurality of processing features according to the measurement values of different measurement items processed by the measurement machine 300 (the measurement values refer to the electrical discharge machining equipment). After the workpiece is processed, the measured value of the workpiece quality), and then the key processing feature values related to the machining accuracy are collected from the machining features to provide the automation server 400 and establish an electric discharge machining accuracy prediction model accordingly to predict the electric discharge machining. The subsequent processing accuracy of the processing equipment 210 .

For example, the automation server 400 uses a neural network (NN)-like method and a regression analysis (Regression Analysis) method (eg, a partial least squares method (PLS)) to establish the electric discharge machining accuracy prediction model. To illustrate, the prediction model used in this embodiment may be the machining accuracy prediction model established by the automatic server disclosed in Patent No. TW I349867, but this is not intended to limit the present invention.

Roughly speaking, the modeling substrate extraction method of this embodiment mainly collects process data (such as electrode consumption value, discharge voltage signal and discharge current signal) of the electrical discharge machining equipment 210 during the discharge process, and then establishes a multi-step process based on the process data. processing features, and applying data pre-processing technology (Data Pre-Processing Technology) to extract key processing feature values that can be used to estimate machining accuracy from these processing features. The key machining characteristic values collected can be mainly provided to the automation server 400 to establish a prediction model of the electric discharge machining accuracy, which is used to predict the machining accuracy of the electric discharge machining equipment 210 .

In detail, FIG. 2A and FIG. 2B are schematic diagrams of establishing a prediction model according to the present invention, respectively. FIG. 2C is a schematic diagram of the input and output of the electric discharge machining accuracy prediction model. FIG. 3A is a flow chart of the establishment of the electric discharge machining accuracy prediction model of the present invention. Figure 3B is a flow chart of building the prediction model of Figure 2A. FIG. 3C is a flow chart of establishing the electric discharge machining accuracy prediction model of FIG. 2B . Here, the model establishment in FIG. 2A corresponds to step S110 in the flow of FIG. 3A , and the model establishment in FIG. 2B corresponds to step S130 in the flow of FIG. 3A .

Please refer to FIG. 2A and FIG. 3A first, in step S110 , the key processing feature value group A of the electrical discharge machining equipment 210 is extracted by the modeling substrate extraction method to establish a preliminary prediction model, wherein the key processing feature value group A includes: Offline electrode consumption value. Next, in step S120, an online prediction model of machining electrode consumption is obtained from the initial prediction model by the online prediction method of machining electrode consumption. Next, in step S130, the key machining feature value group B of the electrical discharge machining equipment 210 is extracted by the modeling substrate extraction method to establish an EDM accuracy prediction model, wherein the key machining feature value group B includes the wires consumed by the machining electrodes. The online electrode consumption value obtained by the prediction model. Finally, in step S140, when the electrical discharge machining equipment 210 is performing subsequent electrical discharge machining, the electrical discharge machining equipment 210 can immediately sense multiple sets of discharge voltage signals and multiple sets of electrical discharge current signals as input values, and transmit them to the electrical discharge machining accuracy The prediction model generates a new machining current value and at least one electric discharge machining accuracy prediction value as an output value, wherein the generated new machining current value can be transmitted to the controller 212, so that the electric discharge machining equipment 210 can be used as a driving The basis for the electrical discharge machining action.

To put it simply, during the EDM process, since the online prediction model of machining electrode consumption can instantly reflect the current state of the machining electrode 211a, the effect of the aforementioned step S140 is that in addition to providing the EDM accuracy prediction value, the operator can In addition to the knowledge, the discharge current (ie, the above-mentioned new machining current value) will also be adjusted accordingly, so that it can be appropriately adjusted according to the consumption of the machining electrode 211a during the electric discharge machining.

Please refer to FIG. 3B , which further describes the relevant detailed steps of step S110 . In step S111 , the electrical discharge machining equipment 210 processes a plurality of workpieces P1 respectively to obtain a plurality of sets of process data, wherein each set of process data includes an offline electrode consumption value, a discharge voltage signal and a discharge current signal. Here, a high voltage probe and a current hook can be used to sense the discharge voltage and discharge current of the electrical discharge machining equipment 210, and the image sensor 222a can be used to measure the electrode consumption value. Next, in step S112, a plurality of machining features are created from the process data.

