An automated process control method and system for cold extrusion processes

Abstract

The application relates to the technical field of cold extrusion, and provides an automatic process control method and system for a cold extrusion process, the method comprising the following steps: in response to an ejection signal of a cold extrusion device, acquiring a displacement distance of an ejection mechanism in the cold extrusion device and an ejection force of the ejection mechanism; according to the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism, determining a surface degradation index of an extrusion die; when the surface degradation index of the extrusion die is less than a preset surface degradation index threshold value, judging whether the ejection mechanism has an ejection abnormality according to the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism, and determining an abnormality cause when the ejection mechanism has the ejection abnormality; and when the ejection mechanism has the ejection abnormality, determining a control strategy of the cold extrusion device based on the abnormality cause of the ejection mechanism. Through the scheme, the efficiency of cold extrusion process control can be improved.

Description

Technical Field

[0001] This application relates to the field of cold extrusion technology, and in particular to an automated process control method and system for cold extrusion. Background Technology

[0002] On multi-station cold extrusion automated production lines, existing control methods typically rely on programmable logic controllers (PLCs) to execute fixed timing logic, that is, by setting fixed delays to control the actions of the ejection mechanism and the picking up of parts by the robotic arm.

[0003] However, during continuous high-speed production, wear on the mold cavity surface increases the product clamping force, and the ejection mechanism may undergo slight deformation under long-term alternating impacts. These physical changes increase the actual demolding resistance, causing a relative lag in the product's ejection time. Traditional timing control mechanically issues robot pick-up commands according to preset times. Because the controller cannot sense the actual workpiece ejection state, the product may not be completely ejected when the robot enters the mold area, leading to collisions, material jamming, or other issues, severely impacting production stability and yield. The core problem is that, under the combined influence of mold wear and ejection mechanism performance degradation, fixed timing control logic cannot accurately reflect the actual physical ejection state of the product, causing timing conflicts between processes. Summary of the Invention

[0004] This application provides an automated process control method and system for cold extrusion, aiming to solve the technical problem in cold extrusion production where the actual demolding state of the product conflicts with the control commands due to die wear and the performance degradation of the ejection mechanism, which in turn causes collisions or jamming of the robot arm.

[0005] The technical solution of this application is as follows: In a first aspect, this application discloses an automated process control method for a cold extrusion process, comprising: responding to an ejection signal from a cold extrusion equipment, acquiring the displacement distance and ejection force of an ejection mechanism in the cold extrusion equipment; the ejection mechanism being used to eject the product from the extrusion die of the cold extrusion equipment; determining the surface degradation index of the extrusion die based on the displacement distance and ejection force of the ejection mechanism; when the surface degradation index of the extrusion die is less than a preset surface degradation index threshold, determining whether there is an ejection abnormality in the ejection mechanism based on the displacement distance and ejection force of the ejection mechanism, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism; the cause of the abnormality includes product skewing and jamming, insufficient pushing distance, or pushing delay; when there is an ejection abnormality in the ejection mechanism, determining a control strategy for the cold extrusion equipment based on the cause of the abnormality in the ejection mechanism.

[0006] Optionally, the surface degradation index of the extrusion die is determined based on the displacement distance and ejection force of the ejection mechanism, including: determining the sliding start time of the product in the extrusion die based on the displacement distance and ejection force of the ejection mechanism; taking the integral of the ejection force within a first preset time after the sliding start time as the actual ejection work; obtaining the reference ejection work; the reference ejection work is the integral of the ejection force within a first preset time after the sliding start time when the cold extrusion equipment is working for the first time; taking the ratio of the first value to the reference ejection work as the surface degradation index; the first value is the difference between the actual ejection work and the reference ejection work.

[0007] Optionally, the sliding start time of the product in the extrusion die is determined based on the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism, including: determining the force gradient curve based on the partial derivative of the ejection force and the displacement distance at the corresponding time; and determining and taking the zero-crossing point in the force gradient curve as the sliding start time of the product in the extrusion die.

[0008] Optionally, based on the displacement distance and ejection force of the ejection mechanism, it is determined whether there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality when there is an ejection abnormality, including: obtaining the working time of the cold extrusion equipment; using the product of the working time and the preset time adjustment coefficient as the first coefficient; using the product of the first coefficient and the preset ejection force change rate as the target ejection force change rate; determining whether the change rate of the ejection force within the first sliding time window is continuously greater than the target ejection force change rate; the duration of the first sliding time window is the second preset duration; the start time of the first sliding time window is the sliding start time; when the change rate of the ejection force within the first sliding time window is continuously greater than the target ejection force change rate, it is determined that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality is product skew and jamming; when the change rate of the ejection force within the first sliding time window is not continuously greater than the target ejection force change rate, based on the preset displacement distance endpoint of the ejection mechanism, the displacement distance of the ejection mechanism, and the ejection force of the ejection mechanism, it is determined whether there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality when there is an ejection abnormality.

[0009] Optionally, based on the preset displacement distance endpoint of the ejection mechanism, the displacement distance of the ejection mechanism, and the ejection force of the ejection mechanism, it is determined whether there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism, including: determining the first moment when the displacement distance of the ejection mechanism is the preset displacement distance; taking the ejection force after the first moment as the final ejection force in the ejection force of the ejection mechanism; taking the product of the surface degradation index and the preset reference ejection force as the reference ejection force adjustment value; determining whether the average value of the final ejection force is greater than the target reference ejection force; the target reference ejection force is the sum of the preset reference ejection force and the reference ejection force adjustment value; when the average value of the final ejection force is greater than the target reference ejection force, it is determined that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality in the ejection mechanism is insufficient pushing distance; when the average value of the final ejection force is less than or equal to the target reference ejection force, it is determined whether there is an ejection abnormality in the ejection mechanism based on the first moment, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism.

[0010] Optionally, determining whether there is an ejection abnormality in the ejection mechanism based on the first moment, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism, includes: taking the duration between the receiving time of the ejection signal and the first moment as the actual ejection motion duration; taking the product of the surface degradation index and the preset degradation coefficient as the ejection motion duration adjustment coefficient; taking the product of the preset ejection motion duration and the ejection motion duration adjustment coefficient as the ejection motion duration adjustment value; determining whether the target ejection motion duration is greater than the actual ejection motion duration; the target ejection motion duration is the sum of the preset ejection motion duration and the ejection motion duration adjustment value; when the target ejection motion duration is greater than the actual ejection motion duration, determining that there is no ejection abnormality in the ejection mechanism; when the target ejection motion duration is less than or equal to the actual ejection motion duration, determining that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality in the ejection mechanism is a shift delay.

[0011] Optionally, when the cause of the anomaly includes a delay, a control strategy for the cold extrusion equipment is determined, including: obtaining the remaining duration of the current production task of the cold extrusion equipment and a first correspondence; the first correspondence includes a one-to-one correspondence between multiple duration ranges and multiple second coefficients; the second coefficient is positively correlated with the upper limit of the corresponding duration range; the second coefficient corresponding to the duration range in the first correspondence is taken as the target second coefficient; the product of the target second coefficient and the actual ejection motion duration is taken as the gripping delay duration of the product gripper in the cold extrusion equipment; and the control strategy for the cold extrusion equipment is determined to be to delay the gripping time of the product gripper by the gripping delay duration.

[0012] Optionally, the cold extrusion equipment also includes a vibration device. When the abnormal cause is insufficient pushing distance, the control strategy of the cold extrusion equipment is determined, including: using the product of the average value of the final ejection force and the preset third coefficient as the vibration frequency adjustment coefficient; using the product of the preset vibration frequency of the vibration device and the vibration frequency adjustment coefficient as the target vibration frequency; and determining the control strategy of the cold extrusion equipment to control the vibration device to vibrate at the target vibration frequency.

[0013] Optionally, when the cause of the abnormality includes product skewing and jamming, a control strategy for the cold extrusion equipment is determined, including: generating an emergency stop command; determining the control strategy for the cold extrusion equipment to send an emergency stop command to the control device of the cold extrusion equipment, and storing the displacement distance and ejection force of the ejection mechanism in the cold extrusion equipment into a non-volatile memory, so that maintenance personnel can analyze the displacement distance and ejection force of the ejection mechanism in the cold extrusion equipment.

