A precision terminal bending forming angle control method

Abstract

The application provides a precision terminal bending forming angle control method, comprising: obtaining a terminal surface image, identifying the terminal surface image to obtain a plating layer thickness, obtaining a crushing critical pressure and an initial bearing capacity according to the plating layer thickness; obtaining a real-time clamping pressure of a current bending forming process, evaluating the real-time clamping pressure and the crushing critical pressure, identifying a local crushing tendency of the plating layer of the clamped area, and generating a first adjustment strategy; obtaining a bending root area image where a bending root mechanical transmission path is located, identifying a dislocation trace of a metal layer along the mechanical transmission path in the bending root area image, and obtaining a bending root slip tendency; if the maximum stress is less than the initial bearing capacity, it is determined that plastic flow does not occur in the bending root, and the bending forming is controlled through a target clamping pressure.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for controlling the bending angle of precision terminals. Background Technology

[0002] Precision terminals, as core components of electronic connectors, directly affect the reliability of electrical contacts and assembly yield due to their forming angle accuracy, playing a crucial role in high-reliability applications such as automotive electronics and communication equipment. Traditional bending processes typically use fixed clamping force parameters combined with mold constraints to control the angle. However, this preset parameter method is difficult to cope with variations in actual production factors such as terminal plating thickness fluctuations and substrate hardness differences, resulting in insufficient consistency in forming angles. A more critical issue is that existing processes lack dynamic response capabilities to the non-linear relationship between clamping force and forming quality. When process parameters deviate from the optimal range, they cannot be adjusted in time, leading to accumulated angle deviations between batches. During bending, the application of clamping force should reduce slippage at the root of the bend by constraining the metal flow in the clamped area, thereby stabilizing the forming angle. However, even slight differences in the thickness of the plating layer on the terminal surface can significantly alter this mechanical transmission path. When a thicker plating area encounters excessive clamping pressure, the plating layer undergoes localized crushing deformation. This crushing not only damages the protective function of the plating layer but, more seriously, alters the stress distribution at the interface between the clamped area and the bending area. The stress concentration points formed by the crushing zone can induce unexpected plastic flow in the substrate during bending, causing the clamping constraints originally intended to suppress slippage to become a cause of angular deviation. This contradiction manifests as a counterintuitive phenomenon on the production floor: when operators increase the clamping force to address the issue of excessive root slippage leading to a smaller angle in a particular batch of terminals, the next batch of terminals with slightly thicker plating exhibits the opposite deviation, with larger angles and indentation defects in the plating. The root cause lies in the overlap between the critical crushing point caused by plating thickness differences and the clamping force range required for slippage control, and this overlap dynamically changes with the plating state of each batch of terminals. How to identify the balance point between the current terminal's plating load-bearing capacity and slippage risk in real time during bending, and adjust the clamping force application strategy accordingly to avoid the simultaneous occurrence of both plating crushing and root slippage failure modes, becomes a key issue in ensuring the stability of the forming angle and the integrity of the plating. Summary of the Invention

[0003] This invention provides a method for controlling the bending angle of precision terminals, comprising: Acquire a terminal surface image, identify the terminal surface image to obtain the coating thickness, and obtain the critical crush pressure and initial load capacity based on the coating thickness; The real-time clamping pressure during the current bending and forming process is obtained, the real-time clamping pressure and the critical crushing pressure are evaluated, the local crushing tendency of the coating in the clamped area is identified, and a first adjustment strategy is generated. According to the first adjustment strategy, the real-time clamping pressure is reduced to obtain an updated clamping pressure. Under the updated clamping pressure, the metal flow at the junction of the clamped area and the bending area is tightened to form a mechanical transmission path at the root of the bend. Obtain an image of the bending root region where the mechanical transmission path of the bending root is located, identify the misalignment traces of the metal layer along the mechanical transmission path of the bending root in the image of the bending root region, and obtain the tendency of the bending root to slip. The tendency of the bending root to slip and the preset critical slip are evaluated, the tendency of the forming angle to become unstable is identified, and a second adjustment strategy of superimposing an anti-slip force on the updated clamping pressure is generated. The target clamping pressure is obtained by superimposing an anti-slip force on the updated clamping pressure according to the second adjustment strategy. Obtain the stress concentration distribution map under the target clamping pressure, extract the maximum stress in the stress concentration distribution map, evaluate the maximum stress and the initial bearing limit, if the maximum stress is less than the initial bearing limit, it is determined that no plastic flow has occurred at the root of the bend, and the bending is controlled by the target clamping pressure.

[0004] Preferably, the step of acquiring a terminal surface image, identifying the plating thickness from the terminal surface image, and obtaining the crushing critical pressure and initial load-bearing limit based on the plating thickness includes: An image of the area where the terminal is clamped is acquired to obtain an image of the terminal surface; Extract the coating area contour from the terminal surface image, and obtain the coating thickness distribution map and coating uniformity according to the gray level difference within the contour. The coating thickness distribution map is used to retrieve the coating pressure threshold value, and the uneven section threshold value is corrected according to the coating uniformity to obtain the crushing critical pressure. The hardness value of the substrate is obtained and superimposed with the pressure bearing value of the corresponding coating to obtain the initial upper limit of the load-bearing capacity.

[0005] Preferably, the step of obtaining the real-time clamping pressure during the current bending process, evaluating the real-time clamping pressure and the critical crushing pressure, identifying the local crushing tendency of the coating in the clamped area, and generating a first adjustment strategy includes: Collect real-time clamping pressure distribution at various locations within the clamped area; The pressure difference is obtained by comparing the real-time clamping pressure distribution with the critical crushing pressure point by point. When the pressure difference is positive, the overpressure zone is marked, and the overpressure rate is obtained according to the ratio of the overpressure zone to the sampling position. Based on the comparison between the overpressure rate and the preset tendency threshold, a first adjustment strategy is generated, which includes a pressure callback benchmark and an overpressure zone location index.

[0006] Preferably, the step of reducing the real-time clamping pressure according to the first adjustment strategy to obtain the updated clamping pressure includes: Based on the pressure pullback benchmark and overpressure zone location index in the first adjustment strategy, the output pressure of the pressure head actuator is adjusted downward according to the pressure pullback benchmark. The adjustment point corresponds to the overpressure zone location index; The pressure values ​​at each location in the clamped area are re-acquired to obtain the updated clamping pressure.

[0007] Preferably, the step of identifying the metal flow tightening at the junction of the clamped area and the bending area under the updated clamping pressure to form a mechanical transmission path at the root of the bend includes: Images of the boundary region are acquired under the updated clamping pressure. Extract the grayscale gradient distribution from the image of the boundary region, identify the texture direction shift, and obtain the direction of metal flow; The texture density along the direction of metal flow is statistically analyzed, and the spatial range of metal flow tightening is determined according to the texture density and a preset density threshold. Connectivity labeling is performed on the pixels within the space and extended along the direction of metal flow to the bend corner contour line to form the mechanical transmission path at the root of the bend.

[0008] Preferably, the step of acquiring an image of the bending root region where the mechanical transmission path at the bending root is located, identifying misalignment traces of the metal layer along the mechanical transmission path at the bending root in the image of the bending root region, and obtaining the bending root slippage tendency includes: Position the imaging field of view according to the mechanical transmission path at the root of the bend, and acquire images of the root area of ​​the bend; The image of the bending root region is sharpened, and a strip-shaped region of interest is extracted along the mechanical transmission path of the bending root. Identify fault trace connected regions within the strip-shaped region of interest where the gray-level gradient exceeds a preset trace gradient threshold. The direction of the trace and the amount of misalignment are determined based on the boundary points at both ends of the connected domain of the misalignment trace, and the directional slippage determination result and slippage amplitude constitute the slippage tendency at the root of the bend.

[0009] Preferably, the assessment of the slippage tendency at the root of the bend and the preset critical slippage, and the identification of the tendency for instability of the forming angle, includes: Based on the directional slip determination result and slip amplitude in the slip tendency at the root of the bend, the slip amplitude is compared with the preset critical slip. When the directional slip determination result is that it exists and the slip amplitude exceeds the preset critical slip, it is determined that there is a tendency for the forming angle to be unstable, and the difference between the slip amplitude and the preset critical slip is recorded as the amplitude exceeding the limit.