It should be noted that the processing characteristics of this embodiment include electrode consumption value, discharge frequency, open circuit ratio, short circuit ratio, average short circuit time, short circuit time standard deviation, average short circuit current, short circuit current standard deviation, average delay time, delay time standard deviation , average discharge peak current, peak current standard deviation, average discharge time, discharge time standard deviation, average discharge energy and discharge energy standard deviation.

In the above machining features, the average delay time and the short circuit ratio are established from the discharge voltage signal, wherein the average delay time is defined as the time from the point when a sufficient open circuit voltage has been established until the voltage pulse crosses the gap between the electrode and the workpiece, and Time difference until discharge current starts. Short circuit ratio (Short circuit ratio) is defined as the number of short circuit pulses (SCP) divided by the number of discharge pulses, where the short circuit pulse is when the open circuit voltage value is continuously lower than the specified voltage threshold during a discharge pulse period, then the current The discharge pulse period is recorded as a short-circuit pulse.

In the above processing features, the discharge frequency, the average discharge peak current and the average discharge time are established from the discharge current signal, wherein the average spark frequency (Average spark frequency) Exceeding the minimum threshold peak value is defined as the occurrence of current sparks, and the discharge frequency is defined as the total number of sparks occurring during the sampling period. The average peak discharge current (Average peak discharge current) is defined as the average of all discharge peak currents (peak current) during the sampling period, where the peak current is the high current value that reaches the workpiece through the electrode during the pulse period. The average discharge current pulse duration (Average discharge current pulse duration) is defined as the average value of all current pulse durations during the sampling period, where the current pulse duration is the time difference between the start point and the end point of the discharge current waveform.

In the above processing features, the average short-circuit time, open-circuit ratio, average discharge energy and average short-circuit current are jointly established according to the discharge current signal and the discharge voltage signal, wherein the average short-circuit time is related to the short-circuit duration, and the short-circuit duration ( Short circuit duration) is defined as when multiple consecutive short circuits occur during a period of discharge pulse (more than two consecutive pulses are required), then the short circuit duration is the period from the first short circuit peak (short circuit peak) to the last The time difference between the peaks of a short-circuit pulse. The open circuit ratio is defined as the number of open circuits divided by the total number of discharge pulses during the sampling period. In a certain pulse time, when the voltage peak ends and the current peak does not rise, it is called an open circuit (Open circuit). If an open circuit occurs, it means that a voltage peak (Ignition Voltage) cannot lead to a subsequent current peak (Discharge Current), and this voltage peak is an invalid pulse. The average discharge energy is mainly used to maintain the stability of the discharge machining process to ensure the processing quality, and the discharge energy (E) formula of the i-th discharge is as follows, where t _ei is the discharge duration, U _i is the discharge voltage, and I _pi is the discharge peak current, this formula assumes that the discharge voltage remains unchanged during the discharge process.

To clarify, according to the foregoing disclosure, the calculation methods of the standard deviation values of the short-circuit time standard deviation, the short-circuit current standard deviation, the delay time standard deviation, the peak current standard deviation, the discharge time standard deviation and the discharge energy standard deviation belong to the present invention. It is well known to those with ordinary knowledge in the technical field, so it is not repeated here.

Please refer to FIG. 3B again. Next, in step S113 , a plurality of workpieces P1 are measured by the measuring machine 300 to obtain a plurality of measurement values, wherein each measurement value is obtained by the electrical discharge machining equipment 210 according to a plurality of sets of process data. Measured values of multiple workpieces P1 processed. For example, the machining electrodes 211 a with different shapes and sizes are used to perform electrical discharge machining on a plurality of workpieces P1 to form blind holes. After completion, the measurement machine 300 is used to measure the size and roughness of these blind holes.

After the measurement of the measurement values is completed, the correlation analysis can be performed in step S114 to obtain a plurality of correlation values between the plurality of processing features and the plurality of measurement values. After finding out the correlation value between the machining feature and the measured value, then step S115 is performed, and the machining feature with the larger correlation value is selected from the correlation value as a plurality of key machining feature values (that is, the aforementioned Key processing feature value group A), wherein, in this embodiment, processing features whose absolute value of correlation value is greater than 0.3 are selected. At this point, it is necessary to complete step S110 , that is, to construct the relevant key processing characteristic values including the offline electrode consumption value as an initial prediction model.