[0014] Secondly, this application also discloses an automated process control system for cold extrusion, comprising: an acquisition device and a processing device; the acquisition device is used to acquire the displacement distance and the ejection force of the ejection mechanism in the cold extrusion equipment in response to the ejection signal of the cold extrusion equipment; the ejection mechanism is used to eject the product from the extrusion die of the cold extrusion equipment; the processing device is used to determine the surface degradation index of the extrusion die based on the displacement distance and the ejection force of the ejection mechanism; the processing device is also used to determine whether there is an ejection abnormality in the ejection mechanism and the cause of the abnormality when the surface degradation index of the extrusion die is less than a preset surface degradation index threshold, based on the displacement distance and the ejection force of the ejection mechanism; the cause of the abnormality includes product skew and jamming, insufficient pushing distance or pushing delay; the processing device is also used to determine the control strategy of the cold extrusion equipment based on the cause of the abnormality when there is an ejection abnormality in the ejection mechanism.

[0015] Beneficial Effects: This application overcomes the shortcomings of traditional fixed-sequence control in responding to equipment performance degradation by acquiring real-time displacement and ejection force data of the ejection mechanism during cold extrusion and constructing a die surface degradation index and dynamic anomaly diagnosis model accordingly. This method can accurately identify faults such as product skewness, insufficient pushing, and delay, and effectively solves robot collision and material jamming problems by dynamically adjusting the gripping time or introducing vibration-assisted control strategies, significantly improving the robustness and product yield of the automated cold extrusion production process. Attached Figure Description

[0016] Figure 1 A flowchart illustrating an automated process control method for a cold extrusion process provided in this application; Figure 2A schematic diagram of the architecture of an automated process control system for a cold extrusion process provided in this application; Figure 3 A flowchart illustrating another automated process control method for cold extrusion provided in this application; Figure 4 A schematic diagram of the architecture of an automated process control system for another cold extrusion process provided in this application. Detailed Implementation

[0017] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application. It should be noted that similar reference numerals and letters in the following drawings indicate similar items; therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0018] To provide a clearer and more detailed explanation of the technical solutions in this application, the basic concepts involved are explained below.

[0019] "Ejection signal" refers to the trigger command issued by the control system of the cold extrusion equipment to drive the ejection mechanism after the extrusion forming is completed and the mold is opened. This signal can be a logic trigger value generated internally by the controller, or it can be an ejection permission command formed after joint confirmation by the mold opening position sensor, the main slide position detection unit, or the process cycle control unit. In other words, the "ejection signal" is not limited to a certain electrical form; its core meaning is that the equipment has entered the process stage where the ejection action is allowed, and the ejection mechanism can begin to push the product away from the extrusion mold.

[0020] The "surface degradation index" is a numerical indicator used to quantify the degree of physical degradation of the surface of an extrusion die cavity due to long-term use, such as wear and increased roughness. This index is preferably dimensionless, reflecting the degree of deviation of the current die condition from its initial reference state. A larger index generally indicates a more significant deterioration in the friction state, surface integrity, or demolding conditions of the die cavity surface compared to its initial state. For example, under the same product, lubrication conditions, and ejection mechanism operation, if more ejection work is required after the workpiece begins to slide, or if the ejection force-displacement relationship exhibits more pronounced resistance characteristics compared to the initial state, the corresponding surface degradation index will increase.

[0021] "Ejection anomaly" refers to an abnormal state during the ejection process where the workpiece fails to detach from the mold cavity as expected due to changes in the mold's condition or deformation of the ejection mechanism itself. This abnormal state does not require the ejection action to completely stop; it also includes situations where, although the ejection mechanism has activated and the position switch has been triggered, the actual demolding state of the workpiece is inconsistent with the expectation. For example, if one side of the workpiece detaches first while the other side remains held, this is considered product misalignment or jamming; if the ejector rod completes its nominal stroke but the workpiece is actually pushed insufficiently, this is considered insufficient pushing distance; if the workpiece eventually detaches, but the detachment time is significantly later than the corresponding time under normal operating conditions, this is considered delayed pushing.

[0022] like Figure 1 As shown, the technical solution disclosed in this embodiment is as follows: S101. In response to the ejection signal of the cold extrusion equipment, the displacement distance of the ejection mechanism in the cold extrusion equipment and the ejection force of the ejection mechanism are obtained; the ejection mechanism is used to eject the product from the extrusion die of the cold extrusion equipment.

[0023] S102. Determine the surface degradation index of the extrusion die based on the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism.

[0024] S103. When the surface degradation index of the extrusion die is less than the preset surface degradation index threshold, determine whether there is an ejection abnormality in the ejection mechanism based on the displacement distance and ejection force of the ejection mechanism, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism; the cause of the abnormality includes product skew and jamming, insufficient pushing distance or pushing delay.

[0025] S104. When there is an ejection abnormality in the ejection mechanism, determine the control strategy for the cold extrusion equipment based on the cause of the abnormality in the ejection mechanism.

[0026] The following will combine Figure 2 This application provides a detailed explanation of the proposed solution.

[0027] The core logic of this automated process control method lies in breaking away from the traditional open-loop control mode based on fixed position and time, and instead adopting closed-loop monitoring and dynamic control based on real-time physical parameters. The system first constructs a data foundation reflecting the dynamic characteristics of the ejection process by real-time acquisition of the displacement distance and ejection force of the ejection mechanism during its operation. Here, "displacement distance" preferably refers to the cumulative advancing distance of the ejection mechanism from receiving the ejection signal to the current moment, and "ejection force" preferably refers to the actual force applied by the ejection mechanism along the ejection direction to the product or force-related components of the product. Subsequently, this dynamic data is used to evaluate the current surface degradation state of the extrusion die, i.e., to calculate the surface degradation index. When the surface degradation index is within an acceptable range (i.e., less than a preset surface degradation index threshold), the system further analyzes the relationship between displacement and ejection force to accurately identify whether any abnormalities have occurred during the ejection process and to specifically classify the causes of these abnormalities. Finally, for different causes of abnormalities, the system automatically matches and outputs corresponding control strategies, thereby achieving adaptive adjustment of the cold extrusion process. The reason for identifying anomalies only when the surface degradation index is less than the preset surface degradation index threshold is that when the mold has degraded to the point of near failure or exceeds the allowable range of use, the mechanical characteristics during the ejection process will deviate from the normal working conditions. At this time, it is more prudent to prioritize mold maintenance, shutdown inspection, or mold replacement strategies. However, when the mold is still within the range of continued use, further subdividing the ejection anomalies can more accurately distinguish whether the general increase in resistance is caused by the surface condition of the mold or whether the abnormal demolding is caused by local problems such as local deformation of the ejector rod or abnormal workpiece posture.

[0028] Regarding the core technical means of "obtaining the displacement distance and ejection force of the ejection mechanism in a cold extrusion device," in some implementations, a high-precision linear displacement sensor can be installed on the drive cylinder or hydraulic cylinder of the ejection mechanism to collect the displacement distance in real time. Simultaneously, a force detection unit is arranged along the force transmission path of the ejection rod to obtain the ejection force. The displacement sensor can be located between the cylinder and the piston rod, or on a slider, connecting plate, or guide seat that moves synchronously with the ejection rod, as long as it can stably reflect the actual displacement of the ejection mechanism along the ejection direction. The force detection unit can be a force sensor installed at the rear end of the ejection rod, between the push plate and the drive component, or in the force path of the fixed support of the ejection mechanism, as long as the measured force change corresponds to the actual ejection force change during the ejection process. To ensure that the displacement data and ejection force data are time-correlated, the control system preferably uses a unified clock to synchronously sample the two types of data, or aligns them according to timestamps after acquisition. If an indirect measurement method is used, the displacement distance can be calculated by collecting the encoder pulse signal of the drive motor, and the jacking force can be indirectly calculated by monitoring the real-time current value of the drive motor and combining it with a pre-calibrated current-torque-thrust conversion relationship. This conversion relationship can be established as follows: during the equipment installation and commissioning phase, the jacking mechanism is operated under different load conditions, and the motor current, transmission output force, and corresponding displacement state are recorded synchronously to form multiple sets of calibration samples; then, based on the transmission structure, reduction mechanism characteristics, lead screw lead, or gear transmission ratio, the correspondence between current change and output thrust is obtained, and this correspondence is stored in the control system for subsequent online estimation.