[0010] Preferably, the step of generating a second adjustment strategy that superimposes an anti-slip force on the updated clamping pressure, and obtaining the target clamping pressure by superimposing the anti-slip force on the updated clamping pressure according to the second adjustment strategy, includes: The anti-slip force increment is obtained by referring to the slip compensation comparison table based on the difference between the slip tendency at the root of the bend and the preset critical slip. Taking the tightening starting point of the mechanical transmission path at the root of the bend as the center, extract the distribution of application points around the tightening starting point; The incremental anti-slip force value is superimposed on the corresponding output position of the application site distribution. The added pressure is added to the updated clamping pressure according to the position to obtain the target clamping pressure.

[0011] Preferably, obtaining the stress concentration distribution map under the target clamping pressure, extracting the maximum stress in the stress concentration distribution map, and evaluating the maximum stress and the initial bearing capacity limit include: Under the clamping pressure of the target, a photoelastic fringe pattern corresponding to the stress distribution at the root of the bend is collected. Perform fringe order identification on the photoelastic fringe pattern, convert the stress value according to the fringe order value, and form the stress concentration distribution map; The stress values ​​in the stress concentration distribution map are traversed, and the peak value is taken as the maximum stress. The difference between the initial upper limit of bearing capacity and the maximum stress is calculated to obtain the bearing capacity margin.

[0012] Preferably, the step of determining that no plastic flow has occurred at the root of the bend if the maximum stress is less than the initial load limit, and controlling the bending shape by the target clamping pressure, includes: When the maximum stress is less than the initial bearing limit, the bending control unit locks the target clamping pressure; The bending control unit sends the target clamping pressure to the pressure head actuator; The pressure head actuator completes the bending and forming output of the current cycle according to the target clamping pressure, and obtains a bent terminal.

[0013] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a method for controlling the bending angle of precision terminals, solving the contradiction between excessive clamping pressure leading to plating collapse and insufficient pressure causing instability in the forming angle in traditional bending processes. This invention identifies the plating thickness by acquiring images of the terminal surface, assesses the real-time clamping pressure based on the critical collapse pressure, and generates a first adjustment strategy to reduce the clamping pressure to avoid localized plating collapse. Simultaneously, after pressure reduction, it identifies the metal flow tightening at the boundary between the clamped area and the bending area, constructs the mechanical transmission path at the bending root, and obtains the slippage tendency at the bending root by identifying the misalignment marks of the metal layer along this path. This generates a second adjustment strategy that superimposes an anti-slip force on the updated clamping pressure, ensuring that forming angle instability is suppressed while preventing plating collapse. Finally, after verifying that no plastic flow has occurred at the bending root using a stress concentration distribution map, the target clamping pressure is output. This achieves coordinated control of plating protection and forming accuracy, effectively improving the quality stability and angle control accuracy of precision terminal bending. Attached Figure Description

[0014] Figure 1 This is a flowchart of a precision terminal bending and forming angle control method according to the present invention.

[0015] Figure 2 This is a schematic diagram of a precision terminal bending forming angle control method according to the present invention.

[0016] Figure 3 This is another schematic diagram of a precision terminal bending forming angle control method according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figures 1-3 This embodiment of a precision terminal bending forming angle control method may specifically include: Step S101: Obtain a terminal surface image, identify the terminal surface image to obtain the coating thickness, and obtain the critical crushing pressure and initial bearing capacity based on the coating thickness.

[0019] An industrial camera is used to acquire images of the terminal clamping area, resulting in terminal surface images. These images are then processed for grayscale conversion and histogram equalization to highlight the grayscale difference between the reflective areas of the coating and the substrate boundary. The Canny edge detection algorithm is used to extract the contour lines of the coating area. Based on the grayscale differences of pixels within the contour lines, thick and thin coating sections are defined, resulting in a coating thickness distribution map and coating uniformity. According to the coating thickness distribution map, pre-stored coating pressure threshold values ​​are retrieved for both thick and thin coating sections to obtain the critical crushing pressure at each location within the clamped area. This critical crushing pressure represents the pressure threshold at which the coating transitions from elastic deformation to plastic crushing. Combined with the coating uniformity, the critical crushing pressure for uneven coating sections is adjusted downwards proportionally to the uniformity deviation, forming a critical crushing pressure distribution. The hardness of the terminal substrate is measured by a hardness tester to obtain the substrate hardness value. The substrate hardness value is then superimposed on the coating bearing pressure value in the crush critical pressure distribution position by position to obtain the upper limit of stress resisting plastic flow in the clamped area. The initial bearing capacity of the clamped area before bending is determined. The initial bearing capacity characterizes the upper limit of resistance to plastic deformation when not subjected to clamping pressure.

[0020] In the bending and forming process of automotive electronic connectors and communication equipment terminals, fluctuations in the plating thickness and differences in substrate hardness of precision terminals can significantly affect the consistency of the forming angle. The following specific implementation method describes the process of determining the load-bearing capacity of the plating in the clamped area before bending and forming. In one embodiment, an industrial camera installed above the bending station acquires images of the terminal clamped area. The industrial camera uses a CMOS sensor with a resolution of no less than 5 megapixels and is equipped with a ring-shaped shadowless light source to eliminate high-gloss reflections on the terminal surface, thereby obtaining an image of the terminal surface. The terminal surface image then enters the image preprocessing stage. First, grayscale processing is performed to convert the three-channel color image into a single-channel grayscale image. Then, histogram equalization is used to stretch the grayscale value distribution range to highlight the grayscale difference between the reflective area of ​​the plating and the boundary of the substrate.

[0021] Specifically, the coated area has a high reflectivity due to the tin or gold plating treatment on the surface, and appears as a brighter pixel segment in the grayscale image, while the exposed substrate area has a relatively dark grayscale value. After equalization, the grayscale difference between the two is amplified.

[0022] For example, the Canny edge detection algorithm is used to extract the contour of the equalized image. The Canny edge detection algorithm first suppresses image noise through Gaussian filtering, then calculates the gradient magnitude and direction of the pixels, refines the edges through non-maximum suppression, and finally uses double threshold detection and edge connection to obtain the complete contour line of the coating area. The gray value of the pixels inside the contour line reflects the thickness of the coating. Pixel segments with higher gray values ​​correspond to thicker coating segments, and pixel segments with lower gray values ​​correspond to thinner coating segments, thus obtaining a coating thickness distribution map. In the workstation calibration stage, calibration samples with known coating thickness specifications are taken, such as tin-plated copper sheets with coating thicknesses of 2 micrometers, 3 micrometers, 5 micrometers, and 8 micrometers. The surface images of the calibration samples are captured under the same light source and camera parameters as the actual acquisition. The gray mean of the coating area of ​​each calibration sample is extracted, and a linear fitting relationship h=a×g+b is established between the gray mean and the known coating thickness, where h is the coating thickness (micrometers), g is the gray mean, and a and b are fitting coefficients. When subsequently acquiring terminal surface images, the average grayscale value of each pixel segment is substituted into the fitting relationship to calculate the corresponding coating thickness value, thus forming a coating thickness distribution map in micrometers. This calibration method is suitable for terminal batches with a single coating material and uniform surface roughness. For batches with mixed materials or significant differences in surface conditions, recalibration is required. The coating uniformity is determined by the following formula: U=max(0,1-σ / μ), where U represents the coating uniformity, μ represents the average grayscale value of pixels within the contour line, σ represents the standard deviation of the grayscale values ​​of pixels within the contour line, and the max function takes the maximum value within the parentheses. When σ / μ is greater than 1, U automatically takes the value of 0, indicating extremely uneven coating. When σ / μ is less than 1, U is calculated as 1-σ / μ. The value of U ranges from 0 to 1. The closer the U value is to 1, the more uniform the coating thickness. During calculation, μ is required to be greater than 5 to ensure measurement validity. Areas below this value are considered invalid measurement areas and excluded from the analysis range.