4A and 4B are flowcharts of establishing the prediction model of FIG. 2B . Please refer to FIG. 4A and FIG. 4B at the same time, and the foregoing step S120 will be described in detail here. In this embodiment, the experimental design of step S210 is performed first, and then step S220 is performed to extract the variables affecting electrode consumption from the machining parameters of the electrical discharge machining equipment 210 . The processing parameters described herein are obtained from the relevant processing features of the aforementioned preliminary prediction model (and especially other key processing feature values other than the aforementioned offline electrode consumption values). Finally, in step S230, the correlation between the machining parameters and the variables affecting the electrode consumption is obtained through correlation analysis, so as to establish a machining electrode consumption prediction model. Please refer to FIG. 4B again. More specifically, the aforementioned step S230 further includes step S231, which judges whether the data is abnormal, and if so, executes step S220 again, that is, re-extracts the variable affecting electrode consumption from the machining parameters of the electrical discharge machining equipment; If not, go to step S232 to determine whether the regression equation is feasible by F-test, if yes, go to subsequent step S233, if not, go back to step S220 to extract the influence from the machining parameters of the electrical discharge machining equipment 210 again Electrode consumption variable; if yes, then carry out residual analysis and judgment in step S233, and then execute step S234, find and remove the electrode consumption variable with extremely low correlation by t-test. At this point, the step S230 for establishing the prediction model of machining electrode consumption is completed.

Next, referring to FIGS. 2B and 3C , after obtaining the machining electrode consumption prediction model from the initial prediction model as described above, it means that the electric discharge machining equipment 210 can obtain the consumption condition of the machining electrode 211 a during the electric discharge machining process (the online electrode consumption value ), so again similar to the process shown in FIG. 3B , and in step S111A shown in FIG. 3C , the electrical discharge machining equipment 210 is used to process a plurality of workpieces P1 respectively to obtain multiple sets of process data, wherein each set of process data includes on-line electrode consumption value, discharge voltage signal and discharge current signal. Next, steps S112 to S114 are performed, as shown in FIG. 3B . Finally, in step S115A, the processing feature with the absolute value of the correlation value greater than 0.3 is selected from the plurality of correlation values as the key processing feature value group B, and is input into step S130 shown in FIG. 3A accordingly, that is, The key machining feature group B including the on-line electrode consumption value is established by modeling the substrate extraction method to establish an EDM prediction model. Finally, as shown in step S140 of FIG. 2C and FIG. 3A , during the electrical discharge machining process, the electrical discharge machining accuracy can be successfully predicted by combining the real-time sensing discharge voltage signal and discharge current signal with the online electrode consumption value. And a new machining current value is generated for the electrical discharge machining equipment 210 to perform subsequent machining, wherein the key machining feature value group B is used to improve the prediction ability of the model.

In this embodiment, the variables affecting electrode consumption mentioned in the aforementioned step S220 include the number of effective discharges, the machining depth and the accumulated machining time, wherein the effective number of discharges is the total number of discharges minus the number of abnormal discharges. Here, the number of abnormal discharges includes the times when the rated voltage (Ue) is lower than the set level (equivalent to abnormal discharge R _A ) during discharge and the discharge delay time (TD) is lower than the standard time (equivalent to abnormal discharge R _B ) send times. FIG. 5 is a schematic diagram of a discharge voltage waveform. Figure 6 provides experimental data affecting electrode consumption variables and effective contact area. Please refer to FIG. 5 and FIG. 6 at the same time. In step S230 of this embodiment, the regression equation constructed by the correlation analysis is: A _eff =A ₁ +A ₂ *N _total -A ₃ *R _B -A ₄ *L +A ₅ *T+A ₆ *L*TA ₇ *R _B *L+A ₈ *N _total *L+A ₉ *N _total *R _B -A ₁₀ N _total *TA ₁₁ *N _total *L*T +A ₁₂ N _total *L*T*R _B ,

The above A _eff is the effective contact area between the machining electrode 211a and the workpiece P1, N _total is the total number of discharges, while R _B is regarded as the number of times that the discharge delay time (TD) is lower than the set time, and L is the machining depth , T is the accumulated processing time. The waveforms of the three voltage signals are shown in Figure 5. The leftmost is the voltage waveform that meets the standard discharge required for processing, with the standard voltage level V _S and the standard discharge delay time _{T S} , and the middle is the abnormal discharge R _A , that is, The discharge voltage waveform whose rated voltage is lower than the standard level VS during discharge, and the abnormal discharge _RB on the right side, that is, the discharge voltage _waveform whose discharge delay time is lower than the standard time _TDS , so the effective discharge times is the total discharge times. Subtract the number of abnormal discharges ( _RA and _RB ) as shown in Figure 5. It should also be noted that the parameters used in the above regression equation can be appropriately changed according to the conditions of the actual processing equipment. For example, if the electrical discharge machining equipment can avoid abnormal discharge when the rated voltage (Ue) is lower than the set level through the circuit control, it does not need to be included in the above regression equation.