[0029] Regarding the core technical approach of "determining the surface degradation index of an extrusion die based on the displacement distance and ejection force of the ejection mechanism," some implementations can extract characteristic quantities from the ejection force versus displacement curve, compare these characteristic quantities with corresponding characteristic quantities in the initial state, and use the comparison results as the surface degradation index after normalization. Here, "initial state" preferably refers to the state when the die is initially put into use, the cavity surface is stable, and it has been confirmed to be in a qualified working condition. Extractable characteristic quantities are not limited to a single peak ejection force, but can also include the force rise slope before sliding begins, the force attenuation amplitude after sliding begins, the average ejection force within a specific displacement range, the ejection work in the initial sliding phase, and the degree of ejection force fluctuation. The reason for using multi-feature joint evaluation is that die surface degradation is not always manifested only as an increase in peak ejection force; sometimes it can also manifest as a delayed sliding start time, a rise in the resistance plateau after sliding, and enhanced local oscillations in the force curve.

[0030] Through the above overall technical solution, this application can effectively solve the misjudgment problem caused by fixed timing control logic in the prior art. Specifically, due to mold wear and ejector rod deformation, demolding resistance increases and effective pushing distance decreases. Traditional control logic relies only on position switch signals and cannot perceive these changes in physical state. However, this application can directly reflect the actual force state and motion trajectory of the workpiece in the mold cavity by acquiring the displacement distance and ejection force of the ejection mechanism in real time. Here, "motion trajectory" does not necessarily rely on visual trajectory reconstruction, but can reflect whether the workpiece demolds as expected through the correspondence between the displacement advancement process and the force change process. For example, during normal demolding, the ejection force will undergo relatively stable phased changes such as contact establishment, static friction accumulation, sliding start, and continuous ejection; while during skew and jamming, the force curve often shows abnormal rise, local jump, or high-level oscillation; when the pushing distance is insufficient, the displacement reaches the nominal endpoint, but the corresponding mechanical release characteristics are not complete; when the pushing is delayed, the sliding start time is significantly delayed compared to normal working conditions.

[0031] By calculating the surface degradation index, the system can quantitatively assess the current state of the mold. Even when the mold has not completely failed but has already degraded to a certain extent, it can further combine displacement and ejection force data to accurately determine whether there are specific anomalies such as product skewness / jamming, insufficient ejection distance, or ejection delay. Based on these accurate anomaly diagnoses, the system can dynamically adjust control strategies, such as delaying robot intervention, extending ejection holding time, performing secondary ejection, reducing the cycle time of the next cycle, issuing maintenance warnings, or prohibiting entry into the next workstation, thereby ensuring that subsequent actions match the actual physical state of the workpiece. In this way, the robot no longer blindly enters based solely on a fixed delay, but determines its intervention time based on whether the workpiece has actually detached from the mold; the ejector rod no longer simply retracts immediately after its stroke end, but determines whether to hold or compensate based on whether the workpiece has been effectively ejected. Therefore, the risk of collision between the robot and incompletely detached workpieces can be significantly reduced, and the phenomenon of workpieces being dragged back into the cavity and jammed can be effectively suppressed, thereby reducing the downtime frequency of automated production lines and ensuring the stability of continuous production and product quality.

[0032] Optionally, such as Figure 3 As shown, the surface degradation index of the extrusion die is determined based on the displacement distance and ejection force of the ejection mechanism, including: S301. Determine the starting moment of the product sliding in the extrusion die based on the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism.

[0033] S302. The integral of the ejection force within the first preset time after the start of the sliding motion is taken as the actual ejection work.

[0034] S303. Obtain the reference ejection work; the reference ejection work is the integral of the ejection force within the first preset time after the sliding start moment when the cold extrusion equipment is working for the first time.

[0035] S304. The ratio of the first value to the reference ejection work is used as the surface degradation index; the first value is the difference between the actual ejection work and the reference ejection work.

[0036] In the process of determining the surface degradation index through displacement distance and ejection force, relying solely on a single peak ejection force or a complex machine learning model may lead to unstable feature extraction or excessively high model training costs. In the actual physical process of cold extrusion demolding, the stress state of the workpiece changes significantly the instant it transitions from a static to a sliding state, and the frictional work done in the initial sliding phase most directly reflects the surface roughness of the mold. Therefore, to more accurately, stably, and cost-effectively quantify the degree of mold surface degradation, an evaluation method based on the principle of physical work is needed.

[0037] This specific implementation scheme assesses surface degradation by calculating the ejection work over a specific time period. First, the system needs to accurately pinpoint the "sliding start moment," which is the instant the workpiece overcomes maximum static friction and begins relative displacement within the mold cavity. The "first preset duration" is preferably set to a time window that covers the main frictional energy consumption stage in the initial sliding phase while minimizing interference from the subsequent large-stroke ejection stage. This time window can be determined through multiple trial runs during equipment debugging. Specifically, under conditions of good mold condition, stable product dimensions, and normal lubrication, the ejection force curves of several normal ejection cycles are continuously collected. The duration of the transition from static friction dominance to stable sliding friction dominance after the start of sliding is observed, and then a duration that stably covers this transition stage is selected as the first preset duration. If the mold release characteristics differ significantly for different product models, corresponding first preset duration parameter tables can be established for different product models and automatically called during production switchovers.

[0038] Subsequently, the system integrates the ejection force over the first preset time period after that moment. This integration can be achieved through discrete sampling and accumulation, that is, accumulating the ejection force values ​​corresponding to each sampling interval within the time window in chronological order to obtain the ejection work characterization value for that time period. To avoid excessive fluctuations in the integration result due to sampling noise, it is preferable to smooth the ejection force data before integration, for example, by using moving average, amplitude limiting filtering, or low-pass filtering. The reference ejection work is a reference value recorded by the equipment under brand-new, ideal conditions. More specifically, the reference ejection work is preferably not the data from a single initial operation, but rather a representative reference value obtained after collecting multiple normal ejection cycles under stable operating conditions during the equipment's initial operation phase. The reason for this is that single data may be affected by lubrication distribution, individual workpiece differences, or instantaneous sensor fluctuations, while using statistical results from multiple cycles as a reference is more stable. By calculating the difference between the actual ejection work and the reference ejection work, and further comparing it with the reference ejection work, the dimensionless surface degradation index can be obtained. The larger the index, the more significant the increase in energy consumed in the initial sliding phase compared to the initial state, which also means a more significant deterioration in the friction state of the mold surface.

[0039] Through the above specific implementation scheme, the system cleverly utilizes the energy consumption characteristics of the workpiece in the initial sliding stage to characterize the mold state. In the initial sliding stage, the surface roughness of the mold has the most intense and pure influence on friction. At this time, the ejection force is mainly used to overcome the interfacial resistance in the initial sliding phase. Integrating the ejection force within this specific time window effectively smooths out transient mechanical fluctuation noise and extracts stable and reliable frictional energy consumption characteristics. Comparing this with the baseline state and calculating the ratio not only reduces the impact of differences in absolute force value calibration between different devices but also intuitively reflects the relative deterioration of the mold surface state.

[0040] Furthermore, because this method focuses on a short time window after the sliding begins, rather than the entire ejection process, it can reduce interference from factors such as ejector rod guiding friction, workpiece inertia changes after leaving the mold, and end-of-mechanism buffering during the subsequent ejection stage. In other words, this method extracts local energy characteristics more directly related to the mold surface state, rather than a comprehensive feature encompassing various subsequent factors. Therefore, this calculation method based on the principle of physical work does not require massive sample data for model training, has clear computational logic, and low resource consumption. It can achieve real-time degradation state assessment within the limited computing power of industrial controllers, thus providing a reliable quantitative basis for subsequent anomaly detection.

[0041] In practical applications, the sliding start time of the product in the extrusion die is determined based on the displacement distance and ejection force of the ejection mechanism. This includes: determining the force gradient curve based on the partial derivative of the ejection force and the displacement distance at the corresponding moment; and determining and using the zero-crossing point in the force gradient curve as the sliding start time of the product in the extrusion die.

[0042] In determining the initiation time of sliding, if only a fixed ejection force threshold or displacement threshold is set to judge whether the workpiece has started to slide, the judgment time will often be inaccurate due to fluctuations in the hardness of different batches of materials or slight changes in lubrication conditions. At the moment of demolding, the workpiece undergoes a sudden change in its stress state, from elastic deformation accumulating energy to overcoming static friction and generating macroscopic sliding. In order to accurately capture this physical abrupt change point in complex dynamic mechanical signals, an analytical method that can sensitively reflect the trend of mechanical state changes is needed.

