Processing control system applied to blister sheet roll pasting work station

Abstract

The application discloses a processing control system applied to a blister sheet roll pasting work station, mainly relates to a blister sheet roll pasting control technical field, and specifically is through a processing control platform to overall plan and coordinate an environment sensing unit, a tension feedback cooperative control unit and a sticky roller critical state identification unit, realizes intelligent and accurate control of a whole process of the blister sheet roll pasting, in the application, the temperature, humidity and electrostatic parameters of the pasting work station are comprehensively monitored and dynamically regulated through the environment sensing unit, sheet material defects caused by environmental factors are avoided, the unwinding speed and the tension change are monitored in real time through the tension feedback cooperative control unit, the correlation model of the speed and the tension is established, the cooperative compensation and accurate regulation of the tension are realized, and the sheet material stretching deformation or shaking is avoided.

Description

Technical Field

[0001] This invention relates to the field of thermoforming sheet roll bonding control technology, specifically to a processing control system applied to a thermoforming sheet roll bonding workstation. Background Technology

[0002] The current thermoforming industry is accelerating its iteration towards high speed, intelligence, and high precision. As the core process in thermoforming product production, the quality of thermoforming sheet rolling and lamination directly determines the pass rate and performance stability of the final product. Environmental conditions, tension control, and roller adhesion during the lamination process are key factors affecting the lamination quality.

[0003] Currently, conventional thermoforming sheet roll bonding workstations mostly employ single-parameter control modes, lacking multi-unit collaborative linkage mechanisms, resulting in numerous technical bottlenecks, specifically as follows: On the one hand, there is an urgent need to address the incomplete environmental monitoring and lack of targeted electrostatic control in conventional systems. Traditional technologies simply monitor temperature and humidity without establishing a mechanism for predicting the electrostatic life cycle and fluctuation trends, thus failing to accurately identify electrostatic anomalies and their causes. On the other hand, traditional tension control lacks feedforward compensation, cannot predict sudden tension changes, and has difficulty in identifying stress layer boundary areas. Traditional technology only uses single feedback control and has not established a correlation model between speed and tension, so it cannot adapt to the tension requirements of different stress layers. Meanwhile, there are also shortcomings in the identification accuracy of sticky rollers in the blister pack bonding workstation, such as low response lag and poor intervention effect. It relies solely on manual or single signal monitoring, fixed thresholds are prone to drift, and there is a lack of hierarchical intervention and closed-loop feedback process.

[0004] To address the aforementioned technical shortcomings, a processing control system for a thermoforming sheet roll bonding workstation is proposed. This system utilizes an environmental sensing unit to achieve comprehensive monitoring of temperature, humidity, and static electricity, establishing a mapping relationship between environmental parameters and static electricity changes. This enables accurate identification of static electricity anomalies, trend prediction, and targeted adjustment of environmental parameters, avoiding sheet defects caused by static electricity. Furthermore, a speed difference-tension change rate correlation model is established to predict tension abrupt changes and accurately identify stress layer boundary zones. An orthogonal constraint discrimination system and dynamic threshold mechanism are constructed to accurately identify the critical state of roller sticking, implement graded intervention, and adaptively optimize intervention parameters, preventing equipment downtime and sheet scrap caused by roller sticking failure. Summary of the Invention

[0005] The purpose of this invention is to solve the problems mentioned above by proposing a processing control system for a blister pack lamination workstation.

[0006] The objective of this invention can be achieved through the following technical solution: a processing control system applied to a blister pack lamination workstation, comprising a processing control platform, wherein the processing control platform is connected via communication: The environmental sensing unit monitors the environment of the blister pack lamination workstation and controls the processing based on the real-time environmental sensing during the lamination stage. The tension feedback and collaborative control unit performs tension feedback and collaborative control on the thermoforming sheet roll bonding process; The sticking roller critical state identification unit performs sticking roller critical state identification.

[0007] Furthermore, the environmental sensing and monitoring process within the environmental sensing unit is as follows: Extract ambient temperature and humidity, and based on the set extraction time and the historical operating cycle of the production line, obtain the ambient temperature range and ambient humidity range for each stage. Within various ranges of ambient temperature and humidity, electrostatic charge is monitored and collected. Simultaneously, based on the time periods of the corresponding ranges, the generation and dissipation times of electrostatic charge are extracted to establish the electrostatic life cycle. Based on the historical operating cycle of the bonding production line, the environmental electrostatic charge at the moment when the bonding process is completed without defects is obtained, and the electrostatic charge range is derived from the extracted environmental electrostatic charge.

[0008] Furthermore, within the current time period, under the fluctuating progress of environmental parameters around the bonding production line, the amount of environmental electrostatic charge at each moment during the bonding stage is obtained. If the amount of environmental electrostatic charge is not within the range of electrostatic charge, then numerical analysis of the amount of electrostatic charge is performed. If the electrostatic charge exceeds the peak value of the electrostatic charge range, a sheet peeling risk signal is generated; if the electrostatic charge does not exceed the valley value of the electrostatic charge range, a sheet adsorption risk signal is generated. If the amount of static charge in the environment is within the static charge range at the corresponding moment during the bonding stage, then a trend analysis of the static charge is performed.

