A method and apparatus for wire sawing
By dynamically calculating the target discharge gap and current cutting offset of the wire EDM wire, and combining it with CNC system control, the problem of balancing local precision and efficiency in precision small structural parts by wire EDM technology is solved, achieving high-precision and low-cost processing results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUIZHOU ZHONGRUI INTELLIGENT EQUIP CO LTD
- Filing Date
- 2026-04-30
- Publication Date
- 2026-06-05
Smart Images
Figure CN122142434A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of wire EDM technology, and in particular to a wire EDM processing method and equipment. Background Technology
[0002] As consumer electronics products become increasingly miniaturized and precision-oriented, the processing accuracy requirements for precision structural components such as mobile phone camera bezels, fingerprint buttons, and card trays are constantly rising. Wire EDM, with its micron-level processing capabilities and lack of limitations due to material hardness, has become one of the main processing methods for such parts.
[0003] Existing wire EDM technology mainly employs two compensation methods: one is to correct the global cutting offset by statistically analyzing the overall dimensional deviation after batch processing; the other is to adjust the tension by monitoring changes in the molybdenum wire size to offset drift caused by wear. However, both of these methods are global operations and cannot provide differentiated compensation for local geometric features of the part (such as inner right angles, small arcs, and thin-walled structures). For precision small structural parts such as mobile phone camera bezels and fingerprint buttons, global compensation often leads to overcutting at corners or undercutting in straight sections, and the trial cutting and adjustment cycle is long, resulting in significant material waste, making it difficult to balance processing accuracy and efficiency. Summary of the Invention
[0004] This application provides a wire EDM processing method and equipment to solve the technical problem that existing wire EDM technology's compensation methods are unable to balance local accuracy and processing efficiency.
[0005] To address the aforementioned technical problems, in a first aspect, this application provides a wire EDM processing method, comprising: The current processing position of the workpiece to be processed is obtained, and the geometric features of the workpiece corresponding to the current processing position are determined. The geometric features include straight line features and circular arc features. Based on the geometric features, the target discharge gap corresponding to the cutting wire is determined. The target discharge gap is used to characterize the shortest distance between the etched surface of the workpiece and the outer surface of the cutting wire. Based on the target discharge gap and the cross-sectional radius of the cutting wire, the current cutting offset of the cutting wire is obtained, wherein the current cutting offset is equal to the sum of the target discharge gap and the cross-sectional radius. Based on the current cutting offset, the cutting wire is controlled to perform cutting processing.
[0006] This wire EDM method incorporates the geometric features of the machining path into the dynamic calculation of the cutting offset, avoiding the overcutting of the arc segment due to concentrated discharge by treating the discharge gap of the cutting wire as a constant. In this method, the current machining position is first acquired in real time and its geometric features are identified: if it is a straight line, a preset reference discharge gap is used; if it is an arc, a larger target discharge gap is obtained through a correction process. Then, the target discharge gap is summed with the cross-sectional radius of the cutting wire to obtain the current cutting offset. This current cutting offset determines the position of the cutting wire's centerline relative to the theoretical contour. Finally, the CNC system controls the cutting wire to perform cutting according to this current cutting offset. This method uses engineering modeling to model the physical coupling between the path's geometric features and the discharge gap, realizing the transformation from a globally constant offset to a geometrically adaptive offset. By adopting this wire EDM method, the machining accuracy of the workpiece's arc features is significantly improved, and the overcutting is significantly reduced. Secondly, the compensation for straight and arc features is independent, avoiding excessive correction of straight features at the expense of efficiency. Furthermore, this method is entirely based on software algorithms, requiring no hardware modification and resulting in low deployment costs. For precision parts with minute arc features, such as mobile phone camera bezels and fingerprint buttons, this method can stably control the contour error within ±0.003mm.
[0007] In one embodiment, determining the target discharge gap corresponding to the cutting wire based on the geometric features includes: If the geometric feature is a straight line feature, then the target discharge gap is set as a preset reference discharge gap; If the geometric feature is a circular arc feature, then a target correction coefficient is determined based on the circular arc feature, and the reference discharge gap is corrected based on the target correction coefficient to obtain the target discharge gap.
[0008] In this embodiment, the target discharge gap for straight features is fixed as the reference discharge gap, ensuring high efficiency and dimensional stability in straight-line machining. For circular features, the reference discharge gap is amplified and corrected using a target correction coefficient, directly compensating for the increased actual discharge gap caused by concentrated discharge at the arc. This distinction avoids over-compensation of straight features resulting in undersized dimensions, and also prevents insufficient compensation from causing edge collapse in circular features. Operators only need to calibrate a reference discharge gap, and the circular features will automatically receive appropriate amplification compensation, simplifying the parameter adjustment process.
[0009] In one embodiment, the arc feature includes a radius of curvature and a central angle. Determining the target correction coefficient based on the arc feature includes: performing a reciprocal transformation on the radius of curvature to obtain a curvature value; normalizing the central angle to obtain an angle ratio value; weighting and fusing the curvature value and the angle ratio value according to the current processing state parameters to obtain a fused feature value; and inputting the fused feature value into a preset nonlinear activation function to obtain the target correction coefficient, wherein the preset nonlinear activation function is monotonically increasing and its output value range is not less than 1.