[0043] This specific implementation scheme introduces the concept of "force gradient," which is a measure of how quickly the ejection force changes relative to the displacement distance. To facilitate implementation in the control system, instead of directly performing continuous analytical differentiation, the discrete rate of change between adjacent sampling points is used to approximate the force gradient. Specifically, the control system acquires a set of displacement distance and ejection force data in each sampling cycle and stores several consecutive sets of data in a sliding data window. Subsequently, the changes in ejection force and displacement at adjacent sampling points are compared in chronological order to obtain the force gradient trend at the current stage. This approach retains the physical meaning of "the slope of the ejection force as a function of displacement" while facilitating real-time execution in the industrial controller. In the initial ejection stage, the ejection mechanism contacts the workpiece and gradually applies thrust. At this point, the workpiece has not yet slipped, and the system mainly exhibits elastic deformation and static friction accumulation. The ejection force increases rapidly with displacement, and the force gradient is typically positive and relatively large. When the thrust reaches the maximum static friction, the workpiece begins to slide. At this point, the displacement continues to increase, but the ejection force shifts from a state dominated by static friction to a state dominated by sliding friction. The force growth trend weakens significantly, and may even experience a brief decrease. Consequently, the force gradient decreases rapidly, exhibiting a change characteristic of turning from positive to zero, then to negative, or rapidly approaching zero from a relatively large positive value. Therefore, the zero-crossing point in the force gradient curve, or the point of abrupt change closest to zero, corresponds, in a physical sense, to the critical moment when the workpiece transitions from rest to sliding.

[0044] Through the specific implementation scheme described above, the system transforms its dependence on absolute values ​​into the capture of changing trends. The force gradient essentially reflects the equivalent stiffness change characteristics during the ejection process. Before sliding, the ejection mechanism continues to advance, but the workpiece has not yet undergone significant relative displacement; the system exhibits strong resistance characteristics, and the ejection force accumulates rapidly with displacement. At the instant of sliding, the interface constraints are broken, and the system's force response changes abruptly. The ejection force no longer continues to rise according to its original trend, thus the force gradient shows a significant inflection point. By calculating the changing trend of the ejection force with respect to displacement and finding the zero-crossing or near-zero-crossing abrupt change points, the system can break free from its dependence on fixed force thresholds. In other words, regardless of whether the current clamping force of the mold is generally high or low, as long as the workpiece undergoes a state transition from stationary to sliding, this transition will leave identifiable trend change characteristics on the force gradient curve. Because the identification basis is "trend abrupt change" rather than "absolute magnitude," this method has better adaptability to material batch fluctuations, slight changes in lubrication, and sensor zero-point drift. Furthermore, once the sliding start moment is accurately identified, the time window used to calculate the actual ejection work can be stably anchored at the stage that truly reflects the frictional state of the mold surface. This avoids incorrectly including the static friction accumulation stage or the subsequent stable ejection stage in the calculation due to deviations in the start moment. As a result, the calculation basis for the entire surface degradation index is more reliable, and subsequent judgments on product skewing, insufficient ejection distance, or ejection delay will be more accurate, thus laying a solid data foundation for the entire automated process control method.

[0045] In some preferred embodiments, determining whether there is an ejection abnormality in the ejection mechanism and the cause of the abnormality when the ejection mechanism has an ejection abnormality is based on the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism, including: obtaining the working time of the cold extrusion equipment; using the product of the working time and a preset time adjustment coefficient as a first coefficient; using the product of the first coefficient and a preset ejection force change rate as a target ejection force change rate; determining whether the change rate of the ejection force within a first sliding time window is continuously greater than the target ejection force change rate; the duration of the first sliding time window is a second preset duration; the start time of the first sliding time window is the sliding start time; when the change rate of the ejection force within the first sliding time window is continuously greater than the target ejection force change rate, determining that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality is product skew and jamming; when the change rate of the ejection force within the first sliding time window is not continuously greater than the target ejection force change rate, determining whether there is an ejection abnormality in the ejection mechanism and the cause of the abnormality when the ejection mechanism has an ejection abnormality is based on the preset displacement distance endpoint of the ejection mechanism, the displacement distance of the ejection mechanism, and the ejection force of the ejection mechanism.

[0046] The "working time" mentioned here preferably refers to the cumulative duration of continuous operation of the cold extrusion equipment since its start-up, or the continuous operating time that has been accumulated since the completion of the most recent mold maintenance, mold replacement, or lubrication system maintenance. Specifically, it can be determined by the equipment controller reading the start-up timing register, maintenance record markers, and production line cycle logs. The "ejection force change rate" mentioned here refers to the speed at which the ejection force changes over time. It focuses not on the instantaneous absolute magnitude of the ejection force, but on whether the ejection force exhibits a continuous and significant rapid upward trend over a short period. For example, during normal demolding, the ejection force may experience a brief increase due to frictional fluctuations, but it usually does not maintain a high increase over multiple consecutive sampling intervals. However, when the product is misaligned or stuck, the ejection force often continues to climb in the early stages of demolding, accompanied by unsmooth displacement or a significantly slower displacement speed.

[0047] This specific implementation plan dynamically adjusts the judgment criteria by introducing the equipment's working time and uses the rate of mechanical change within a sliding time window to identify jamming characteristics. First, the system acquires the cumulative continuous working time of the cold extrusion equipment since the current startup or mold maintenance. This working time can be obtained in the following ways: the equipment's main control unit starts timing after the startup self-test is completed; when maintenance personnel perform a "mold maintenance completed" or "lubrication system maintenance completed" confirmation operation on the touchscreen, the system automatically resets and restarts the accumulation; if the production line is connected to the manufacturing execution system, it can also read the task start time, mold maintenance timestamp, and downtime records to correct the continuous working time, thereby avoiding the miscounting of long downtime periods as valid working time. As the working time increases, factors such as mold thermal expansion and lubricant consumption will cause changes in the normal rate of increase in demolding resistance. Therefore, the system multiplies the working time by a pre-set time adjustment coefficient to obtain a first coefficient. This time adjustment coefficient is an empirical value used to quantify the amplification effect of working time on the rate of change of ejector force.

[0048] Subsequently, after the sliding start moment, the system opens a first sliding time window with a duration of a second preset time, and continuously slides this window over time. The "first sliding time window" here refers to a short observation interval of fixed length with a continuously shifting starting point, used to determine whether the output force exhibits a stable and abnormal upward trend within a continuous short period. "Continuously greater than" is preferably understood as meaning that, within the continuous sampling segment covered by the first sliding time window, the rate of change of the output force is always higher than the target rate of change of the output force, or, in engineering implementation, after allowing a very small number of fluctuations not exceeding the noise tolerance, the overall performance still meets the continuous exceedance characteristic. Within each window, the system calculates the actual rate of change of the output force. To ensure stability, the actual rate of change can be obtained using the following process: first, collect output force data at a fixed sampling period; perform de-scratching on the raw data, for example, using median filtering, amplitude limiting filtering, or short-window averaging to filter out electrical noise and instantaneous spikes; then read the output force changes corresponding to the start and end positions of the window to obtain the actual rate of change within that window. When a workpiece becomes misaligned and jammed, its interference with the mold causes the ejection force to rise sharply and continuously within a short period of time. This upward trend significantly exceeds the resistance fluctuations under normal operating conditions. Therefore, when the system detects that the rate of change of the ejection force within the first sliding time window is continuously greater than the dynamically calculated target rate of change of the ejection force, it can be decisively determined that product misalignment and jamming have occurred. Conversely, if there is no continuous abnormally high growth rate, it indicates that jamming has not occurred, and the system will proceed to the investigation process for other abnormal causes. This judgment method is effective because the increase in ejection force caused by normal lubrication decay and temperature rise usually manifests as a slow drift, characterized by an overall baseline rise or a slight increase in local acceleration. However, product misalignment and jamming will cause local geometric interference in the early stage of demolding, resulting in a continuous increase in high-intensity force. Therefore, by combining "dynamic compensation for working time + judgment of continuous over-limit in the first sliding time window", it is possible to distinguish between normal operating condition drift and true jamming.