[0023] It should be noted that the critical crushing pressure represents the critical pressure point at which the coating transitions from elastic deformation to plastic crushing. If the critical pressure is exceeded, indentation defects will appear on the coating.

[0024] In one possible implementation, the critical crushing pressure is corrected based on the coating uniformity. If the coating uniformity U at a certain location is lower than a preset uniformity threshold of 0.85, the location is determined to be a region with uneven coating. During the calibration phase, crushing tests are performed on tin-plated samples with different uniformity levels. When U drops below 0.85, the measured crushing pressure in the thin-coated area decreases by more than 15% compared to the nominal value, which affects the safety of subsequent clamping pressure settings. Therefore, 0.85 is used as the lower limit of uniformity for triggering reduction correction. The critical crushing pressure at the location is reduced according to the uniformity deviation. The reduced pressure value P' = P0 * U, where P0 represents the critical crushing pressure before the location correction, and P' represents the critical crushing pressure after the location correction, thereby forming the critical crushing pressure distribution across the entire clamped area. Engineering experience shows that a lower coating uniformity U indicates a greater thickness difference between thin and thick coating areas, resulting in a smaller effective load-bearing area in the thin coating area and a corresponding decrease in the local crushing critical pressure. In the absence of a precise laminated mechanical model, using U as a linear reduction coefficient is an engineering approximation of the reduced load-bearing capacity in the thin coating area. During the calibration phase, crushing tests were conducted on coating samples with different uniformity levels, verifying that the deviation between U×P0 and the measured crushing pressure was within 15%. Preferably, the terminal substrate is subjected to offline random sampling hardness measurement using a Leeb hardness tester or a Vickers hardness tester to obtain the substrate hardness value, which reflects the terminal copper alloy substrate's ability to resist indentation deformation. The hardness value of the substrate is converted into the yield stress of the substrate according to the hardness-yield strength conversion relationship in the mechanics of materials. For copper alloy substrates, the conversion relationship adopts an approximate conversion of yield stress equal to Vickers hardness value multiplied by 3.0. The converted substrate yield stress is then superimposed on the coating bearing pressure value at each position in the crushing critical pressure distribution. The superimposed stress value characterizes the upper limit of stress of the coating and substrate in resisting plastic deformation, thereby determining the initial bearing capacity of the clamped area before bending.

[0025] It is understood that the initial load-bearing capacity characterizes the upper limit of the plastic deformation resistance of the clamped area when it is not subjected to clamping pressure, providing a reference for the initial setting of clamping pressure in the bending forming process.

[0026] Step S102: Obtain the real-time clamping pressure of the current bending forming process, evaluate the real-time clamping pressure and the critical crushing pressure, identify the local crushing tendency of the coating in the clamped area, and generate the first adjustment strategy.

[0027] A piezoelectric pressure sensor installed at the clamping station of the bending die samples the current bending process in real time, obtaining the real-time clamping pressure distributed at various positions in the clamped area. The real-time clamping pressure reflects the loading state of each point on the clamping surface in the form of a time-series data stream. The time-series data stream is smoothed by a sliding window to filter out sampling noise interference, resulting in a stable real-time clamping pressure distribution. Based on the real-time clamping pressure distribution and the aforementioned critical crushing pressure distribution, the pressure difference between the real-time clamping pressure and the critical crushing pressure is calculated point by point for each position in the clamped area. If the pressure difference at a certain position is positive, the position is determined to have entered the overpressure zone. The ratio of the number of overpressure zone positions in the clamped area to the total number of positions is counted to obtain the overpressure rate. The overpressure rate characterizes the local crushing tendency of the coating in the clamped area. Based on the local crushing tendency of the coating, if the overpressure rate exceeds the preset tendency threshold, the excess amount of the real-time clamping pressure relative to the critical crushing pressure is calculated point by point for the overpressure zone location. The maximum value of the excess amount is taken as the pressure correction benchmark, and a first adjustment strategy for reducing the output pressure of the clamping head is obtained. The first adjustment strategy includes the pressure correction benchmark and the overpressure zone location index.

[0028] In the bending process of automotive electronic connectors and communication equipment terminals, based on the aforementioned critical pressure distribution for crushing in the clamped area, the following specific implementation describes the acquisition of real-time clamping pressure during bending, the identification of local crushing tendency of the plating layer, and the formation process of the first adjustment strategy. In one implementation, a piezoelectric pressure sensor array installed at the clamping station of the bending die samples the current bending process in real time. The piezoelectric pressure sensor array is uniformly arranged tangentially along the clamping surface, with the sensitive surface of each sensor unit facing the normal to the clamping surface. Each sensor unit covers a sub-region of the clamped area, the size of which matches the width of the terminal clamped area. The sampling frequency is not less than 1000 Hz, obtaining the real-time clamping pressure distributed at various positions within the clamped area. The real-time clamping pressure reflects the loading state of each point on the clamping surface in the form of a time-series data stream. The data stream is uploaded to the bending control unit via an industrial Ethernet for subsequent comparison.

[0029] Specifically, the piezoelectric pressure sensor operates based on the piezoelectric effect of quartz crystals. When the clamping head applies a load to the clamped area of ​​the terminal, a charge proportional to the pressure is generated on the surface of the piezoelectric crystal. This charge is converted into a voltage signal by a charge amplifier, and then converted into a pressure value by analog-to-digital conversion. The raw sampled data is subject to sampling noise due to factors such as clamping mechanism vibration and electromagnetic interference. Directly using unprocessed raw data can lead to false alarms in subsequent judgments. Therefore, a sliding window smoothing process is performed on the time-series data stream. The window width is set to 10 to 20 sampling points. The sampled values ​​within the window are averaged and used as the output value at the center of the window. The window slides point by point along the time axis, thereby obtaining a stable real-time clamping pressure distribution.

[0030] Further, based on the real-time clamping pressure distribution and the aforementioned critical crushing pressure distribution, a point-by-point comparison is performed for each position within the clamped area. Specifically, for position i within the clamped area, the real-time clamping pressure at that position is denoted as Pr(i), and the critical crushing pressure is denoted as Pc(i), with the pressure difference being ΔP(i) = Pr(i) - Pc(i). If ΔP(i) is positive, it indicates that the real-time loading pressure at that position has exceeded the critical point where the coating transitions from elastic deformation to plastic crushing, and the position is determined to have entered the overpressure zone; if ΔP(i) is negative or zero, the position is determined to be within the safe loading range. After point-by-point traversal, the set of overpressure zone positions is obtained.

[0031] In one possible implementation, the overpressure rate R = N1 / N is obtained by statistically analyzing the ratio of the number of overpressure locations N1 within the clamped area to the total number of locations N. The overpressure rate characterizes the proportion of locations within the clamped area where the plating bears supercritical loads. A higher value indicates a wider range of locations where indentation defects occur in the plating, directly serving as a quantitative expression of the plating's local crushing tendency. Preferably, the preset tendency threshold is set in stages according to the terminal plating type and thickness specifications. For tin-plated terminals with a plating thickness of 3 micrometers or more, the preset tendency threshold is 0.10. Tin plating is relatively soft and has good ductility; after local overpressure, the tin layer can redistribute stress through plastic rheology, allowing for a relatively high overpressure rate. During the calibration stage, indentation detection is performed on tin-plated samples under different overpressure rates. When the overpressure rate is below 10%, the indentation depth does not exceed 1 micrometer and does not affect subsequent assembly. For gold-plated terminals or thin plating specifications with a plating thickness of less than 3 micrometers, the preset tendency threshold is 0.05. Gold plating has a higher hardness than tin, and thin plating layers have a small load-bearing capacity, making them prone to irreversible indentation after overpressure. During the calibration phase, detectable indentations appear when the overpressure rate exceeds 5%. Therefore, 0.05 is chosen as a more stringent threshold, reflecting the extremely small allowable range of overpressure tolerance. If the overpressure rate exceeds a preset tendency threshold, the formation process of the first adjustment strategy is triggered; if the overpressure rate is within the threshold, the current clamping pressure remains unchanged.