[0009] Furthermore, the normal state of electrostatic charge is marked as the normal state of charge, and the electrostatic fluctuation trend of the current bonding production line area is established based on the collected data of electrostatic charge value fluctuations. The electrostatic fluctuation trend is then analyzed. Starting from the normal charging moment, if the static charge value within the static fluctuation trend continues to fluctuate far from the static charge range, or if the static charge in the production line area within the static fluctuation trend is not within the static charge range and continues to increase at intervals when it recovers to the range, a trend risk signal is generated and sent to the processing control platform. If the static charge value does not fluctuate continuously away from the static charge range within the static fluctuation trend, and the static charge in the production line area within the static fluctuation trend is within the static charge range or recovers to the range at intervals lower than the interval threshold, then a trend stabilization signal is generated and sent to the processing control platform.

[0010] Furthermore, after receiving the trend risk signal, the processing control platform obtains the numerical fluctuation trend of environmental parameters within the historical operating cycle, and establishes the mapping relationship between the numerical fluctuation trend of environmental parameters and the trend of electrostatic charge based on the electrostatic charge trend of the bonding production line area collected at the same time, and by comparing the data collected in a distributed time period. When a trend risk signal is generated, the trend of environmental parameter value fluctuation is obtained based on the current electrostatic charge trend mapping relationship, and used as the direction of environmental control. The cause data of the environmental parameter value fluctuation trend is set as the environmental control object.

[0011] Furthermore, the tension feedback collaborative control process in the tension feedback collaborative control unit is as follows: The unwinding angular velocity and real-time tension are monitored in real time. The difference between the unwinding line speed and the main traction line speed is calculated as a speed feedforward signal. At the same time, the current tension is received as a feedback signal. A speed difference-tension change rate correlation model is established. Based on the speed difference-tension change rate correlation model, the floating span of the tension change rate when the speed difference exceeds the set range is obtained. When the real-time execution is performed, if the speed difference fluctuates and exceeds the set range, the actual floating span of the tension change rate is extracted, and the change trajectory of the thermoforming sheet roll is determined based on the tension change rate. If the change trajectory exceeds the correction area set by the bonding equipment, it is inferred that there is a bonding risk in the blister pack roll, a bonding deviation signal is generated and sent to the processing control platform. If the speed difference does not fluctuate and the actual floating span of the tension change rate exceeds the set floating span threshold, it is inferred that the blister pack roll has entered the boundary area of ​​different stress layers, a stress layer change signal is generated and sent to the processing control platform.

[0012] Furthermore, the specific operation process of the sticky roller critical state identification unit is as follows: Acquire the raw waveform signal collected by the acoustic emission sensor and the temperature sequence collected by the temperature sensor; The first feature mining is performed on the original waveform signal to obtain the zero crossover rate time series that characterizes the drastic change in pulse interval. The second feature mining is performed on the original waveform signal to obtain the short-time fluctuation entropy series that characterizes the drastic change in energy envelope shape. The temperature sequence is differentially processed to obtain the temperature change rate sequence; Based on the logic and constraints between the zero crossover rate time series, the short-time fluctuation entropy series, and the temperature change rate series, a true critical state flag is generated. Based on the historical trend of the zero crossover rate time series, the energy detection threshold at the current moment is dynamically generated. The current energy envelope amplitude, the current value of the zero crossover rate time series, the current value of the short-time fluctuation entropy series, and the energy detection threshold are mapped into a three-dimensional intermediate state coordinate vector. Based on the distance relationship between the three-dimensional intermediate state coordinate vector and the preset centroids of normal and failed clusters, a continuous risk index is calculated. Based on the risk index and the true critical state indicator, a graded intervention instruction is generated.

[0013] Furthermore, the specific steps of performing the first feature mining on the original waveform signal to obtain the zero crossover rate time series include: The original waveform signal is bandpass filtered to retain the frequency band from 15kHz to 40kHz, resulting in a filtered signal. The energy envelope of the filtered signal is extracted using the short-time energy detection method, and the arrival time of each pulse is identified based on an adaptive threshold of background noise. The time interval between the arrival times of adjacent pulses is calculated to obtain a pulse interval sequence. The pulse interval sequence is then subjected to first-order difference to obtain a difference sequence. The number of zero-crossings in the difference sequence is counted within a 1-second sliding time window, and the zero-crossing rate is obtained by dividing the number of zero-crossings by 1 second.

[0014] Furthermore, the step of dynamically generating the energy detection threshold at the current moment based on the historical trend of the zero crossover rate time series specifically includes: Calculate the mean of the zero crossover rate time series over the most recent 3 seconds and its first derivative over the most recent 0.6 seconds. Based on the relative position of the mean with respect to the minimum and maximum zero crossover rates under normal operating conditions, calculate the first adjustment factor. The first adjustment factor is negatively correlated with the mean. Based on the negative part of the first derivative, the second adjustment factor is calculated using an exponential function. The second adjustment factor has a positive exponential relationship with the magnitude of the negative value of the first derivative. The benchmark threshold is multiplied by the product of the first and second adjustment factors to obtain the dynamic energy detection threshold.