[0049] The working principle of this specific implementation scheme is to transform static mechanical threshold judgment into dynamic trend judgment based on time windows, and deeply integrate the time dimension characteristics of equipment operation status. By calculating the first coefficient through working time, the system can automatically compensate for the normal increase in resistance caused by long-term equipment operation, avoiding frequent false alarms in the later stages of production. At the same time, the use of a first sliding time window to monitor the rate of change of ejection force can effectively filter out transient mechanical spikes caused by sensor noise or local micro-interference. Only when the ejection force shows a continuous and sharp upward trend is it identified as skew jamming. This multi-dimensional comprehensive judgment mechanism enables the system to accurately identify the real jamming signal when facing complex cold extrusion demolding conditions. Its unique technical effect is that, in response to the normal resistance fluctuation problem caused by lubrication decay and temperature changes in continuous high-speed batch production, as mentioned above, this scheme gives the control system adaptive immunity, greatly improving the robustness and accuracy of product skew jamming identification, thus enabling accurate diagnosis at a very early stage of jamming, preventing further jamming from causing serious mold damage or ejection mechanism overload. Furthermore, since this scheme is based on actual sampling data, dynamic compensation thresholds, and window persistence criteria, its judgment results can not only be used for real-time control, but also as a basis for subsequent maintenance decisions. For example, it can statistically analyze the frequency of skewness and jamming of a certain mold under different working durations, and reverse the adjustment coefficient of the duration and the preset ejection force change rate, thereby continuously improving the system's adaptability under similar working conditions.

[0050] Optionally, based on the preset displacement distance endpoint of the ejection mechanism, the displacement distance of the ejection mechanism, and the ejection force of the ejection mechanism, it is determined whether there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism, including: determining the first moment when the displacement distance of the ejection mechanism is the preset displacement distance; taking the ejection force after the first moment as the final ejection force in the ejection force of the ejection mechanism; taking the product of the surface degradation index and the preset reference ejection force as the reference ejection force adjustment value; determining whether the average value of the final ejection force is greater than the target reference ejection force; the target reference ejection force is the sum of the preset reference ejection force and the reference ejection force adjustment value; when the average value of the final ejection force is greater than the target reference ejection force, it is determined that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality in the ejection mechanism is insufficient pushing distance; when the average value of the final ejection force is less than or equal to the target reference ejection force, it is determined whether there is an ejection abnormality in the ejection mechanism based on the first moment, and the cause of the abnormality when there is an ejection abnormality in the ejection mechanism.

[0051] The “preset displacement distance endpoint” mentioned here preferably refers to the target position of the ejection stroke predefined in the equipment control program. This position is usually determined based on the workpiece structure, mold cavity depth, demolding safety margin, and robot gripping space. The “final ejection force” mentioned here refers to the ejection force data that the system continues to collect after the ejection mechanism reaches the preset displacement distance. It reflects whether the workpiece has been freed from the mold constraint and whether there is still abnormal residual resistance after freeing.

[0052] This specific implementation scheme determines whether the workpiece has truly detached from the mold by analyzing the residual ejection force at the end of the ejection stroke. First, the system monitors the displacement distance of the ejection mechanism in real time. When the displacement distance reaches a preset distance (usually corresponding to the nominal end point or near the end point of the ejection stroke), this moment is recorded as the first moment. This displacement distance can be directly acquired by a linear displacement sensor mounted on the ejection mechanism, or it can be obtained by counting and converting using a servo encoder. To reduce the impact of mechanical transmission backlash and reverse return error, the actual measured linear displacement signal is preferably used as the primary judgment quantity, and the driver feedback position is used as the verification quantity. Under normal circumstances, when the ejection mechanism reaches this position, the workpiece should have completely detached from the mold cavity. At this point, the ejection mechanism only needs to overcome its own mechanical friction, and the ejection force should quickly decrease to a lower baseline level. Therefore, the system extracts the ejection force data after the first moment and defines it as the final ejection force.

[0053] To ensure the comparability of the final ejection force, the sampling interval after the first moment can be fixed in advance as a short time interval or a short displacement interval, as long as this interval is located at the end of the ejection stroke and can stably cover the stage where the workpiece should have detached from the mold. Due to the degradation of the mold surface, even if the workpiece is completely detached, its residual frictional resistance at the moment of detachment and afterwards will increase due to the increase in mold roughness. To avoid misjudging this normal increase in resistance caused by mold degradation as insufficient pushing distance, the system uses the surface degradation index calculated above to dynamically compensate the judgment benchmark. The "surface degradation index" mentioned here preferably reflects the comprehensive degradation state of increased mold cavity surface roughness, coating wear, adhesive wear, or localized scratches; its source can be a quantitative indicator obtained above based on the characteristics of the ejection force curve, operating cycle statistics, mold maintenance records, or manual inspection results. Specifically, the surface degradation index is multiplied by the preset benchmark ejection force (i.e., the residual thrust at the end of the stroke under ideal conditions) to obtain the benchmark ejection force adjustment value. The preset reference ejection force is preferably obtained by continuously collecting residual ejection force data of multiple batches of qualified workpieces after the first moment, under conditions of good mold condition, no wear on the ejector rod, and normal lubrication. After removing outliers, a stable level is selected as the reference. Then, the preset reference ejection force is added to this adjustment value to obtain the target reference ejection force after dynamic compensation.

[0054] Finally, the system calculates the average value of the final ejection force and compares it with the target reference ejection force. To improve the stability of the average value, the final ejection force can be deburred and limited before calculating the average value to avoid amplifying judgment errors due to local vibration and electrical signal jitter. If the average value of the final ejection force is significantly greater than the target reference ejection force, it indicates that there is still a large demolding resistance after the ejection mechanism reaches the nominal endpoint. This means that the workpiece has not completely detached from the mold cavity, thus accurately determining that insufficient pushing distance has occurred. Conversely, it indicates that the workpiece has basically detached, and the system will continue to check for pushing delay. The fundamental reason for this judgment logic is that once the workpiece is truly freed from the mold constraint, the main load acting on the ejection mechanism will change from "demolding resistance between the workpiece and the mold" to "the running resistance of the mechanism itself," and there is a significant difference in the magnitude and duration of the two. Therefore, the residual force level at the end of the stroke is a direct basis for identifying insufficient pushing distance, and the introduction of the surface degradation index allows this basis to adapt to the normal rise caused by mold aging.

[0055] The working principle of this specific implementation scheme lies in utilizing the abrupt change in the mechanical boundary conditions at the moment the workpiece leaves the mold, combined with dynamic compensation of the mold surface condition, to achieve accurate identification of insufficient ejection distance. When the workpiece completely leaves the mold, the demolding resistance disappears instantly, and the ejection force drops sharply; however, if the ejection distance is insufficient, part of the workpiece remains in the mold, and the ejection force remains at a relatively high level. By calculating the average value of the final ejection force, the mechanical fluctuations at the end of the stroke can be effectively smoothed. More importantly, a surface degradation index is introduced to calculate the reference ejection force adjustment value, allowing the target reference ejection force to adaptively increase as the mold ages. Its unique technical effect is that, addressing the implicit insufficient ejection distance problem caused by the slight plastic bending of the ejector rod mentioned above, this scheme breaks through the blind spot of traditional position control logic. It no longer blindly trusts the position sensor's positioning signal, but instead uses the actual force state at the end of the stroke to infer the actual position of the workpiece. Meanwhile, the dynamic compensation mechanism effectively decouples the coupling effects of mold surface degradation and mechanism deformation on the end ejection force, avoiding false alarms caused by normal mold wear and ensuring high sensitivity and reliability in diagnosing insufficient ejection distance. Furthermore, the diagnostic results can be linked with the maintenance process. For example, when the ejection distance is determined to be insufficient for several consecutive cycles, the system can prompt the user to check the straightness of the ejector rod, the end wear morphology, the installation coaxiality, and any foreign objects remaining in the mold guide section, so that the diagnostic results can directly serve the maintenance positioning.

[0056] In practical applications, determining whether the ejection mechanism has an ejection anomaly based on the first moment, and the cause of the anomaly when it exists, includes: using the time between the receiving time of the ejection signal and the first moment as the actual ejection motion duration; using the product of the surface degradation index and the preset degradation coefficient as the ejection motion duration adjustment coefficient; using the product of the preset ejection motion duration and the ejection motion duration adjustment coefficient as the ejection motion duration adjustment value; determining whether the target ejection motion duration is greater than the actual ejection motion duration; the target ejection motion duration is the sum of the preset ejection motion duration and the ejection motion duration adjustment value; when the target ejection motion duration is greater than the actual ejection motion duration, determining that the ejection mechanism does not have an ejection anomaly; when the target ejection motion duration is less than or equal to the actual ejection motion duration, determining that the ejection mechanism has an ejection anomaly, and the cause of the anomaly is a shift delay.