In addition, as shown in FIG. 6 , the variables that affect the electrode consumption are verified by experiments, that is, the total number of discharges N _total , the number of abnormal discharges ( _RA and R _B ), the machining depth L, the accumulated machining time T and the effective The correlation of the contact area (A _eff ), which is indicated by the correlation index (FITS) in the table, from which it can be clearly seen that as the effective contact area (A _eff ) decreases, the correlation index (FITS) also will decrease accordingly, which means that these variables affecting electrode consumption are indeed positively correlated with the effective contact area. At the same time, for the electrical discharge machining process, the shape (including geometric shape and size) of the machining electrode 211a is one of the main factors affecting the electrical discharge machining accuracy. Therefore, in this embodiment, the effective contact area can be obtained immediately during the machining process. Master the conditions of electrical discharge machining accuracy. Accordingly, the present embodiment can grasp the current situation of the machining electrode 211 a in real time by using the online prediction method (and its model) of the machining electrode, and accordingly propose a corresponding adaptation solution.

To sum up, in the above-mentioned embodiments of the present invention, the on-line prediction method of machining electrode consumption is used to extract the variables affecting electrode consumption from the machining parameters of electric discharge machining equipment through its experimental design, and through correlation analysis The correlation between the machining parameters and the variables affecting the electrode consumption is obtained, and a model that can predict the effective contact area between the machining electrode and the workpiece is obtained. In this way, the effective contact area obtained from the model can represent the discharge machining capability that the current discharge electrode can perform on the workpiece, and it is also equivalent to reflect the consumption level of the discharge electrode after the previous processing, so the relevant control system is The relevant machining parameters of the electric discharge machining equipment can be adjusted adaptively according to the current electric discharge machining capability of the electric discharge electrode, so as to be adapted to the machining process required by the workpiece and gain the machining accuracy.

200: Prediction system for electrical discharge machining accuracy 210: Electric Discharge Machining Equipment 211: Spindle 211a: Machining Electrodes 220: Data processing module 221: Control feedback unit 222: Sensing unit 223: Processing unit 300: Measuring machine 400: Automation Server P1: Workpiece S110～S140, S111～S115, S111A: Steps S210～S230, S231～S234: Steps

FIG. 1 is a schematic diagram of a prediction system of electric discharge machining accuracy according to an embodiment of the present invention. 2A and 2B are schematic diagrams of establishing a prediction model according to the present invention, respectively. FIG. 2C is a schematic diagram of the input and output of the electric discharge machining accuracy prediction model. FIG. 3A is a flow chart of the establishment of the electric discharge machining accuracy prediction model of the present invention. Figure 3B is a flow chart of building the prediction model of Figure 2A. FIG. 3C is a flow chart of establishing the electric discharge machining accuracy prediction model of FIG. 2B . 4A and 4B are flowcharts of establishing the prediction model of FIG. 2B . FIG. 5 is a schematic diagram of a discharge voltage waveform. Figure 6 provides experimental data affecting electrode consumption variables and effective contact area.

S210, S220, S230: steps

Claims (11)