[0010] In this embodiment, the curvature value and angle ratio value are fused into a continuous feature value. This feature value is then mapped using a monotonically increasing nonlinear activation function with an output value range not less than 1 to obtain a smooth target correction coefficient. Compared to piecewise threshold judgment, this method avoids abrupt changes in the correction coefficient at the threshold, ensuring a continuous transition in offset between adjacent processing segments and preventing tool marks. The saturation characteristic of the nonlinear activation function prevents the correction coefficient from increasing infinitely at extremely small arcs, thus ensuring process stability.
[0011] In one embodiment, the processing state parameters include the real-time feed rate of the cutting wire, and the weighted fusion of the curvature value and the angle ratio value based on the current processing state parameters includes: A first weighting coefficient and a second weighting coefficient are determined based on the real-time feed rate, wherein the first weighting coefficient and the second weighting coefficient are respectively the weighting coefficients corresponding to the curvature value and the angle ratio value, the first weighting coefficient is positively correlated with the feed rate, and the second weighting coefficient is negatively correlated with the feed rate; The curvature value and the angle ratio value are weighted and fused according to the first weighting coefficient and the second weighting coefficient.
[0012] In this embodiment, the discharge concentration effect is more severe during high-speed feed, and the increased weight of the curvature value makes the correction coefficient more sensitive to radius changes; during low-speed feed, the arc length effect becomes more prominent, and the increased weight of the angle ratio value can more accurately compensate for heat accumulation. This dynamic weight allocation allows the same arc to obtain optimal compensation at different feed speeds, avoiding insufficient compensation at excessively high speeds or excessive compensation at excessively low speeds due to fixed weights, thus improving process robustness.
[0013] In one embodiment, the processing state parameters further include the loss coefficient of the cutting metal wire, and the weighted fusion of the curvature value and the angle ratio value based on the current processing state parameters further includes: The first weighting coefficient is corrected based on the loss coefficient to obtain the third weighting coefficient, wherein the larger the loss coefficient of the cutting metal wire, the greater the increase of the third weighting coefficient relative to the first weighting coefficient; The second weight coefficient is corrected based on the third weight coefficient to obtain the fourth weight coefficient, wherein the sum of the third weight coefficient and the fourth weight coefficient is equal to 1; The curvature value and the angle ratio value are weighted and fused according to the third weighting coefficient and the fourth weighting coefficient.
[0014] In this embodiment, the diameter and stiffness of the cutting wire decrease after wear, making it more prone to vibration and secondary discharge due to the discharge reaction force at curved sections, thus exacerbating overcutting. A larger wear coefficient leads to a greater increase in the curvature value weight, making the correction coefficient more sensitive to the degree of bending and compensating for the additional overcutting risk caused by wire wear. Normalization of the weight sum ensures dimensional stability of the fused feature value. This feature couples wear degree with bending compensation, extending the effective service life of the cutting wire and maintaining consistent processing accuracy throughout the entire wire diameter variation cycle.
[0015] In one embodiment, the processing state parameters further include a discharge frequency, and the weighted fusion of the curvature value and the angle ratio value based on the current processing state parameters further includes: The third weighting coefficient is corrected according to the discharge frequency to obtain the fifth weighting coefficient, wherein the higher the discharge frequency, the greater the increase of the fifth weighting coefficient relative to the third weighting coefficient; The fourth weight coefficient is corrected based on the fifth weight coefficient to obtain the sixth weight coefficient, wherein the sum of the fifth weight coefficient and the sixth weight coefficient is equal to 1; The curvature value and the angle ratio value are weighted and fused according to the fifth weighting coefficient and the sixth weighting coefficient.
[0016] In this embodiment, the higher the discharge frequency (the smaller the pulse interval), the greater the input energy per unit time, leading to more severe heat accumulation and secondary discharge at the arc, and a significant increase in the risk of overcutting. Therefore, increasing the curvature value weight further amplifies the correction coefficient, fully compensating for the additional gap increment caused by high-frequency discharge. This feature enables the compensation model to respond to real-time changes in discharge energy, thus adapting to processing scenarios with different power supply parameter settings. For roughing using high frequencies, bending compensation is automatically increased; for finishing using low frequencies, compensation automatically decreases, achieving an intelligent balance between processing efficiency and accuracy.
[0017] In one implementation, before inputting the fused feature values into a preset nonlinear activation function to obtain the target correction coefficient, the method further includes: Obtain the contour deviation amount, which is used to characterize the deviation between the actual contour and the theoretical contour of the machined workpiece at the arc feature position; Based on the contour deviation, the target parameters of the preset nonlinear activation function are iteratively corrected, wherein the target parameters include the slope and / or saturation threshold of the preset nonlinear activation function.
[0018] In this embodiment, different batches of materials and machine tools with different wear conditions exhibit variations, meaning that the activation function parameters calibrated in a single step may not be optimal. By measuring the arc deviation after actual cutting and adjusting the slope or saturation threshold of the activation function in reverse, the model parameters can gradually approach the ideal value, achieving optimization while machining. This self-learning capability significantly shortens the process debugging cycle for new materials and can automatically compensate for machine tool aging and environmental impacts during long-term operation, ensuring continuous high-precision machining.