[0057] The "pushing delay" mentioned here differs from the aforementioned "insufficient pushing distance." The former emphasizes that although the workpiece eventually reaches its position, the arrival time is too late, while the latter emphasizes that the workpiece has not been pushed to its proper position after the ejection mechanism reaches its nominal endpoint. Therefore, in the diagnostic sequence, it is necessary to first rule out skewness and jamming, then rule out insufficient pushing distance, and finally make a judgment on pushing delay to avoid confusion between diagnostic objects. The "receiving time of the ejection signal" mentioned here is preferably the time point when the ejection execution unit confirms successful reception after the equipment control system sends the ejection action command to the ejection mechanism. This time point can come from one or more of the following: the moment when the PLC output action position changes, the moment when the servo drive command is issued and confirmed, and the moment when the hydraulic valve or pneumatic valve is enabled.

[0058] This specific implementation scheme accurately identifies the delay phenomenon by comparing the actual ejection time with the dynamically adjusted allowable time. First, the system records the initial moment of receiving the ejection signal and calculates the time difference between this initial moment and the first moment determined earlier (i.e., the moment the ejection mechanism reaches the preset displacement distance), using this difference as the actual ejection motion duration. This duration accurately reflects the time taken to complete the entire ejection action under the current working conditions. In actual acquisition, the PLC can write a timestamp when the output ejection command bit flips, and the displacement monitoring module can write a timestamp when the displacement distance first reaches the preset displacement distance. The control program then reads the difference between the two as the actual ejection motion duration. If the equipment adopts a distributed control structure, the industrial clock can be used to synchronize the time of each module before calculation. Since mold surface degradation inevitably leads to a certain degree of increased resistance and speed reduction, the system cannot use an absolutely fixed standard time to measure whether there is a delay, otherwise, a large number of false alarms will occur. Therefore, the system again uses the surface degradation index, multiplying it by a preset degradation coefficient to obtain the ejection motion duration adjustment coefficient. The preset degradation coefficient is used to characterize the theoretical influence weight of the degree of mold degradation on the ejection speed.

[0059] Next, the preset ejection motion duration under ideal conditions is multiplied by the adjustment coefficient to calculate the ejection motion duration adjustment value. This preset ejection motion duration is generally derived from the statistical results of multiple qualified ejection cycles collected when the mold is in good condition, lubrication is normal, and the ambient temperature is stable, with a typical duration value from a stable production stage being preferred. Adding the preset ejection motion duration to this adjustment value yields the maximum reasonable time allowed under the current mold condition, i.e., the target ejection motion duration. Finally, the system compares the target ejection motion duration with the actual ejection motion duration. If the target ejection motion duration is still greater than the actual ejection motion duration, it indicates that the current ejection time is within the reasonable fluctuation range allowed by mold degradation, and the system determines that there is no ejection abnormality, and the entire ejection process is completed smoothly. Conversely, if the actual ejection motion duration exceeds the dynamically relaxed target ejection motion duration, it indicates that the ejection process has encountered unexpected resistance, resulting in severe lag, and the system determines that a delay has occurred. The reason this scheme works effectively is that the rate decrease caused by surface degradation usually increases gradually with the degree of degradation, which manifests as a predictable and tolerable time extension. However, abnormal delays will cause the actual time to exceed this reasonable tolerance range. Therefore, by using the method of "preset ejection movement duration + degradation compensation", the effects of normal aging and abnormal time lags can be distinguished.

[0060] The working principle of this specific implementation scheme lies in constructing a dynamic time tolerance model based on the physical state of the mold. It recognizes that a moderate extension of ejection time as the mold ages is a normal phenomenon consistent with physical laws. Through joint calculation of the surface degradation index and the preset degradation coefficient, the system assigns a reasonable elastic range to the preset ejection motion duration. Only when the actual time exceeds the upper limit of this elastic range is it considered a true delay. Its unique technical effect is that, addressing the aforementioned problem of motion timing misalignment caused by increased demolding resistance, this scheme effectively distinguishes between normal physical decay and abnormal motion lag. It avoids the drawback of frequently triggering false delay alarms in the later stages of the mold's lifespan when using a single fixed time threshold. By accurately locking the delay, the system can provide accurate time references for subsequent robotic gripping actions, thereby completely breaking the rigid constraints of fixed timing control logic and ensuring the efficient and coordinated operation of the automated production line throughout the entire mold lifecycle. Furthermore, this dynamic time tolerance model is easy to migrate between different product specifications. It can be adapted under the same diagnostic framework simply by recalibrating the preset ejection motion duration and preset degradation coefficient for different workpieces.

[0061] In a preferred embodiment of this application, when the cause of the abnormality includes a delay, determining the control strategy for the cold extrusion equipment includes: obtaining the remaining duration of the current production task of the cold extrusion equipment and a first correspondence; the first correspondence includes a one-to-one correspondence between multiple duration ranges and multiple second coefficients; the second coefficient is positively correlated with the upper limit of the corresponding duration range; the second coefficient corresponding to the duration range in the first correspondence is taken as the target second coefficient; the product of the target second coefficient and the actual ejection motion duration is taken as the gripping delay duration of the product gripper in the cold extrusion equipment; and the control strategy for the cold extrusion equipment is to control the gripping time of the product gripper to be delayed by the gripping delay duration.

[0062] The "remaining time of the current production task" mentioned here preferably refers to the remaining time of the current batch order, the current production work order, or the current production schedule task until the planned completion time, or it can be the remaining processing time estimated according to the current cycle time. The "first correspondence" mentioned here is a pre-established one-to-one correspondence between a set of remaining time ranges and a second coefficient, used to select different safety margins based on the urgency of the task. Among them, "the second coefficient is positively correlated with the upper limit of the corresponding time range" means that the closer the remaining time range is to the end of the task, the larger the capture delay margin given by the system, so as to reduce the risk of line stoppage due to anomalies in the final stage.

[0063] This specific implementation plan dynamically calculates the optimal intervention time for the product gripper by introducing the remaining duration of the production task and combining it with the actual ejection time. First, the system communicates with the manufacturing execution system (MES) or the local task scheduling module to obtain the remaining duration of the production task currently being executed by the cold extrusion equipment. This remaining duration can be obtained in the following ways: if the production line is connected to the MES, the planned completion time, completed quantity, target quantity, and real-time cycle time of the work order are directly read and converted to obtain the remaining duration; if not connected to the upper-level system, the equipment controller estimates it based on the locally set target output, completed count, and the average cycle time per unit over a recent period. Simultaneously, the system pre-stores a first correspondence, which establishes a mapping between a series of remaining duration ranges and a second coefficient. This second coefficient is a time amplification factor, and its core design logic is that the second coefficient is positively correlated with the upper limit of the corresponding duration range. That is, the longer the remaining duration (the heavier the task), the smaller the corresponding second coefficient, and the system tends to use a more compact delay to maintain capacity; the shorter the remaining duration (the task is about to end), the larger the corresponding second coefficient, and the system tends to use a more generous delay to maintain stability. The first correspondence is preferably established by combining historical production data and safety risk assessment. Specifically, it can be done as follows: First, divide the remaining time into several ranges according to the task stage, such as the task end interval, the task middle interval, and the task beginning interval; then, when a delay occurs in each interval, analyze the impact of different grasping delay margins on grasping success rate, collision rate, and cycle time loss; finally, process engineers and production managers jointly determine the second coefficient corresponding to each time range and write it into the control system in the form of a parameter table.

[0064] Subsequently, the system multiplies the target second coefficient by the actual ejection motion duration calculated earlier to calculate the final gripping delay duration. The actual ejection motion duration is used as the delay calculation benchmark because it directly reflects the true degree of lag in the ejection action within the current cycle. Combining it with the target second coefficient allows the final delay duration to reflect both the severity of the current anomaly and the different preferences for safety and efficiency at each stage of the task. Finally, the system generates a control command to shift the original gripping time of the product gripper backward by this gripping delay duration, ensuring that the gripper only performs gripping after the workpiece is fully in place and its posture is stable. This control command can be implemented by modifying the robot arm trigger delay parameters via the PLC, sending a new gripping start time offset to the robot controller, or temporarily blocking the original gripping trigger signal and re-issuing the gripping permission signal after the delay. If the system simultaneously detects a continuous increase in the shifting delay, it can also prompt a reduction in equipment cycle time or schedule maintenance while delaying the gripping to prevent the anomaly from escalating.