An online prediction method for machining electrode consumption, suitable for electric discharge machining equipment, comprising: experimental design; extracting variables affecting electrode consumption from machining parameters of the electric discharge machining equipment; and obtaining the machining parameters and all the parameters through correlation analysis The correlation of the variables affecting the electrode consumption is established to establish a model that can predict the effective contact area between the machining electrode and the workpiece, wherein the variables affecting the electrode consumption include the number of effective discharges, the machining depth, and the accumulated machining time. The online prediction method for machining electrode consumption according to claim 1, wherein the number of effective discharges is the total number of discharges minus the number of abnormal discharges. The online prediction method for machining electrode consumption according to claim 2, wherein the abnormal discharge times include the times when the rated voltage (Ue) is lower than the set level during the discharge and the times when the discharge delay time (TD) is less than the set time. . The online prediction method for machining electrode consumption according to claim 3, wherein the regression equation generated by the correlation analysis method is: A _eff =A ₁ +A ₂ *N _total -A ₃ *R _B -A ₄ *L+ A ₅ *T+A ₆ *L*TA ₇ *R _B *L+A ₈ *N _total *L+A ₉ *N _total *R _B -A ₁₀ N _total *TA ₁₁ *N _total *L*T+ A ₁₂ N _total *L*T*R _B , where A _eff is the effective contact area, N _total is the total number of discharges, _RB is the number of times the discharge delay time (TD) is lower than the set time, and L is the The machining depth, T is the accumulated machining time. The online prediction method for machining electrode consumption according to claim 1, wherein the correlation analysis includes: judging whether the data is abnormal; judging whether the regression equation is feasible by F-test; judging by residual analysis; and A t-test was used to identify and remove variables that had very low correlations affecting electrode consumption. A method for predicting machining accuracy, comprising: extracting a key machining feature value group A of an electrical discharge machining device by a modeling substrate extraction method to establish a preliminary prediction model, wherein the key machining feature value group A includes an offline electrode consumption value Obtaining an online prediction model of machining electrode consumption from the initially built prediction model by using the online prediction method for machining electrode consumption according to any one of claim 1 to claim 5; extracting electrical discharge with the modeling base extraction method A key machining feature value group B of the machining equipment is used to establish an EDM accuracy prediction model, wherein the key machining feature value group B includes the online electrode consumption values obtained by the online prediction model of machining electrode consumption. The method for predicting machining accuracy according to claim 6, further comprising: Immediately sense multiple sets of discharge voltage signals and multiple sets of discharge current signals of the electrical discharge machining equipment as input values, and send them to the electrical discharge machining accuracy prediction model to generate machining current and at least one electrical discharge machining accuracy prediction value as output values . The method for predicting machining accuracy according to claim 6, wherein the step of extracting the key machining feature value group A of the electrical discharge machining equipment comprises: using the electrical discharge machining equipment to process a plurality of workpieces respectively to obtain multiple sets of process data , wherein each set of process data includes the offline electrode consumption value, the discharge voltage signal and the discharge current signal; the multiple sets of process data are used to establish a plurality of processing features; the measurement equipment is used to measure the plurality of workpieces to obtain a plurality of measurement values, wherein each measurement value is a measurement value of the plurality of workpieces processed by the electrical discharge machining equipment according to the plurality of sets of process data; and a correlation analysis is performed to obtain the plurality of processing features a plurality of correlation values with the plurality of measurement values; and selecting a machining feature with an absolute value of the correlation value greater than 0.3 from the plurality of correlation values as the plurality of key machining feature values. The method for predicting machining accuracy according to claim 8, wherein the measurement value is a roughness measurement value of the plurality of workpieces, and the plurality of machining characteristics include the offline electrode consumption value, discharge frequency, open circuit ratio, short-circuit ratio, average short-circuit time, short-circuit time standard deviation, average short-circuit current, short-circuit current standard deviation, average delay time, delay time standard deviation, average discharge peak current, peak current standard deviation, average Average discharge time, standard deviation of discharge time, average discharge energy, and standard deviation of discharge energy. The method for predicting machining accuracy according to claim 6, wherein the step of extracting the key machining feature value group B of the electrical discharge machining equipment comprises: processing a plurality of workpieces with the electrical discharge machining equipment respectively to obtain multiple sets of process data , wherein each set of process data includes the on-line electrode consumption value, the discharge voltage signal and the discharge current signal; multiple sets of process data are used to establish multiple processing features; the multiple workpieces are measured with a measuring device to obtain multiple measurement values, wherein each measurement value is a measurement value of the plurality of workpieces processed by the electrical discharge machining equipment according to the plurality of sets of process data; and a correlation analysis is performed to obtain the plurality of processing features a plurality of correlation values with the plurality of measurement values; and selecting a machining feature with an absolute value of the correlation value greater than 0.3 from the plurality of correlation values as the plurality of key machining feature values. The method for predicting machining accuracy according to claim 10, wherein the measurement value is a roughness measurement value of the plurality of workpieces, and the plurality of machining characteristics include the wire electrode consumption value, discharge frequency, open circuit ratio, short circuit ratio, average short circuit time, short circuit time standard deviation, average short circuit current, short circuit current standard deviation, average delay time, delay time standard deviation, average discharge peak current, peak current standard deviation, average discharge time, discharge time standard deviation , the average discharge energy and the standard deviation of the discharge energy.

Images