[0032] Specifically, for each location within the overpressure zone, the excess of the real-time clamping pressure relative to the critical crushing pressure is calculated. This excess is the value of the aforementioned pressure difference ΔP(i) at the overpressure zone location. A pressure correction amount is determined for each location within the overpressure zone. For the i-th overpressure location, the pressure correction amount is set as the actual excess ΔP(i) plus a safety margin δ. The safety margin δ ranges from 5% to 15% of the excess to ensure that the coating completely exits the critical crushing point after adjustment. No adjustment is made when the excess is less than 0.5 MPa to avoid frequent pressure adjustments due to minor fluctuations. A first adjustment strategy for reducing the output pressure of the clamping head is obtained. This first adjustment strategy includes an array of pressure correction amounts corresponding to each overpressure location and a set of overpressure zone location indices. The clamping head actuator adjusts the output pressure loading value for different points of action according to the first adjustment strategy.

[0033] It is understood that, through the above implementation method, the first adjustment strategy can reflect the coating crushing risk status of the clamped area under the current bending cycle, and provide a direct input basis for the bending control unit to reduce the real-time clamping pressure.

[0034] Step S103: According to the first adjustment strategy, the real-time clamping pressure is reduced to obtain the updated clamping pressure. Under the updated clamping pressure, the metal flow at the junction of the clamped area and the bending area is identified to tighten, forming a mechanical transmission path at the root of the bend.

[0035] Based on the pressure callback benchmark and overpressure zone location index set in the first adjustment strategy, a downward adjustment operation is performed on the output pressure of the pressure head actuator. The downward adjustment range of the pressure head output pressure is equal to the pressure callback benchmark, and the distribution of the downward adjustment points corresponds to the overpressure zone location index set. After the downward adjustment is completed, the pressure values ​​at each position of the clamped area are re-acquired to obtain the updated clamping pressure. The updated clamping pressure forms a new loading distribution at each position of the clamped area. Under the updated clamping pressure, an image of the boundary between the clamped area and the bending area is acquired using a high frame rate industrial camera to obtain the boundary area image. The boundary area image is then processed by grayscale conversion and histogram equalization. The Sobel operator is used to extract the grayscale gradient distribution, and the metal surface texture is identified to offset along the clamping surface towards the bending area to obtain the metal flow direction. The number of texture lines per unit length along the metal flow direction is counted to obtain the texture density. If the texture density exceeds a preset density threshold, it is determined that there is metal flow tightening at that location. All tightening locations are collected to obtain the spatial range of metal flow tightening. Based on the spatial range of the metal flow tightening, connected component labeling is performed on the pixels within the tightening range in the boundary region image. Using the tightening start point as the seed point, a region growing algorithm is used to extend the tightening connected component along the direction of the metal flow, stopping at the bend corner contour line. This connects the tightening start point on the clamped region side with the tightening end point on the bend root region side, forming a mechanical transmission path at the bend root. The mechanical transmission path at the bend root is represented by a strip-shaped connected component, indicating the direction of mechanical transmission from the clamping constraint to the bend root.

[0036] In the bending and forming process of automotive electronic connectors and communication equipment terminals, based on the aforementioned first adjustment strategy, the following specific implementation describes the application process of updated clamping pressure, the identification of metal flow tightening at the interface, and the formation process of the mechanical transmission path at the bending root. In one implementation, the bending control unit receives the first adjustment strategy and, according to the pressure callback reference and the overpressure zone position index set in the first adjustment strategy, issues a pressure reduction command to the pressure head actuator. The pressure head actuator is typically driven by a servo electric cylinder or a servo hydraulic cylinder. After receiving the pressure reduction command, the servo driver adjusts the output torque or hydraulic pressure according to the pressure callback reference. The reduction range of the pressure head output pressure is equal to the pressure callback reference, and the distribution of the reduction action points corresponds to the overpressure zone position index set. After the reduction is completed, the aforementioned piezoelectric pressure sensor array re-acquires the pressure values ​​at each position of the clamped area to obtain the updated clamping pressure.

[0037] It should be noted that the updated clamping pressure characterizes the loading state of the coating after it has passed the critical point of crushing. A new loading distribution is formed at various locations within the clamped area, causing the pressure that previously exceeded the critical crushing pressure to fall back to a safe loading range. Furthermore, under the updated clamping pressure, images are captured at the boundary between the clamped area and the bending area using a high-frame-rate industrial camera. The high-frame-rate industrial camera has a sampling frame rate of no less than 500 frames per second and a resolution of no less than 3 million pixels. The lens is aimed at the side view of the boundary, and a coaxial light source is used to eliminate high reflectivity interference from the metal surface, resulting in an image of the boundary area. This boundary area image reflects the surface texture of the metal at the boundary under the bending cycle; the texture is a visible representation left on the surface by the plastic flow within the metal.

[0038] Specifically, grayscale and histogram equalization are performed on the boundary region image to obtain a grayscale-enhanced boundary region image.

[0039] For example, the Sobel operator is used to extract the gray-level gradient distribution. The Sobel operator calculates the gray-level gradient in the horizontal and vertical directions using two 3x3 convolution kernels, respectively. The horizontal gray-level gradient is denoted as Gx, and the vertical gray-level gradient is denoted as Gy. The gradient magnitude G of a pixel is calculated according to... The gradient direction θ is determined by θ = arctan2(Gy, Gx), and the value of θ ranges from negative π to positive π. This direction reflects the normal direction of the texture lines perpendicular to the surface.

[0040] In one possible implementation, for each pixel in the boundary region image, the gradient direction is divided into intervals of 10 degrees. The number of pixels in each interval is counted to obtain a gradient direction histogram. The interval with the highest percentage in the histogram corresponds to the dominant texture direction of the metal surface at the boundary. The dominant texture direction points from the clamping surface to the bending area, reflecting the macroscopic trend of metal flow along the clamping surface to the bending area during bending, thus obtaining the metal flow direction. The physical meaning of the metal flow direction is that after the clamping head applies pressure to the clamped area, the metal in the constrained area undergoes plastic extrusion towards the adjacent bending area; the extrusion direction is the metal flow direction. The metal flow direction is not only reflected in the texture direction but also in the change in texture density. The transition from sparse to dense areas at the boundary is consistent with the flow direction. Preferably, samples are taken from the boundary region image along the metal flow direction, with the sampling length being the number of pixels corresponding to 1 mm. The number of texture lines crossed within this sampling length is counted to obtain the texture density. The number of pixels is calculated from a known-size calibration board captured by a high-frame-rate industrial camera during the calibration phase. The calibration method is consistent with the image scale calibration method of the lateral industrial camera in S104. The pixel distance of the known spacing marks on the calibration board in the image is divided by the actual spacing to obtain the number of pixels per millimeter. The reference benchmark is the average texture density of a stable region more than 3 millimeters away from the boundary within the clamped area. No metal flow occurs in this stable region during bending, and the average texture density is denoted as D0. If the texture density at a certain position exceeds 1.5 times D0, it is determined that there is metal flow tightening at that position. All positions that meet the tightening judgment are collected to obtain the spatial range of metal flow tightening. Specifically, connected component marking is performed on the pixels within the spatial range of metal flow tightening, and adjacent tightened pixels are merged into a tightened connected component using the 8-neighborhood connectivity rule. The tightening starting point is taken as the boundary point closest to the clamped area in the tightened connected region. The region growing algorithm is used to extend the tightened connected region along the direction of metal flow. The region growing algorithm takes the tightening starting point as the seed point and takes the absolute value of the gray value difference between the adjacent pixel and the seed point as not exceeding the preset growth threshold as the merging condition. The preset growth threshold is taken as 10% of the overall gray dynamic range of the boundary area image after gray enhancement. The connected region range is iteratively expanded.