[0019] In one embodiment, before weighted fusing the curvature value and the angle ratio value according to the current processing state parameters to obtain the fused feature value, the method further includes: acquiring a real-time image of the cutting metal wire; determining the real-time diameter of the cutting metal wire based on the real-time image; and determining the loss coefficient of the cutting metal wire based on the real-time diameter and the initial diameter of the cutting metal wire.
[0020] In this embodiment, traditional manual measurements suffer from long intervals and large subjective errors, making it impossible to reflect changes in wire diameter in a timely manner. By utilizing a vision system to automatically take pictures, detect edges, and calculate diameters during processing gaps, non-contact, high-frequency updates of the loss coefficient are achieved. The real-time and accurate loss coefficient provides a reliable data foundation for weight correction, avoiding compensation mismatch caused by lag in wire diameter data.
[0021] In one embodiment, determining the geometric features of the workpiece to be processed corresponding to the current processing position includes: acquiring processing path parameters of the workpiece to be processed; determining the interpolation type of the current processing position based on the processing path parameters, wherein the interpolation type is used to characterize whether the trajectory of the current processing position is a straight line feature or a circular arc feature; and determining the geometric features of the workpiece to be processed corresponding to the current processing position based on the interpolation type.
[0022] In this embodiment, feature recognition is achieved directly using existing G-code instructions in the CNC system, eliminating the need for additional sensors or complex algorithms, thus enabling zero-latency and zero-cost geometric classification. This analysis process can be performed before or in real-time, without affecting processing efficiency, and is 100% accurate. Compared to methods relying on vision-based processing path recognition, this solution avoids the environmental sensitivity and computational latency of the vision system, making it more suitable for high-speed continuous processing scenarios and providing a low-complexity, high-reliability front-end foundation for the entire dynamic compensation scheme.
[0023] Secondly, this application provides a wire EDM processing device, including a processor and a memory, wherein the memory is used to store a computer program, and the computer program, when executed by the processor, implements the above-described wire EDM processing method.
[0024] This wire cutting equipment, by employing the aforementioned wire cutting method, can significantly improve the wire cutting accuracy and reliability of precision structural components such as mobile phone camera bezels, fingerprint buttons, and card trays. Attached Figure Description
[0025] Figure 1 This is a schematic flowchart illustrating the wire EDM processing method according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a wire EDM processing equipment shown in an embodiment of this application; Figure 3 This is another structural schematic diagram of the wire cutting processing equipment shown in the embodiments of this application. Detailed Implementation
[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.
[0027] In traditional wire EDM, the cutting offset of the cutting wire is usually set to a fixed value. However, when machining arc segments, the discharge gap of the cutting wire is not constant, but increases as the radius of curvature of the arc decreases, and the size of the central angle also affects the degree of discharge concentration. This application constructs a target correction coefficient related to the radius of curvature and the central angle, which enables dynamic adjustment of the discharge gap of the cutting wire, thereby obtaining a more accurate cutting offset under different geometric features.
[0028] Please refer to Figure 1 , Figure 1 This is a schematic flowchart illustrating a wire EDM method provided in an embodiment of this application. Figure 1 As shown, the wire cutting method of this embodiment includes steps S11 to S14, which are described in detail below: Step S11: Obtain the current processing position of the workpiece to be processed, and determine the geometric features of the workpiece to be processed corresponding to the current processing position. The geometric features include straight line features and circular arc features.
[0029] In this step, as the processing time of the workpiece increases, the processing position of the workpiece changes continuously, and the geometric features at the processing position also change continuously.
[0030] The machining path parameters of the workpiece to be processed consist of continuous program segments. The interpolation type of each program segment determines the trajectory of the cutting wire, i.e., the geometric features of the workpiece corresponding to the current machining position. There are significant differences in the discharge characteristics of the cutting wire corresponding to straight-line features and circular-arc features. With straight-line features, the discharge gap of the cutting wire is stable; while with circular-arc features, the discharge of the cutting wire is concentrated, and the gap increases. Early identification of the geometric features of the workpiece corresponding to the current machining position lays the foundation for subsequent differentiated discharge gap compensation of the cutting wire. Through code-level parsing, zero-latency geometric feature recognition can be achieved, avoiding the delays and errors of peripherals such as vision sensors.
[0031] CNC machining code, or G-code, is a standardized language for describing tool movements. G01 represents linear interpolation, and G02 / G03 represent circular interpolation. These parameters directly reflect the geometric characteristics of the current program segment, achieving zero-delay recognition without the need for external sensors.
[0032] In some embodiments, when determining the geometric features of the workpiece to be processed corresponding to the current processing position, the processing path parameters of the workpiece to be processed are first obtained; then, based on the processing path parameters, the interpolation type of the current processing position is determined, whereby the interpolation type is used to characterize whether the trajectory of the current processing position is a straight line feature or a circular arc feature; finally, the geometric features of the workpiece to be processed corresponding to the current processing position are determined based on the interpolation type.