[0065] The working principle of this specific implementation scheme lies in the deep integration of the underlying equipment physical state (actual ejection motion duration) with the top-level production scheduling information (remaining time). The actual ejection motion duration provides an objective benchmark for the degree of delay, while the second coefficient, dynamically selected based on the remaining time, endows the system with the ability to intelligently balance efficiency and safety. Through a positively correlated mapping relationship, the system can automatically switch to a conservative mode with a high safety margin at the end of the task to prevent the entire batch from being delayed due to equipment failure at the last moment; while in the early or middle stages of the task, it maintains a relatively aggressive cycle control to maximize equipment utilization. Its unique technical effect is that, in response to the risk of robot collision caused by the aforementioned delay, this scheme not only provides an adaptive collision avoidance method, but more importantly, it breaks the limitations of single equipment control and achieves collaborative optimization of equipment-level anomaly handling and workshop-level production scheduling. This flexible and intelligent delay control strategy enables automated production lines to achieve the highest operational stability with the lowest production capacity cost when facing motion lag caused by mold degradation, greatly improving the intelligence level of automated process control in cold extrusion. Furthermore, since the generation of the capture delay duration is directly derived from the actual ejection motion duration and the first correspondence, the control strategy has clear data basis and traceability, which facilitates retrospective analysis in production line review, parameter optimization, and quality incident analysis.

[0066] This application further proposes that the cold extrusion equipment also includes a vibration device, and when the abnormal cause includes insufficient pushing distance, the control strategy of the cold extrusion equipment is determined, including: using the product of the average value of the final ejection force and the preset third coefficient as the vibration frequency adjustment coefficient; using the product of the preset vibration frequency of the vibration device and the vibration frequency adjustment coefficient as the target vibration frequency; and determining the control strategy of the cold extrusion equipment to control the vibration device to vibrate at the target vibration frequency.

[0067] This specific implementation scheme introduces high-frequency micro-amplitude vibration to disrupt the residual static friction between the workpiece and the die. In terms of hardware configuration, the cold extrusion equipment pre-installs a vibration device on the base of the extrusion die, the demolding push plate, or the fixed support of the ejection mechanism. This vibration device is preferably a high-frequency pneumatic vibrator, a piezoelectric ceramic vibrator, or an eccentric motor vibrator. When the system determines that the pushing distance is insufficient, it means that although the workpiece has been mostly ejected, it is still locally constrained by friction within the die cavity. At this point, the "average value of the final ejection force" calculated earlier accurately reflects the magnitude of this residual constraint resistance. The greater the residual resistance, the tighter the workpiece is stuck, requiring stronger excitation energy to loosen it.

[0068] Therefore, the system multiplies the average value of the final ejection force by a preset third coefficient to calculate the vibration frequency adjustment coefficient. This preset third coefficient is a proportional conversion factor used to convert mechanical dimensions into frequency adjustment weights; its value can be obtained through demolding tests during the equipment commissioning phase. Subsequently, the system multiplies the preset vibration frequency of the vibration device under standard operating conditions by this adjustment coefficient to obtain the dynamically generated target vibration frequency. Finally, the control system sends a command to the drive unit of the vibration device to start vibration at the target vibration frequency. In practical applications, the vibration duration can be set to an extremely short safety window (e.g., 0.5 to 1 second). Accompanied by high-frequency micro-vibration, the static friction between the workpiece and the mold contact surface is instantly converted into dynamic friction, allowing the workpiece to smoothly slide off the mold under its own weight or a small residual thrust, completely detaching from the mold.

[0069] The unique technical advantage of this specific implementation scheme lies in providing a flexible and adaptive demolding compensation mechanism. Traditional methods often resort to secondary hard pushing when the pushing distance is insufficient, which can easily lead to deformation of thin-walled workpieces or localized tearing of the mold cavity. This scheme cleverly utilizes the physical characteristics of vibration-induced friction locking and can dynamically match the vibration frequency according to the magnitude of residual resistance. Higher resistance results in a higher frequency and stronger excitation force; lower resistance results in a lower frequency, avoiding excessive vibration that could affect equipment lifespan. This non-destructive intervention strategy not only effectively eliminates the risk of robot collisions due to insufficient pushing distance but also significantly reduces the number of downtime alarms caused by minor demolding failures, significantly improving the self-healing capability and continuous operation rate of automated production lines.

[0070] In a preferred embodiment of this application, when the cause of the abnormality includes product skewing and jamming, the control strategy of the cold extrusion equipment is determined, including: generating an emergency stop command; determining the control strategy of the cold extrusion equipment to send an emergency stop command to the control device of the cold extrusion equipment, and storing the displacement distance and ejection force of the ejection mechanism in the cold extrusion equipment into a non-volatile memory, so that maintenance personnel can analyze the displacement distance and ejection force of the ejection mechanism in the cold extrusion equipment.

[0071] This specific implementation plan establishes a dual response mechanism of "immediate loss mitigation and data preservation." Once the system, through the dynamic time window monitoring logic described above, confirms a continuous and abnormal increase in the rate of change of the ejector force and determines that the product is skewed and stuck, the internal logic processing unit will generate an emergency stop command with the highest priority. This command is quickly sent to the underlying control equipment of the cold extrusion equipment (such as a programmable logic controller (PLC) or motion controller), forcibly cutting off the power source of the ejector mechanism. For example, for a servo-driven ejector mechanism, the servo enable is immediately deactivated and the brake is triggered; for a hydraulically driven ejector mechanism, the proportional directional valve is immediately controlled to return to the neutral position and the unloading circuit is opened, thereby preventing the ejector rod from continuing to apply destructive thrust within milliseconds. While executing the physical emergency stop, the system simultaneously initiates the data preservation process. The system will package and write the ejector mechanism displacement distance sequence and ejector force sequence, which were collected at high frequency at the moment of the stuck incident and for a period of time prior, along with the corresponding timestamps, into non-volatile memory. The non-volatile memory here can be a solid-state drive inside an industrial control computer, a Flash memory card for PLC expansion, or an electrically erasable programmable read-only memory (EEPROM). Because non-volatile memory has the characteristic of not losing data when power is lost, even if the equipment triggers a power outage due to severe overload, this critical "crime scene" data can still be safely saved.

[0072] The unique technical advantage of this specific implementation scheme lies in its maximum protection of the expensive extrusion die and core transmission mechanism from irreversible physical damage. Through millisecond-level emergency stop response, the system limits the destructive force of skewing and jamming to an extremely early stage. More importantly, by storing high-precision force and displacement data in non-volatile memory, it's equivalent to equipping the cold extrusion equipment with a highly reliable "black box." When maintenance personnel intervene, there's no need to blindly disassemble the die; simply retrieving the data curves from the memory allows for precise analysis of the point of force abrupt change and displacement stagnation at the time of jamming. This provides objective and detailed quantitative evidence for analyzing the root cause of the fault (such as local lubrication failure, severe wear on one side of the die, uneven billet hardness, or poor initial alignment of the ejector rod), thereby significantly shortening the time for troubleshooting and equipment restoration.

[0073] Based on the above-mentioned automated process control methods for cold extrusion, such as Figure 4As shown, this application also proposes an automated process control system for cold extrusion, including: an acquisition device and a processing device; the acquisition device is used to acquire the displacement distance and the ejection force of the ejection mechanism in the cold extrusion equipment in response to the ejection signal of the cold extrusion equipment; the ejection mechanism is used to eject the product from the extrusion die of the cold extrusion equipment; the processing device is used to determine the surface degradation index of the extrusion die based on the displacement distance and the ejection force of the ejection mechanism; the processing device is also used to determine whether there is an ejection abnormality in the ejection mechanism and the cause of the abnormality when the surface degradation index of the extrusion die is less than a preset surface degradation index threshold, based on the displacement distance and the ejection force of the ejection mechanism; the cause of the abnormality includes product skew and jamming, insufficient pushing distance or pushing delay; the processing device is also used to determine the control strategy of the cold extrusion equipment based on the cause of the abnormality when there is an ejection abnormality in the ejection mechanism.