[0041] It should be noted that in the Sobel gradient distribution of the aforementioned boundary region image, pixels with gradient magnitude G exceeding a preset edge threshold, when connected, form a bend and corner contour line. The preset edge threshold T... edge The method for determining this is as follows: First, statistically analyze the gradient magnitude histogram of the boundary region image, and then calculate the gradient mean μ. G and standard deviation σ G Then according to T edge =μ G +k×σ GThe calculations show that the coefficient k ranges from 1.2 to 2.0. For workpieces with bending angles between 60 and 90 degrees, the difference between the gray-level gradient at the bend corner contour and the background gradient is significant, and k=1.5 is sufficient for effective contour extraction. For workpieces with bending angles between 90 and 120 degrees, large-angle bending causes the metal at the corner to stretch and thin, reducing surface reflectivity and narrowing the difference between the gray-level gradient at the contour and the background gradient. Therefore, k needs to be increased to 1.8 to suppress the background gradient from being misjudged as an edge. The specific calibration method for k is as follows: During the calibration stage, images of samples with known bending angles are acquired, and the true edge pixels on the bend corner contour are manually labeled. The distribution of the gradient amplitude of the true edge pixels is statistically analyzed, and the 5th percentile of the lower limit of the distribution is taken as the value of μ. G +k×σ G The closest calculated value for k is used as the calibration value for that angle setting. The bend corner contour line is the outermost geometric bend boundary of the metal surface on one side of the bend area, serving as the stopping criterion for region growth. This threshold determination method can adaptively adjust according to the image grayscale characteristics of different workpieces, ensuring that the contour line extraction accuracy meets the requirements of subsequent mechanical analysis.

[0042] It is understood that the region growth extends to the bend corner contour line and stops there, thereby connecting the tightening start point on the clamped region side and the tightening end point on the bend root region side, forming a mechanical transmission path at the bend root. The mechanical transmission path at the bend root is characterized in the form of a band-shaped connected domain, representing the direction of mechanical transmission from the clamping constraint to the bend root, providing a spatial positioning basis for subsequent identification of the slip tendency of the bend root region.

[0043] Step S104: Obtain an image of the bending root region where the mechanical transmission path is located, identify the misalignment traces of the metal layer along the mechanical transmission path in the bending root region image, and obtain the slippage tendency of the bending root.

[0044] Based on the spatial orientation of the mechanical transmission path at the root of the bend, the imaging field of view of the bend root region is located. A side-facing industrial camera is used to simultaneously acquire images of the bend root region, obtaining an image covering the mechanical transmission path. Grayscale and image sharpening processes are then performed on the bend root region image to enhance the contrast of fine traces on the metal surface, resulting in a sharpened image of the bend root region. A strip-shaped region of interest is extracted from the sharpened bend root region image along the mechanical transmission path. The Sobel operator is used to extract the grayscale gradient from the strip-shaped region of interest. Pixel connected regions with gradient magnitudes exceeding a preset trace gradient threshold are identified as fault trace connected regions. The two boundary points of the fault trace connected region are taken as the start and end points. The angle between the direction connecting the start and end points and the tangent of the mechanical transmission path is recorded as the trace direction. The coordinate difference between the start and end points in the direction perpendicular to the mechanical transmission path is taken as the misalignment value. Based on the trace direction and the misalignment value, the connected regions of each misalignment trace along the force transmission path are summarized. If the ratio of the number of misalignment traces whose trace direction falls within ±15 degrees along the tangential direction of the force transmission path to the total number of misalignment traces exceeds a preset direction ratio threshold, it is determined that there is directional slippage at the root of the bend along the force transmission path. The maximum value among the misalignment values ​​is taken as the slippage amplitude. The directional slippage determination result and the slippage amplitude together constitute the slippage tendency at the root of the bend.

[0045] In the bending forming process of automotive electronic connectors and communication equipment terminals, based on the aforementioned mechanical transmission path at the bending root, the following specific implementation methods describe the image acquisition, misalignment trace identification, and formation process of the bending root slippage tendency in the bending root region. In one implementation method, based on the spatial orientation of the mechanical transmission path at the bending root in the boundary area image, the endpoint coordinates of the mechanical transmission path at the bending root are back-projected onto the actual spatial coordinate system of the bending station. The imaging field of view of the bending root region is located centered on the coordinate back-projection result. A lateral industrial camera positioned on the side of the bending station performs synchronous image acquisition of the bending root region. The sampling sequence of the lateral industrial camera is aligned with the bending cycle, and the resolution is not less than 5 megapixels, resulting in an image of the bending root region covering the mechanical transmission path. During the calibration phase, the lateral industrial camera photographs a calibration plate with known actual dimensions, obtaining the conversion relationship between pixels and actual length as an image scale for subsequent size conversion of misalignment values.

[0046] Specifically, the misalignment marks on the surface of the metal layer at the root of the bend represent surface morphology changes on the order of micrometers. Directly acquired images show low grayscale contrast between these marks and the background metal substrate, making direct extraction difficult. The image of the bend root region is converted to grayscale by converting the color image to a single-channel grayscale image. Then, image sharpening is performed using a Laplacian sharpening template to increase the high-frequency components of the grayscale image. The subtle unevenness on the metal layer surface manifests as localized grayscale abrupt changes after sharpening, resulting in a sharpened image of the bend root region. Further, a strip-shaped region of interest is extracted from the sharpened bend root region image along the mechanical transmission path. The centerline of the mechanical transmission path serves as the axis of the strip-shaped region, and its width is 1.2 times the width of the mechanical transmission path, thus covering the adjacent metal layers on both sides of the mechanical transmission path. The extraction of the strip-shaped region of interest is achieved through masking operations. The grayscale of the original image is retained inside the mask, while the grayscale outside the mask is set to zero.

[0047] In one possible implementation, the Sobel operator is used to extract gray-level gradients for pixels within the strip-shaped region of interest, resulting in gradient magnitude distribution and gradient direction distribution. The gradient magnitude reflects the intensity of local gray-level abrupt changes, and the gradient direction reflects the direction of local texture.

[0048] For example, the preset trace gradient threshold is determined through offline calibration. Images of the bending root region are acquired on qualified terminal samples in a non-slip state, and the average gradient magnitude is calculated. Three times this average magnitude is taken as the preset trace gradient threshold. This benchmark does not fluctuate with the current batch trace density. Pixels with gradient magnitudes exceeding the preset trace gradient threshold are identified as trace pixels. All trace pixels are merged using an 8-neighborhood connectivity rule. The pixel length is converted to the actual length using the image scale. Isolated connected regions with a length less than 20 micrometers are considered noise and discarded. Connected regions with a length of 20 micrometers or more are retained as misalignment trace connected regions. The misalignment trace connected region reflects the visible representation left on the surface by microscopic misalignment occurring tangentially along the mechanical transmission path between adjacent metal layers. Specifically, for each misalignment trace connected region, the two boundary points of the misalignment trace connected region are extracted as the start and end points. The start point is the boundary pixel closest to the clamped area in the misalignment trace connected region, and the end point is the boundary pixel closest to the bending corner in the misalignment trace connected region. Connecting the starting point and the ending point yields a trace line. The angle between the trace line and the tangent of the mechanical transmission path is taken as the trace direction. The trace direction reflects the skewness of the connected domain of the displacement trace relative to the mechanical transmission path.

[0049] It should be noted that the difference between the coordinate projections of the starting point and the ending point along the direction perpendicular to the force transmission path is recorded as the misalignment value. The misalignment value reflects the displacement component of the connected domain of the misalignment trace along the normal direction of the force transmission path. The value is in pixels and converted to actual size according to the image scale. Preferably, the connected domains of each misalignment trace along the force transmission path are summarized and statistically analyzed. The number of connected domains of misalignment traces whose trace direction falls within ±15 degrees along the tangential direction of the force transmission path is recorded as N2, the total number of connected domains of misalignment traces is N3, and the direction ratio R2 = N2 / N3. If the direction ratio R2 exceeds the preset direction ratio threshold, which is 0.60 for the precision terminal bending process, it is determined that there is directional slippage along the force transmission path at the root of the bend; if the direction ratio R2 is less than the preset direction ratio threshold, it is determined that the misalignment traces at the root of the bend are scattered and do not constitute directional slippage. Specifically, in the scenario where directional slippage is determined to exist, the misalignment values ​​of all connected domains of the misalignment traces are traversed, and the maximum value among all misalignment values ​​is taken as the slippage amplitude. The slippage amplitude reflects the most severe degree of misalignment that occurs tangentially along the mechanical transmission path at the root of the bend.