[0033] For example, before machining begins, the CAD model of the workpiece is imported into the CNC system, which automatically generates ISO standard G-code machining path parameters. During machining, the CNC system obtains the coordinates of the current machining position by reading the currently executing program segment in real time. Simultaneously, the CNC system analyzes the interpolation type in the program segment; if the interpolation type is G01, the trajectory of the current machining position is determined to be a straight line; if the interpolation type is G02 or G03, the trajectory of the current machining position is determined to be an arc. For arc features, the system further extracts its radius of curvature and central angle. For example, machining a 90° arc with a radius of 2mm, the system records the radius of curvature as 2mm and the central angle as π / 2 radians. Since determining the geometric features of the current machining position requires no additional sensors and is entirely based on code analysis, reliability is high.
[0034] Step S12: Determine the target discharge gap corresponding to the cutting wire based on the geometric features. The target discharge gap is used to characterize the shortest distance between the workpiece etching surface and the outer surface of the cutting wire.
[0035] In this step, straight and circular arc features correspond to different discharge gaps. During straight segment machining, the discharge gap remains relatively constant, and the reference value can be accurately calibrated through experimental cutting. However, due to the concentrated electric field and numerous secondary discharges in the circular arc segment, the actual discharge gap is 10%-50% larger than that of the straight segment. The smaller the radius of the circular arc feature, the more severe the accumulation of discharge ions at the corners. Therefore, a correction coefficient greater than 1 needs to be introduced to increase the target discharge gap, thereby increasing the offset and compensating for the overcutting tendency.
[0036] In some embodiments, during step S12, if the geometric feature of the workpiece to be processed corresponding to the current processing position is a straight line, the target discharge gap is set to a preset reference discharge gap. The reference discharge gap can be obtained by measuring the groove width after experimentally cutting a straight section of the workpiece. If the geometric feature of the workpiece to be processed corresponding to the current processing position is an arc, a target correction coefficient is determined based on the arc feature, and the reference discharge gap is corrected based on the target correction coefficient to obtain the target discharge gap. By distinguishing between straight and arc features, and correcting the reference discharge gap when the geometric feature is an arc, the overcutting caused by underestimation of the discharge gap corresponding to the arc feature in the traditional method is avoided. At the same time, standard compensation is maintained at the straight feature, balancing efficiency and accuracy.
[0037] Furthermore, the reference discharge gap = (actual groove width - diameter of the cutting wire) / 2, and the target discharge gap = reference discharge gap × target correction factor. For example, if the reference discharge gap is 0.02 mm and the target correction factor is 1.3, then the target discharge gap is 0.026 mm.
[0038] In this step, the arc features include the radius of curvature and the central angle. The radius of curvature directly reflects the degree of arc bending; the more severe the bending, the stronger the discharge concentration effect, and the larger the required target correction coefficient. The central angle affects the arc length and discharge duration; the larger the central angle, the more significant the heat accumulation, and the larger the required target correction coefficient. However, the importance of the radius of curvature and the central angle varies with the processing conditions. For example, under different processing conditions such as speed and wire loss, the importance of the radius of curvature and the central angle is different; therefore, weighted fusion is introduced. A preset nonlinear activation function ensures that the output is ≥1 and gradually saturates, avoiding extreme values.
[0039] In some embodiments, when determining the target correction coefficient based on the arc characteristics, the radius of curvature is first transformed inversely to obtain a curvature value, which can characterize the degree of curvature at the current processing position; then, the central angle is normalized to obtain an angle ratio value, which can characterize the arc length ratio; next, the curvature value and the angle ratio value are weighted and fused according to the current processing state parameters to obtain a fused feature value; finally, the fused feature value is input into a preset nonlinear activation function to obtain the target correction coefficient, wherein the preset nonlinear activation function is monotonically increasing and the output value range is not less than 1.
[0040] For example, if the radius of curvature R = 0.5 mm, then the curvature value C = 1 / 0.5 = 2.0; if the central angle α = 90° = π / 2, then the normalized angle ratio value β = α / π = 0.5. First, the first and second weighting coefficients are obtained based on the current processing state parameters. Then, the curvature value and the angle ratio value are weighted and fused according to the first and second weighting coefficients to obtain the fused feature value. If the first weighting coefficient w1 = 0.7 and the second weighting coefficient w2 = 0.3, the fused feature value S = w1 × C + w2 × β = 1.55. A preset nonlinear activation function is used, and the target correction coefficient K = =1 + 0.915 = 1.915. If the fused eigenvalue S is small, the target correction coefficient K approaches 1; if the fused eigenvalue S is large, the target correction coefficient K approaches 2. In practice, the parameters of the preset nonlinear activation function can be adjusted according to material experiments.
[0041] In this step, the slope of the nonlinear activation function determines the response sensitivity, and the saturation threshold determines the maximum correction. Different materials and equipment exhibit varying dynamic characteristics, so the parameters of the nonlinear activation function calibrated in one step may not be optimal. By measuring the arc deviation of the processed workpiece and adjusting the target of the nonlinear activation function in reverse, the model can continuously approach the ideal state, forming a closed-loop self-optimization.
[0042] In some embodiments, before inputting the fused feature values into a preset nonlinear activation function to obtain the target correction coefficient, the contour deviation can be obtained first. The contour deviation is used to characterize the deviation between the actual contour and the theoretical contour of the processed workpiece at the arc feature position. Based on the contour deviation, the target parameters of the preset nonlinear activation function are iteratively corrected, wherein the target parameters include the slope and / or saturation threshold of the preset nonlinear activation function.