[0074] In practical industrial deployments, the acquisition devices in this system can consist of a high-frequency data acquisition card, a linear displacement sensor, a pressure sensor, or a servo drive's data communication interface, responsible for stably and synchronously acquiring underlying physical signals in harsh electromagnetic environments. The processing devices can be implemented by an industrial control computer with a real-time operating system, a high-performance programmable logic controller (PLC), or an edge computing gateway, internally running the data processing, feature extraction, dynamic compensation, and logical judgment algorithms described above.

[0075] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. 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.

Claims

1. An automated process control method for cold extrusion, characterized in that, include: In response to the ejection signal of the cold extrusion equipment, the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism in the cold extrusion equipment are obtained; The ejection mechanism is used to eject the product from the extrusion die of the cold extrusion equipment; The surface degradation index of the extrusion die is determined based on the displacement distance and ejection force of the ejection mechanism. When the surface degradation index of the extrusion die is less than the preset surface degradation index threshold, the ejection mechanism is judged to have an ejection abnormality based on the displacement distance and ejection force of the ejection mechanism, and the cause of the abnormality when the ejection mechanism has an ejection abnormality; the cause of the abnormality includes product skew and jamming, insufficient pushing distance or pushing delay. When there is an ejection abnormality in the ejection mechanism, the control strategy for the cold extrusion equipment is determined based on the cause of the abnormality.

2. The automated process control method for cold extrusion according to claim 1, characterized in that, The surface degradation index of the extrusion die is determined based on the displacement distance and ejection force of the ejection mechanism, including: The moment when the product begins to slide in the extrusion die is determined based on the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism. The integral of the ejection force within the first preset time after the start of the sliding motion is taken as the actual ejection work. Obtain the reference ejection work; the reference ejection work is the integral of the ejection force within the first preset time after the sliding start moment when the cold extrusion equipment is working for the first time. The ratio of the first value to the reference ejection work is used as the surface degradation index; the first value is the difference between the actual ejection work and the reference ejection work.

3. The automated process control method for cold extrusion according to claim 2, characterized in that, Based on the displacement distance and ejection force of the ejection mechanism, determine the starting moment of the product sliding in the extrusion die, including: The force gradient curve is determined based on the partial derivative of the ejection force with the displacement distance at the corresponding moment. The zero-crossing point in the force gradient curve is determined and taken as the starting moment of the product sliding in the extrusion die.

4. The automated process control method for cold extrusion according to claim 3, characterized in that, Based on the displacement distance and ejection force of the ejection mechanism, determine whether there is an ejection abnormality in the ejection mechanism, and if so, the cause of the abnormality, including: Obtain the operating time of the cold extrusion equipment; The product of the working hours and the preset duration adjustment coefficient is used as the first coefficient; The product of the first coefficient and the preset rate of change of top output force is taken as the target rate of change of top output force. Determine whether the rate of change of top output force within the first sliding time window is continuously greater than the target rate of change of top output force; the duration of the first sliding time window is the second preset duration; the start time of the first sliding time window is the sliding start time; If the rate of change of the ejection force is continuously greater than the rate of change of the target ejection force within the first sliding time window, it is determined that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality is product skew and jamming. If the rate of change of the ejection force does not continuously exceed the rate of change of the target ejection force within the first sliding time window, the ejection mechanism is judged to have an ejection abnormality based on the preset displacement distance endpoint of the ejection mechanism, the displacement distance of the ejection mechanism, and the ejection force of the ejection mechanism, and the cause of the abnormality when the ejection mechanism has an ejection abnormality.

5. The automated process control method for cold extrusion according to claim 4, characterized in that, Based on the preset displacement endpoint of the ejection mechanism, the displacement distance of the ejection mechanism, and the ejection force of the ejection mechanism, determine whether there is an ejection abnormality in the ejection mechanism, and the reasons for the abnormality when such an abnormality exists, including: The first moment when the displacement distance of the ejection mechanism is determined to be the preset displacement distance; The ejection force after the first moment in the ejection force of the ejection mechanism is taken as the final ejection force; The product of the surface degradation index and the preset benchmark top output force is used as the benchmark top output force adjustment value; Determine whether the average value of the final jacking force is greater than the target reference jacking force; the target reference jacking force is the sum of the preset reference jacking force and the reference jacking force adjustment value. When the average value of the final ejection force is greater than the target reference ejection force, it is determined that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality in the ejection mechanism is insufficient pushing distance. When the average value of the final ejection force is less than or equal to the target reference ejection force, the ejection mechanism is judged to have an ejection abnormality based on the first moment, and the cause of the abnormality when the ejection mechanism has an ejection abnormality.

6. The automated process control method for cold extrusion according to claim 5, characterized in that, Based on the initial assessment of whether there is an ejection abnormality in the ejection mechanism, and the reasons for such abnormalities, the following should be considered: The duration between the receiving time of the ejection signal and the first moment is taken as the actual ejection motion duration; The product of the surface degradation index and the preset degradation coefficient is used as the adjustment coefficient for the ejection motion duration; The product of the preset ejection motion duration and the ejection motion duration adjustment coefficient is used as the ejection motion duration adjustment value; Determine whether the target ejection motion duration is greater than the actual ejection motion duration; the target ejection motion duration is the sum of the preset ejection motion duration and the ejection motion duration adjustment value. When the duration of the target ejection motion is longer than the actual ejection motion duration, it is determined that there is no ejection abnormality in the ejection mechanism; When the target ejection motion duration is less than or equal to the actual ejection motion duration, it is determined that there is an ejection abnormality in the ejection mechanism, and the cause of the abnormality in the ejection mechanism is a push delay.

7. The automated process control method for cold extrusion according to claim 6, characterized in that, When abnormal causes include delays, determine the control strategy for the cold extrusion equipment, including: Obtain the remaining duration of the current production task of the cold extrusion equipment and the first correspondence; the first correspondence includes a one-to-one correspondence between multiple duration ranges and multiple second coefficients; the second coefficient is positively correlated with the upper limit of the corresponding duration range; The second coefficient corresponding to the duration range of the remaining duration in the first correspondence is taken as the target second coefficient; The product of the target second coefficient and the actual ejection motion time is used as the gripping delay time of the product gripper in the cold extrusion equipment; The control strategy for the cold extrusion equipment is determined to be to delay the gripping time of the product gripper by a certain delay duration.

8. The automated process control method for cold extrusion according to claim 5, characterized in that, The cold extrusion equipment also includes a vibration device. When abnormalities occur, such as insufficient traction distance, a control strategy for the cold extrusion equipment is determined, including: The product of the average value of the final ejection force and the preset third coefficient is used as the vibration frequency adjustment coefficient. The product of the preset vibration frequency of the vibration device and the vibration frequency adjustment coefficient is taken as the target vibration frequency. The control strategy for the cold extrusion equipment is determined to be to control the vibration device to vibrate at the target vibration frequency.

9. The automated process control method for cold extrusion according to claim 1, characterized in that, When abnormal causes include product skewing and jamming, determine the control strategy for the cold extrusion equipment, including: Generate an emergency stop command; The control strategy for the cold extrusion equipment is determined to be to send an emergency stop command to the control device of the cold extrusion equipment, and to store the displacement distance and ejection force of the ejection mechanism in the cold extrusion equipment into a non-volatile memory, so that maintenance personnel can analyze the displacement distance and ejection force of the ejection mechanism in the cold extrusion equipment.

10. An automated process control system for a cold extrusion process, characterized in that, include: Acquisition device and processing device; A data acquisition device is used to acquire the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism in the cold extrusion equipment in response to the ejection signal of the cold extrusion equipment. The ejection mechanism is used to eject the product from the extrusion die of the cold extrusion equipment; A processing device for determining the surface degradation index of an extrusion die based on the displacement distance of the ejection mechanism and the ejection force of the ejection mechanism; The processing device is also used to determine whether there is an ejection abnormality in the ejection mechanism and the cause of the abnormality when the surface degradation index of the extrusion die is less than the preset surface degradation index threshold, based on the displacement distance and ejection force of the ejection mechanism; the cause of the abnormality includes product skew and jamming, insufficient pushing distance or pushing delay. The processing device is also used to determine the control strategy of the cold extrusion equipment based on the cause of the abnormality when there is an abnormality in the ejection mechanism.

Images