[0050] It is understandable that the directional slip determination result is used as the first field and the slip amplitude is used as the second field to form a binary form of bending root slip tendency. The bending root slip tendency, in the form of qualitative determination plus quantitative amplitude, characterizes the trend of misalignment of the root metal relative to the matrix along the mechanical transmission path tangentially, providing a direct basis for the subsequent assessment of the instability tendency of the forming angle.

[0051] Step S105: Evaluate the tendency of slippage at the root of the bend and the preset critical slippage, identify the tendency of instability of the forming angle, generate a second adjustment strategy that superimposes anti-slip force on the updated clamping pressure, and obtain the target clamping pressure by superimposing anti-slip force on the updated clamping pressure according to the second adjustment strategy.

[0052] Based on the directional slip determination result and slip amplitude in the slip tendency at the bending root, a numerical comparison is performed between the slip amplitude and the preset critical slip. The preset critical slip adopts the same length dimension as the slip amplitude, and the maximum allowable misalignment value at the bending root of the qualified terminal sample is taken. If the directional slip determination result is present and the slip amplitude exceeds the preset critical slip, it is determined that there is a tendency for the forming angle to be unstable during the bending forming process, and the difference between the slip amplitude and the preset critical slip is recorded as the amplitude exceeding the limit. Based on the amplitude exceeding the limit, a pre-established slip compensation comparison table is consulted. This table is established through offline tests of the bending process. Each entry in the table records an amplitude exceeding the limit segment and the corresponding anti-slip force increment. An anti-slip force increment matching the amplitude exceeding the limit is obtained. Taking the tightening starting point of the bending root mechanical transmission path on the clamped area side as the center, pixel positions within a 3mm radius around the tightening starting point are taken as the application point distribution. This yields a second adjustment strategy for superimposing the anti-slip force on the updated clamping pressure. The second adjustment strategy includes the anti-slip force increment and the application point distribution. According to the second adjustment strategy, the pressure head actuator adds the anti-slip force increment to the corresponding positions of the application point distribution. The added pressure head output pressure is added point-by-point to the aforementioned updated clamping pressure to obtain the target clamping pressure. The target clamping pressure exhibits a loading distribution higher than the updated clamping pressure in the vicinity of the tightening starting point, while maintaining the updated clamping pressure unchanged at other positions.

[0053] In the bending and forming process of automotive electronic connectors and communication equipment terminals, based on the aforementioned tendency of slippage at the bending root, the following specific implementation method describes the identification of the tendency of instability in the forming angle, the formation of a second adjustment strategy, and the formation process of the target clamping pressure. In one implementation method, the bending control unit, as the core execution module of the forming process, integrates a data acquisition interface, a logic judgment circuit, and a parameter retrieval function, and is responsible for real-time monitoring of the bending state and generating adjustment commands. The control unit reads the binary content of the tendency of slippage at the bending root, extracts the directional slippage judgment result as the first component, the slippage amplitude as the second component, and then retrieves the preset critical slippage from the pre-stored parameter area. The preset critical slippage adopts the same length dimension as the slippage amplitude, with the unit being micrometers, and the value is taken from the upper limit of the allowable misalignment at the bending root of the qualified terminal sample. This upper limit reflects the maximum degree of root misalignment allowed when the terminal bending forming angle is still in the stable range.

[0054] Specifically, the slippage amplitude is compared with the preset critical slippage. If the directional slippage determination result is that it exists and the slippage amplitude exceeds the preset critical slippage, it is determined that the bending forming process has a tendency to lose forming angle stability; if the directional slippage determination result is that it does not exist, or the slippage amplitude does not exceed the preset critical slippage, it is determined that the bending forming process is in a stable range, and the aforementioned updated clamping pressure remains unchanged. Further, a pre-established slippage compensation lookup table is consulted to obtain the matching anti-slip force value increment. The slip compensation reference table was established through offline tests of the bending process. The offline test process was as follows: several terminal samples of the same specification were selected, and an initial clamping pressure was deliberately set to be too low for each sample, so that different degrees of initial misalignment were generated in the root area of ​​the bend after bending. The initial misalignment values ​​of each sample were measured and recorded. Then, known force increments were added to the pressure head actuator step by step. After each addition of force increment, the misalignment in the root area of ​​the bend was measured. When the misalignment dropped below the preset critical slip, the addition was stopped. The total force increment added from the initial misalignment to the point below the critical slip was taken as the anti-slip force increment corresponding to the initial misalignment. In this way, an empirical correspondence between the amplitude exceeding the limit section and the anti-slip force increment was established, and the reference table entries were compiled and stored in the parameter library of the bending control unit.

[0055] In one possible implementation, the lookup table entries are key-value pairs of amplitude over-limit segments and anti-slip force increments.

[0056] For example, an amplitude exceedance of less than 5 micrometers corresponds to an increase in anti-slip force of 0.5 MPa, an amplitude exceedance between 5 and 10 micrometers corresponds to an increase in anti-slip force of 1.0 MPa, and an amplitude exceedance of more than 10 micrometers corresponds to an increase in anti-slip force of 1.5 MPa. This mapping relationship is determined by empirical data on the additional pressure required for different displacement amounts in the aforementioned offline tests. Based on the segment into which the amplitude exceedance falls, the corresponding entry is read to obtain the increase in anti-slip force.

[0057] Preferably, the pixel positions within a 3 mm radius around the tightening starting point of the mechanical transmission path at the root of the bend on the clamped area side are taken as the application point distribution. This radius is determined based on stress cloud map analysis in offline tests. When the radius of the additional pressure application range is less than 2 mm, the suppression effect is insufficient; when it is greater than 4 mm, it will generate excessive compressive stress on the non-misaligned area. 3 mm is the optimal radius. The application point distribution covers the pressure head application surface in the vicinity of the tightening starting point. This range corresponds to the entry section of the mechanical transmission path into the clamped area side. Applying additional anti-slip force here can directly suppress the occurrence of misalignment in the root of the bend. The increment of the anti-slip force value and the application point distribution together constitute a second adjustment strategy for superimposing the anti-slip force on the updated clamping pressure.

[0058] Specifically, according to the second adjustment strategy, the bending control unit adds an incremental anti-slip force value to the output of the pressure head actuator at the corresponding position of the application point distribution. This additional action is achieved by controlling the partitioned pressure servo valve of the pressure head actuator. The partitioned pressure servo valve is a multi-channel hydraulic control device that divides the pressure head's working surface into 16 independent control zones. Each zone is equipped with an independent proportional pressure valve and a pressure sensor. The control unit achieves independent setting of pressure from 0 to 10 MPa by sending current signals to the proportional pressure valves of each zone. The pressure sensors provide feedback on the actual output pressure, forming a closed-loop control with a control accuracy of 0.1 MPa. Based on the zone number covered by the application point distribution, the bending control unit sends an incremental control signal to the proportional pressure valve of the corresponding zone, causing the output pressure of that zone to be superimposed with the incremental anti-slip force value on top of the original updated clamping pressure. The added pressure head output pressure is added to the aforementioned updated clamping pressure point by point according to the corresponding position. The position output pressure within the application point distribution is the sum of the updated clamping pressure and the increment of the anti-slip displacement force value. The position output pressure outside the application point distribution remains unchanged with the updated clamping pressure, thus obtaining the target clamping pressure.

[0059] It is understood that the target clamping pressure exhibits a loading distribution that is higher than the updated clamping pressure in the neighborhood of the tightening starting point, while the updated clamping pressure remains unchanged in other positions, providing a basis for the loading distribution for the subsequent evaluation of the plastic flow state at the bending root.

[0060] Step S106: Obtain the stress concentration distribution map under the target clamping pressure, extract the maximum stress in the stress concentration distribution map, evaluate the maximum stress and the initial bearing capacity, identify that no plastic flow has occurred at the root of the bend, and output the target clamping pressure to control the bending forming.