[0043] For example, the preset nonlinear activation function is expressed as: target correction coefficient After machining the first piece, measure the deviation Δ between the actual and theoretical dimensions at the arc (overcutting is positive). If Δ > 0, it indicates that the target correction coefficient K is too large, and b or a should be decreased; if Δ < 0, it indicates that the target correction coefficient K is too small, and b or a should be increased. The activation function parameters are iteratively optimized using the gradient descent method. For example, for a nonlinear activation function... The update rule for parameter b is as follows: After multiple iterations, b and a converge to their optimal values. , The learning rate is a preset small positive number (e.g., 0.001). This represents the measured deviation (overshear is positive, undershear is negative). When Δ>0, If the value is positive, b decreases, thus slightly reducing the target correction coefficient K; conversely, b increases. After multiple iterations, b converges to the optimal value.
[0044] In this step, the faster the feed speed of the cutting wire, the longer the arc length traversed per unit time, the more intense the discharge concentration effect, and the curvature value has a dominant influence on overcutting, so the weight of the curvature value should be increased; conversely, the slower the feed speed of the cutting wire, the more significant the influence of the central angle, so the weights need to be balanced.
[0045] In some embodiments, the processing state parameters include the real-time feed rate of the cutting wire. When weighting and fusing the curvature value and the angle ratio value according to the current processing state parameters, a first weighting coefficient and a second weighting coefficient are first determined according to the real-time feed rate. The first weighting coefficient and the second weighting coefficient are the weighting coefficients corresponding to the curvature value and the angle ratio value, respectively. The first weighting coefficient is positively correlated with the feed rate, and the second weighting coefficient is negatively correlated with the feed rate. Then, the curvature value and the angle ratio value are weighted and fused according to the first weighting coefficient and the second weighting coefficient.
[0046] For example, set the base speed for cutting the metal wire V0 = 100 mm / min, the real-time feed speed is V, and the first weighting coefficient w1 = The first weighting coefficient w1 ∈ [0,1] increases with the real-time feed rate V. The second weighting coefficient w2 = 1 - w1. If the real-time feed rate V = 50 mm / min, the first weighting coefficient w1 ≈ 0.33 and the second weighting coefficient w2 ≈ 0.67; if the real-time feed rate V = 200 mm / min, the first weighting coefficient w1 ≈ 0.88 and the second weighting coefficient w2 ≈ 0.12, and the fused feature value S = w1 × C + w2 × β. The target correction coefficient K is continuously differentiable with respect to the fused feature value S, avoiding the jumps caused by the segmented threshold.
[0047] In this step, the diameter and stiffness of the cutting wire decrease after wear, making it more prone to jitter and secondary discharge due to the discharge reaction force at the arc, thus exacerbating overcutting. Therefore, the greater the wear of the cutting wire, the greater the weight of the curvature value should be, while the weight of the angle ratio value should be reduced accordingly. Normalizing the sum of the weights ensures the dimensional stability of the fused feature values.
[0048] In some embodiments, the processing state parameters also include the loss coefficient of the cutting metal wire. When weighting and fusing the curvature value and the angle ratio value according to the current processing state parameters, the first weight coefficient can be corrected according to the loss coefficient to obtain a third weight coefficient. The larger the cutting metal wire loss coefficient, the greater the increase of the third weight coefficient relative to the first weight coefficient. The second weight coefficient is corrected according to the third weight coefficient to obtain a fourth weight coefficient. The sum of the third weight coefficient and the fourth weight coefficient is equal to 1. Finally, the curvature value and the angle ratio value are weighted and fused according to the third weight coefficient and the fourth weight coefficient.
[0049] For example, the initial diameter of the cutting wire. =0.18mm, current diameter d=0.176mm, loss coefficient =0.0222. Let the first weighting coefficient w1 = 0.7, then the third weighting coefficient w3 = w1 × (1 + η) = 0.7 × 1.0222 = 0.7155, clamping it to no more than 1. The fourth weighting coefficient w4 = 1 - w3 = 0.2845. The fused eigenvalue S = w3 × C + w4 × β. The larger the loss coefficient, the greater the correction magnitude; for example, when the loss coefficient η = 0.1, the third weighting coefficient w3 = 0.77.
[0050] In this step, the cutting wire continuously wears down during processing, gradually reducing its diameter. Traditional manual measurements are spaced out and cannot reflect the wear in real time. Using a vision system, images can be taken during processing gaps or outside the wire, and the diameter can be quickly obtained through image processing, providing an accurate wear coefficient.
[0051] In some embodiments, before obtaining the fused feature value by weighted fusion of curvature value and angle ratio value based on the current processing state parameters, a real-time image of the cutting wire can be obtained first, and the real-time diameter of the cutting wire can be determined based on the real-time image of the cutting wire; then, the loss coefficient of the cutting wire can be determined based on the real-time diameter of the cutting wire and the initial diameter of the cutting wire.
[0052] For example, a high-resolution CCD camera and backlight are installed at appropriate locations on the cutting equipment. After each workpiece is processed, the cutting wire is stopped, triggering the camera to capture a real-time image. Edge detection algorithms are used to extract the two edges of the cutting wire, the pixel distance is calculated, and this distance is calibrated and converted into the current diameter *d* of the cutting wire. The loss coefficient of the cutting wire is also considered. ,in This is the initial diameter. The frequency of measuring the real-time diameter of the cutting wire can be adjusted according to the processing load, for example, once every 10 pieces.