[0061] Under the target clamping pressure, a photoelastic coating is applied to the surface of the terminal bending root using a photoelastic coating method. The coating is then photographed during the bending process using a polarized light imaging device, resulting in a photoelastic fringe pattern reflecting the stress distribution at the bending root. Fringe order identification is performed on the photoelastic fringe pattern, and the fringe order value is converted to the stress value at that location according to the photoelastic fringe calculation formula. The stress values ​​at each location are then collected to form a stress concentration distribution map. The stress values ​​at each location are traversed from the stress concentration distribution map, and the peak value among all stress values ​​is taken as the maximum stress. A numerical comparison is performed between the maximum stress and the aforementioned initial load-bearing capacity to obtain the load-bearing margin. The load-bearing margin is the difference between the initial load-bearing capacity and the maximum stress. The sign of the load-bearing margin reflects the current plastic flow risk state at the bending root. Based on the load-bearing margin, if the load-bearing margin is positive, it is determined that no plastic flow has occurred at the root of the bend. The bending control unit locks the target clamping pressure and sends it to the pressure head actuator. The pressure head actuator completes the bending and forming output of the current cycle according to the target clamping pressure, and obtains a bent terminal with a forming angle in a stable range.

[0062] In the bending and forming process of automotive electronic connectors and communication equipment terminals, based on the aforementioned target clamping pressure, the following specific implementation method describes the acquisition of the stress concentration distribution map at the bending root, the evaluation of the maximum stress and initial load-bearing capacity, and the output process of the target clamping pressure. In the preparation stage before the target clamping pressure is formally applied to the clamped area of ​​the terminal, a photoelastic coating is applied to the surface of the terminal bending root using a spraying process. The preparation stage refers to the no-load period when the terminal enters the bending die but the pressure head has not yet been pressed down. The photoelastic coating is a transparent polymer film with birefringence, and the coating thickness ranges from 0.1 mm to 0.3 mm. The coating adheres tightly to the terminal substrate.

[0063] It should be noted that for miniature terminals with a bending root width of less than 2 mm, a photoelastic coating with a thickness of 0.1 to 0.3 mm will significantly alter the mechanical boundary conditions at the bending root. In this case, the following alternative solution is used: a witness piece of the same specification is placed next to the bending station. The witness piece is not subjected to clamping pressure. Only after bending is it placed on the photoelastic coating measurement station to apply a static load equivalent to the target clamping pressure, and the stress distribution is obtained through photoelastic fringes. This alternative solution uses a witness piece instead of the actual terminal for photoelastic measurement, avoiding the mechanical interference of the coating on the actual bending process. For conventional terminals with a bending root width greater than 5 mm, the coating thickness is negligible relative to the terminal size, and the coating is directly applied to the terminal surface for online measurement. When the terminal undergoes bending deformation under the target clamping pressure, the stress state at the bending root is transmitted to the interior of the photoelastic coating, causing the coating to produce a birefringent phase delay proportional to the principal stress difference. Based on this, the maximum stress σmax at the bending root can be analyzed. The maximum stress σmax is then numerically compared with the initial load-bearing capacity σ0 of the clamped area before bending, yielding the load-bearing margin ΔM = σ0 - σmax. If the load-bearing margin is positive, it is determined that no plastic flow has occurred at the root of the bend, and the current target clamping pressure has taken into account both the anti-collapse of the coating and the anti-slip of the root, while not exceeding the stress limit of the substrate's resistance to plastic deformation. If the load-bearing margin is zero or negative, it is determined that the root of the bend is in a state of plastic flow risk. The bending control unit stops the current cycle execution process and reverts to the second adjustment strategy formation stage to re-form the target clamping pressure, cyclically executing the acquisition of the stress concentration distribution map and the comparison of the maximum stress until the load-bearing margin returns to a positive value.

[0064] Specifically, the photoelastic coating during the bending process is imaged using a polarized light imaging device. This device consists of a white light source, a polarizer, an analyzer, and an industrial camera arranged sequentially along the optical axis. The photoelastic coating being imaged is placed in the optical path between the polarizer and the analyzer, whose polarization directions are orthogonal. The imaging process is triggered synchronously with the bending cycle, resulting in a photoelastic fringe pattern reflecting the stress distribution at the root of the bend. This pattern displays alternating bright and dark stripes, with each stripe corresponding to a specific principal stress difference level. Furthermore, stripe order recognition is performed on the photoelastic stripe pattern. An image thinning algorithm is used to locate the center line of each stripe, and each stripe is numbered outward from the zero-order center along the direction perpendicular to the bend corner contour line to obtain the stripe order value N at each position. The stripe order value and the stress value at that position are converted according to the photoelastic stripe calculation formula σ=N×f÷t, where σ represents the principal stress difference at that position, N represents the stripe order value at that position, f represents the material stripe value of the photoelastic coating, and t represents the thickness of the photoelastic coating. Both f and t are recorded as known parameters in the parameter library of the bending control unit during the coating application stage.

[0065] It should be noted that the formula for calculating photoelastic fringes is a classic formula in photoelastic mechanics. The material fringe value f is given in the product manual of the photoelastic coating, and the thickness t is obtained by on-site measurement with a coating thickness gauge. The stress value at each position of the photoelastic fringes is obtained by converting the values ​​at each position of the bending root. Specifically, the stress values ​​at each position of the bending root are collected to form a stress concentration distribution map. The stress concentration distribution map is a two-dimensional numerical matrix that represents the spatial distribution of the principal stress difference in the bending root region. Each element in the matrix corresponds to a spatial position at the bending root, and the element value is the stress value at that position. The clustering of elements with larger values ​​reflects the stress concentration area.

[0066] In one possible implementation, the stress values ​​at each location are traversed from the stress concentration distribution map, and the peak value among all stress values ​​is taken as the maximum stress σmax. The maximum stress reflects the most severe stress level borne by the bending root. A numerical comparison is performed between the maximum stress and the aforementioned initial bearing capacity, denoted as σ0, to obtain the bearing margin ΔM = σ0 - σmax. The sign of the bearing margin reflects the current plastic flow risk state at the bending root. Preferably, based on the bearing margin, if the bearing margin is positive, it is determined that no plastic flow has occurred at the bending root, and the current target clamping pressure avoids coating crushing and inhibits root slippage, while not exceeding the upper limit of the substrate's resistance to plastic deformation stress. The bending control unit latches the target clamping pressure and sends it to the pressure head actuator. The pressure head actuator completes the bending forming output of the current cycle according to the target clamping pressure, resulting in a bent terminal with a forming angle in a stable range.

[0067] It is understood that, through the above implementation method, the target clamping pressure takes into account both the coating bearing capacity and slip suppression in the same cycle, the stress level at the root of the bend is below the initial bearing capacity, and the terminal output by bending is within the process requirement range of precision terminals in terms of forming angle and coating integrity.

[0068] In step S107, if the maximum stress is less than the initial bearing capacity, it is determined that no plastic flow has occurred at the root of the bend, and the bending is controlled by the target clamping pressure.

[0069] The maximum stress is extracted from the stress concentration distribution map. A numerical comparison is performed between the maximum stress and the initial bearing capacity. If the maximum stress is less than the initial bearing capacity, it is determined that no plastic flow has occurred at the root of the bend. Based on the determination that no plastic flow has occurred at the root of the bend, the bending control unit locks the target clamping pressure and sends it to the pressure head actuator. The pressure head actuator completes the bending and forming output according to the target clamping pressure.

[0070] In the bending forming process of automotive electronic connectors and communication equipment terminals, based on the aforementioned stress concentration distribution diagram and initial load-bearing capacity, the following specific implementation method describes the determination of the plastic flow state at the bending root and the execution process of issuing the target clamping pressure. In one implementation method, the bending control unit traverses the values ​​of each element in the two-dimensional numerical matrix corresponding to the stress concentration distribution diagram, and takes the peak value among all element values ​​as the maximum stress. The maximum stress reflects the most severe principal stress difference level borne in the bending root region. The spatial location of the maximum stress generally falls near the bending corner contour line or in the neighborhood of the tightening end point of the mechanical transmission path.