[0053] In this step, the higher the discharge frequency of the cutting wire and the smaller the pulse interval, the greater the energy input per unit time, leading to more severe heat accumulation and secondary discharge at the arc, and increasing the risk of overcutting. Therefore, a larger curvature value weight is required. This correction, combined with speed and wire loss corrections, forms a multi-factor adaptive mechanism.
[0054] In some embodiments, the processing state parameters also include the discharge frequency. When weighting and fusing the curvature value and the angle ratio value according to the current processing state parameters, the third weighting coefficient is first corrected according to the discharge frequency to obtain the fifth weighting coefficient. The higher the discharge frequency, the greater the increase of the fifth weighting coefficient relative to the third weighting coefficient. Then, the fourth weighting coefficient is corrected according to the fifth weighting coefficient to obtain the sixth weighting coefficient. The sum of the fifth weighting coefficient and the sixth weighting coefficient is equal to 1. Finally, the curvature value and the angle ratio value are weighted and fused according to the fifth weighting coefficient and the sixth weighting coefficient.
[0055] For example, setting the reference frequency for the discharge of the cutting metal wire to f0 = 50kHz, the current frequency to f = 70kHz, and the frequency factor γ = f / f0 = 1.4. With a preset correction intensity coefficient q = 0.2, the fifth weighting coefficient w5 = w3 × (1 + q × γ) = 0.7155 × (1 + 0.28) = 0.7155 × 1.28 = 0.9158. Furthermore, the sixth weighting coefficient w6 = 1 - 0.9158 = 0.0842. The fusion value S = w5 × C + w6 × β. It can be seen that at high frequencies, the curvature value has an absolutely dominant weight, fully compensating for the discharge concentration effect.
[0056] Step S13: Based on the target discharge gap and the cross-sectional radius of the cutting wire, obtain the current cutting offset of the cutting wire, wherein the current cutting offset is equal to the sum of the target discharge gap and the cross-sectional radius.
[0057] In this step, during wire EDM, the actual kerf width is equal to the cross-sectional diameter of the cutting wire plus twice the target discharge gap. To ensure the workpiece contour matches the design, the current cutting offset is equal to the sum of the target discharge gap and the cross-sectional radius. This formula is common knowledge in the field of wire EDM, but this embodiment compensates for discharge changes caused by the curvature of the cutting path by dynamically adjusting the current cutting offset. The current cutting offset represents the distance the centerline of the cutting wire must offset outward (for external cutting) or inward (for internal cutting) to obtain the theoretical contour. The CNC system stores the current cutting offset in the tool compensation register and automatically applies it when generating the machining trajectory.
[0058] Step S14: Control the cutting wire to perform cutting processing according to the current cutting offset.
[0059] In this step, the cutting equipment's tool compensation function automatically adjusts the movement of the cutting wire based on the current cutting offset and machining direction. This step dynamically calculates the current cutting offset and injects it into the compensation mechanism in real time, ensuring the cutting wire always moves to the correct position required by the theoretical contour. The cutting wire can be, for example, molybdenum wire or copper wire.
[0060] The cutting equipment writes the calculated current cutting offset into the offset register and activates the tool radius compensation function. During machining, the programmed trajectory is automatically offset inward or outward by the current cutting offset according to the current path direction, generating actual control point coordinates. The servo system drives the XY table to move according to the offset trajectory, and the cutting wire performs electrical discharge cutting along the compensated path. For example, for counterclockwise machining of the outer contour, G41 left compensation is used, and the actual trajectory is offset outward from the contour by the current cutting offset. Simultaneously, the system continuously monitors whether the current program segment has ended. If it has, it returns to read the next program segment and repeats the above process, achieving dynamic compensation for the entire path. This process is fully automated, requiring no manual intervention, and achieves real-time adaptive control of the entire path. It eliminates the need for pauses or manual adjustments, significantly improving the consistency and automation of batch processing.
[0061] This wire EDM method incorporates the geometric features of the machining path into the dynamic calculation of the cutting offset, avoiding the overcutting of the arc segment due to concentrated discharge by treating the discharge gap of the cutting wire as a constant. In this method, the current machining position is first acquired in real time and its geometric features are identified: if it is a straight line, a preset reference discharge gap is used; if it is an arc, a larger target discharge gap is obtained through a correction process. Then, the target discharge gap is summed with the cross-sectional radius of the cutting wire to obtain the current cutting offset. This current cutting offset determines the position of the cutting wire's centerline relative to the theoretical contour. Finally, the CNC system controls the cutting wire to perform cutting according to this current cutting offset. This method uses engineering modeling to model the physical coupling between the path's geometric features and the discharge gap, realizing the transformation from a globally constant offset to a geometrically adaptive offset. By adopting this wire EDM method, the machining accuracy of the workpiece's arc features is significantly improved, and the overcutting is significantly reduced. Secondly, the compensation for straight and arc features is independent, avoiding excessive correction of straight features at the expense of efficiency. Furthermore, this method is entirely based on software algorithms, requiring no hardware modification and resulting in low deployment costs. For precision parts with minute arc features, such as mobile phone camera bezels and fingerprint buttons, this method can stably control the contour error within ±0.003mm.