[0071] Specifically, a numerical comparison is performed between the maximum stress and the initial bearing capacity. If the maximum stress is less than the initial bearing capacity, it is determined that no plastic flow has occurred at the bend root, and the metal matrix at the bend root is still in the elastic deformation range. If the maximum stress is not less than the initial bearing capacity, it is determined that the bend root is in a state of plastic flow risk. The bending control unit stops the current cycle's execution flow and reverts to the second adjustment strategy formation stage. It re-queries the slip compensation lookup table to obtain a smaller anti-slip force increment and re-forms the target clamping pressure. The acquisition of the stress concentration distribution map and the comparison of the maximum stress are performed cyclically until the maximum stress is less than the initial bearing capacity. Further, under the determination that no plastic flow has occurred at the bend root, the bending control unit latches the target clamping pressure and sends the loading value and application point distribution of the target clamping pressure to the pressure head actuator via the industrial bus. The pressure head actuator completes the bending forming output of the current cycle according to the loading distribution of the target clamping pressure, resulting in a bent terminal with a stable forming angle and intact coating.

[0072] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for controlling the bending angle of precision terminals, characterized in that, include: Acquire a terminal surface image, identify the terminal surface image to obtain the coating thickness, and obtain the critical crush pressure and initial load capacity based on the coating thickness; The real-time clamping pressure during the current bending and forming process is obtained, the real-time clamping pressure and the critical crushing pressure are evaluated, the local crushing tendency of the coating in the clamped area is identified, and a first adjustment strategy is generated. According to the first adjustment strategy, the real-time clamping pressure is reduced to obtain an updated clamping pressure. Under the updated clamping pressure, the metal flow at the junction of the clamped area and the bending area is tightened to form a mechanical transmission path at the root of the bend. Obtain an image of the bending root region where the mechanical transmission path of the bending root is located, identify the misalignment traces of the metal layer along the mechanical transmission path of the bending root in the image of the bending root region, and obtain the tendency of the bending root to slip. The tendency of the bending root to slip and the preset critical slip are evaluated, the tendency of the forming angle to become unstable is identified, and a second adjustment strategy of superimposing an anti-slip force on the updated clamping pressure is generated. The target clamping pressure is obtained by superimposing an anti-slip force on the updated clamping pressure according to the second adjustment strategy. Obtain the stress concentration distribution map under the target clamping pressure, extract the maximum stress in the stress concentration distribution map, evaluate the maximum stress and the initial bearing limit, if the maximum stress is less than the initial bearing limit, it is determined that no plastic flow has occurred at the root of the bend, and the bending is controlled by the target clamping pressure.

2. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The process of acquiring a terminal surface image, identifying the plating thickness from the terminal surface image, and obtaining the critical crushing pressure and initial load-bearing limit based on the plating thickness includes: An image of the area where the terminal is clamped is acquired to obtain an image of the terminal surface; Extract the coating area contour from the terminal surface image, and obtain the coating thickness distribution map and coating uniformity according to the gray level difference within the contour. The coating thickness distribution map is used to retrieve the coating pressure threshold value, and the uneven section threshold value is corrected according to the coating uniformity to obtain the crushing critical pressure. The hardness value of the substrate is obtained and superimposed with the pressure bearing value of the corresponding coating to obtain the initial upper limit of the load-bearing capacity.

3. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The process of obtaining the real-time clamping pressure during the current bending process, evaluating the real-time clamping pressure and the critical crushing pressure, identifying the local crushing tendency of the coating in the clamped area, and generating a first adjustment strategy includes: Collect real-time clamping pressure distribution at various locations within the clamped area; The pressure difference is obtained by comparing the real-time clamping pressure distribution with the critical crushing pressure point by point. When the pressure difference is positive, the overpressure zone is marked, and the overpressure rate is obtained according to the ratio of the overpressure zone to the sampling position. Based on the comparison between the overpressure rate and the preset tendency threshold, a first adjustment strategy is generated, which includes a pressure callback benchmark and an overpressure zone location index.

4. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The step of reducing the real-time clamping pressure according to the first adjustment strategy to obtain the updated clamping pressure includes: Based on the pressure pullback benchmark and overpressure zone location index in the first adjustment strategy, the output pressure of the pressure head actuator is adjusted downward according to the pressure pullback benchmark. The adjustment point corresponds to the overpressure zone location index; The pressure values ​​at each location in the clamped area are re-acquired to obtain the updated clamping pressure.

5. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The process of identifying the tightening of metal flow at the junction of the clamped area and the bending area under the updated clamping pressure, forming a mechanical transmission path at the root of the bend, includes: Images of the boundary region are acquired under the updated clamping pressure. Extract the grayscale gradient distribution from the image of the boundary region, identify the texture direction shift, and obtain the direction of metal flow; The texture density along the direction of metal flow is statistically analyzed, and the spatial range of metal flow tightening is determined according to the texture density and a preset density threshold. Connectivity labeling is performed on the pixels within the space and extended along the direction of metal flow to the bend corner contour line to form the mechanical transmission path at the root of the bend.

6. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The step of acquiring an image of the bending root region where the mechanical transmission path at the bending root is located, identifying the misalignment traces of the metal layer along the mechanical transmission path at the bending root in the image of the bending root region, and obtaining the slippage tendency at the bending root includes: Position the imaging field of view according to the mechanical transmission path at the root of the bend, and acquire images of the root area of ​​the bend; The image of the bending root region is sharpened, and a strip-shaped region of interest is extracted along the mechanical transmission path of the bending root. Identify fault trace connected regions within the strip-shaped region of interest where the gray-level gradient exceeds a preset trace gradient threshold. The direction of the trace and the amount of misalignment are determined based on the boundary points at both ends of the connected domain of the misalignment trace, and the directional slippage determination result and slippage amplitude constitute the slippage tendency at the root of the bend.

7. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The assessment of the slippage tendency at the root of the bend and the preset critical slippage, and the identification of the tendency for instability of the forming angle, includes: Based on the directional slip determination result and slip amplitude in the slip tendency at the root of the bend, the slip amplitude is compared with the preset critical slip. When the directional slip determination result is that it exists and the slip amplitude exceeds the preset critical slip, it is determined that there is a tendency for the forming angle to be unstable, and the difference between the slip amplitude and the preset critical slip is recorded as the amplitude exceeding the limit.

8. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The generation of a second adjustment strategy that superimposes an anti-slip force on the updated clamping pressure, and the superimposition of the anti-slip force on the updated clamping pressure according to the second adjustment strategy to obtain the target clamping pressure, includes: The anti-slip force increment is obtained by referring to the slip compensation comparison table based on the difference between the slip tendency at the root of the bend and the preset critical slip. Taking the tightening starting point of the mechanical transmission path at the root of the bend as the center, extract the distribution of application points around the tightening starting point; The incremental anti-slip force value is superimposed on the corresponding output position of the application site distribution. The added pressure is added to the updated clamping pressure according to the position to obtain the target clamping pressure.

9. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, The process of obtaining the stress concentration distribution map under the target clamping pressure, extracting the maximum stress from the stress concentration distribution map, and evaluating the maximum stress relative to the initial bearing capacity includes: Under the clamping pressure of the target, a photoelastic fringe pattern corresponding to the stress distribution at the root of the bend is collected. Perform fringe order identification on the photoelastic fringe pattern, convert the stress value according to the fringe order value, and form the stress concentration distribution map; The stress values ​​in the stress concentration distribution map are traversed, and the peak value is taken as the maximum stress. The difference between the initial upper limit of bearing capacity and the maximum stress is calculated to obtain the bearing capacity margin.

10. The method for controlling the bending angle of a precision terminal according to claim 1, characterized in that, If the maximum stress is less than the initial load limit, it is determined that no plastic flow has occurred at the root of the bend, and the bending is controlled by the target clamping pressure, including: When the maximum stress is less than the initial bearing limit, the bending control unit locks the target clamping pressure; The bending control unit sends the target clamping pressure to the pressure head actuator; The pressure head actuator completes the bending and forming output of the current cycle according to the target clamping pressure, and obtains a bent terminal.

Images