[0062] Figure 2 This is a schematic diagram of a wire EDM (Electrical Discharge Machining) equipment provided in one embodiment of this application. Figure 2 As shown, the wire cutting processing equipment 20 of this embodiment includes: at least one processor 21 ( Figure 2 (Only one is shown in the diagram), memory 22, and computer program 23 stored in the memory 22 and executable on the at least one processor 21, wherein the processor 21 executes the computer program 23 to implement the steps in any of the above method embodiments.
[0063] The wire EDM processing equipment 20 may include, but is not limited to, a processor 21 and a memory 22. Those skilled in the art will understand that... Figure 2 This is merely an example of wire EDM processing equipment 20 and does not constitute a limitation on the wire EDM processing equipment 20. It may include more or fewer components than shown, or combine certain components, or different components, such as input / output devices, network access devices, etc.
[0064] The processor 21 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.
[0065] In some embodiments, the memory 22 may be an internal storage unit of the wire EDM processing equipment 20, such as a hard disk or memory of the wire EDM processing equipment 20. In other embodiments, the memory 22 may be an external storage device of the wire EDM processing equipment 20, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the wire EDM processing equipment 20. Furthermore, the memory 22 may include both internal storage units and external storage devices of the wire EDM processing equipment 20. The memory 22 is used to store operating systems, applications, boot loaders, data, and other programs, such as the program code of computer programs. The memory 22 can also be used to temporarily store data that has been output or will be output.
[0066] In addition, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in any of the above method embodiments.
[0067] This application provides a computer program product that, when run on a computer device, enables the computer device to execute the steps described in the various method embodiments above.
[0068] Figure 3 This is another structural schematic diagram of a wire EDM processing equipment provided in one embodiment of this application. (See attached diagram.) Figure 3 As shown, the wire EDM processing equipment of this embodiment includes a wire routing system, a processing tank, an XY axis mechanism, a processing fluid treatment device, a filter, a display box, and a control unit cabinet.
[0069] The wire routing system includes a wire spool, guide rollers, a tension adjustment mechanism, and a cutting wire. The wire routing system supports, guides, and drives the cutting wire to reciprocate at a certain speed and tension. In this application, the cross-sectional radius of the cutting wire directly affects the calculation of the current cutting offset; simultaneously, the cutting wire loss coefficient fed back by the tension adjustment mechanism can be used to correct the weighted fusion weight of the curvature value and the angle ratio value, thereby improving the compensation accuracy of the arc feature.
[0070] The processing tank is used to hold the working fluid, such as deionized water or emulsion, which immerses or washes the cutting area during processing, serving to cool, remove chips, and stabilize the discharge. The structural stability of the processing tank ensures the relative positional accuracy between the workpiece and the cutting wire, enabling the reliable execution of determining the target discharge gap and current cutting offset based on geometric features in the embodiments of this application.
[0071] The XY-axis mechanism is a cross-shaped worktable driven by a servo motor, used to achieve precise movement of the workpiece in the horizontal plane. This mechanism performs the step of controlling the cutting wire to perform cutting processing according to the current cutting offset in the embodiments of this application, that is, by moving the workpiece, the cutting wire moves along the path after the theoretical contour is offset, thereby obtaining a high-precision processing contour.
[0072] The processing fluid treatment device includes a circulating pump, a filter, and an ion exchanger, used to maintain the cleanliness, resistivity, and temperature of the working fluid. Stable working fluid performance ensures consistent discharge gaps. In the embodiments of this application, both the reference discharge gap and the target discharge gap rely on a stable processing environment provided by the processing fluid treatment device, thereby ensuring the effectiveness of the compensation model under different processing conditions.
[0073] Filters are typically internal pressure or wound filter elements used to remove electrolytic corrosion products and impurities from the working fluid. The filter pressure and replacement frequency affect the flowability and insulation of the working fluid, thus influencing the actual value of the discharge gap. In this embodiment, the execution of iterative correction of the nonlinear activation function parameters indirectly depends on the working fluid quality maintained by the filter, in order to avoid abnormal discharge deviations caused by impurities.
[0074] The monitor box is equipped with a touch-screen LCD display for real-time display of processing status, parameter settings, trajectory drawing, and other information. Operators can use the monitor box to adjust the reference discharge gap, observe the contour deviation of the processed workpiece, and manually intervene in the slope or saturation threshold of the activation function, thereby achieving human-machine collaborative process optimization.
[0075] The control unit cabinet houses the CNC system, processing power supply, and motion controller, serving as the core hardware carrier for executing each step of the wire cutting method according to the embodiments of this application. The processor in the control unit cabinet runs computer programs to complete geometric feature recognition, inverse curvature transformation, central angle normalization, weighted fusion, nonlinear activation mapping, and real-time calculation and output of cutting offset.
[0076] Through the coordinated operation of the aforementioned components, the wire EDM processing equipment provided in this application can reliably achieve dynamic discharge gap compensation based on geometric features, significantly improving the processing accuracy and consistency of small arc features.
[0077] In the several embodiments provided in this application, it will be understood that each block in the flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the figures. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved.
[0078] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0079] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application for those skilled in the art.
Claims
1. A wire EDM machining method, characterized in that, include: Obtain the current processing position of the workpiece to be processed, and determine the geometric features of the workpiece to be processed corresponding to the current processing position, wherein the geometric features include straight line features and circular arc features; The target discharge gap corresponding to the cutting wire is determined based on the geometric features. The target discharge gap is used to characterize the shortest distance between the workpiece etching surface and the outer surface of the cutting wire. Based on the target discharge gap and the cross-sectional radius of the cutting wire, the current cutting offset of the cutting wire is obtained, wherein the current cutting offset is equal to the sum of the target discharge gap and the cross-sectional radius; The cutting wire is controlled to perform cutting processing based on the current cutting offset.
2. The wire EDM method as described in claim 1, characterized in that, Determining the target discharge gap corresponding to the cutting metal wire based on the geometric features includes: If the geometric feature is a straight line feature, then the target discharge gap is set as a preset reference discharge gap; If the geometric feature is a circular arc feature, then a target correction coefficient is determined based on the circular arc feature, and the reference discharge gap is corrected based on the target correction coefficient to obtain the target discharge gap.
3. The wire EDM processing method as described in claim 2, characterized in that, The arc feature includes the radius of curvature and the central angle, and the determination of the target correction coefficient based on the arc feature includes: The curvature value is obtained by performing a reciprocal transformation on the radius of curvature. The central angle is normalized to obtain the angle ratio value; The curvature value and the angle ratio value are weighted and fused according to the current processing status parameters to obtain a fused feature value; The fused feature values are input into a preset nonlinear activation function to obtain the target correction coefficient. The preset nonlinear activation function is monotonically increasing and its output value range is not less than 1.
4. The wire EDM method as described in claim 3, characterized in that, The processing status parameters include the real-time feed speed of the cutting wire, and the weighted fusion of the curvature value and the angle ratio value based on the current processing status parameters includes: A first weighting coefficient and a second weighting coefficient are determined based on the real-time feed rate, wherein the first weighting coefficient and the second weighting coefficient are respectively the weighting coefficients corresponding to the curvature value and the angle ratio value, the first weighting coefficient is positively correlated with the feed rate, and the second weighting coefficient is negatively correlated with the feed rate; The curvature value and the angle ratio value are weighted and fused according to the first weighting coefficient and the second weighting coefficient.
5. The wire EDM method as described in claim 4, characterized in that, The processing status parameters also include the loss coefficient of the cutting metal wire, and the weighted fusion of the curvature value and the angle ratio value based on the current processing status parameters further includes: The first weighting coefficient is corrected based on the loss coefficient to obtain the third weighting coefficient, wherein the larger the loss coefficient of the cutting metal wire, the greater the increase of the third weighting coefficient relative to the first weighting coefficient; The second weight coefficient is corrected based on the third weight coefficient to obtain the fourth weight coefficient, wherein the sum of the third weight coefficient and the fourth weight coefficient is equal to 1; The curvature value and the angle ratio value are weighted and fused according to the third weighting coefficient and the fourth weighting coefficient.
6. The wire EDM machining method as described in claim 5, characterized in that, The processing state parameters also include the discharge frequency, and the weighted fusion of the curvature value and the angle ratio value based on the current processing state parameters further includes: The third weighting coefficient is corrected according to the discharge frequency to obtain the fifth weighting coefficient, wherein the higher the discharge frequency, the greater the increase of the fifth weighting coefficient relative to the third weighting coefficient; The fourth weight coefficient is corrected based on the fifth weight coefficient to obtain the sixth weight coefficient, wherein the sum of the fifth weight coefficient and the sixth weight coefficient is equal to 1; The curvature value and the angle ratio value are weighted and fused according to the fifth weighting coefficient and the sixth weighting coefficient.
7. The wire EDM method as described in claim 3, characterized in that, Before inputting the fused feature values into a preset nonlinear activation function to obtain the target correction coefficient, the method further includes: Obtain the contour deviation amount, which is used to characterize the deviation between the actual contour and the theoretical contour of the machined workpiece at the arc feature position; Based on the contour deviation, the target parameters of the preset nonlinear activation function are iteratively corrected, wherein the target parameters include the slope and / or saturation threshold of the preset nonlinear activation function.
8. The wire EDM method as described in claim 5, characterized in that, Before the step of weightedly fusing the curvature value and the angle ratio value according to the current processing state parameters to obtain the fused feature value, the method further includes: Acquire a real-time image of the cutting metal wire, and determine the real-time diameter of the cutting metal wire based on the real-time image; The loss coefficient of the cutting wire is determined based on the real-time diameter and the initial diameter of the cutting wire.
9. The wire EDM method as described in claim 1, characterized in that, The determination of the geometric features of the workpiece to be processed corresponding to the current processing position includes: Obtain the processing path parameters of the workpiece to be processed; Based on the processing path parameters, the interpolation type of the current processing position is determined, and the interpolation type is used to characterize whether the trajectory of the current processing position is a straight line or a circular arc. The geometric features of the workpiece to be processed corresponding to the current processing position are determined based on the interpolation type.
10. A wire cutting processing equipment, characterized in that, It includes a processor and a memory, the memory being used to store a computer program that, when executed by the processor, implements the wire EDM machining method as described in any one of claims 1 to 9.