[0015] Furthermore, the specific steps for generating tiered intervention instructions based on the risk index and the true critical state indicator include: When the risk index is below 0.2, a shutdown alarm command is generated; When the risk index is greater than or equal to 0.2 and less than 0.4, and the true critical state flag is true, an instantaneous idle intervention command with a lifting stroke of 0.5 mm is generated; When the risk index is greater than or equal to 0.4 and less than 0.7, and the true critical state flag is true, a release agent spraying intervention instruction is generated that is linearly negatively correlated with the spraying time and the risk index.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. By employing a distributed point-based acquisition method to capture the static electricity distribution in the production line area and establish the static electricity lifecycle, it can accurately identify abnormalities and fluctuations in static charge, predict risks such as sheet peeling and adsorption deformation in advance, and avoid sheet scrapping due to static interference and environmental parameter fluctuations. By establishing a mapping relationship between environmental parameter fluctuations and static electricity changes, it achieves targeted and forward-looking environmental control, and can quickly determine the control direction and control object based on abnormal static electricity trends, thereby improving environmental control efficiency. At the same time, it adapts to the environmental requirements of different materials and different bonding stages, and optimizes the parameter range by combining historical defect-free data, significantly improving bonding accuracy and work efficiency, and reducing the impact of environmental factors on bonding quality, thus solving the core pain points of traditional technologies such as incomplete environmental monitoring and lack of targeted static electricity control.

[0017] 2. It can detect fluctuations in the difference between the unwinding speed and the main traction speed in advance, predict the trend of tension changes, and avoid problems such as sheet plastic deformation and vibration caused by sudden tension changes. This ensures the uniformity of subsequent heating and the accuracy of forming. It can also accurately identify the boundary area of ​​different stress layers of the thermoforming sheet roll, adjust the unwinding speed accordingly, adapt to changes in stress layers, and reduce unnecessary speed adjustments and sheet waste. When a bonding deviation occurs, a correction mechanism is quickly activated. If correction is not possible, abnormal sheets are promptly cut off to avoid continuous scrapping of subsequent sheets and improve material utilization. At the same time, it is compatible with thermoforming sheet rolls of different batches and thicknesses, has strong versatility, and a fast response speed. It can meet the needs of high-speed bonding scenarios, solve the technical problems of traditional tension control such as lack of feedforward compensation, difficulty in stress layer identification, and lag in response, and ensure the continuity and stability of the bonding process.

[0018] 3. Combining the multi-dimensional characteristics of AE signals and temperature change rate signals, it can distinguish effective signals from external interference, avoiding misjudgment and missed judgment of the critical state of sticking to the roller. The recognition accuracy is much higher than that of traditional single signal monitoring methods. Through a dynamic threshold generation mechanism, it adapts to the adhesion characteristics of different materials and batches of sheets to solve the recognition failure problem caused by fixed threshold drift. A graded intervention mechanism is established to perform differentiated intervention actions according to the degree of sticking risk (deep critical state, shallow critical state), which avoids production interruption caused by excessive intervention and ensures the intervention effect, effectively curbing the development of sticking to the roller and reducing losses such as sheet tearing and equipment downtime. Through closed-loop feedback modulation of intervention parameters, the intervention effect is continuously optimized. At the same time, historical data is updated to calibrate the recognition model, improving the stability and accuracy of long-term use. This addresses the technical problems of lagging sticking roller recognition and poor intervention effect in the blister sheet roll bonding workstation, ensuring the continuous and efficient operation of the workstation. Attached Figure Description

[0019] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0020] Figure 1 This is a system principle block diagram of the present invention; Figure 2 This is a flowchart of the environmental sensing unit method in this invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] In existing technologies, processing equipment relies on an integrated roll-to-roll processing model, comprising four core processes: surface transparent film processing, bottom white shell processing, double-layer roll-to-roll lamination, and finished single-piece cutting. Each process is interconnected and seamlessly integrated to complete the entire processing flow from raw material rolls to finished single pieces. The specific process logic is as follows: First, the transparent film processing process is carried out. The equipment performs precise cutting and positioning hole processing on the transparent film roll in sequence, and then recycles it in the form of a whole roll to complete the pretreatment of the product's surface structure. Next, the bottom white shell processing process is carried out. The white shell roll is first softened by constant temperature heating, and then the required structure is processed by compression molding. Matching positioning holes are processed simultaneously. After processing, it is also recycled in the whole roll to lay the foundation for the product's bottom support. The transparent film rolls and white shell rolls processed in the above two processes are then fed into the double-layer roll lamination process simultaneously. After precise alignment and heating activation, they are tightly laminated and then recycled as a whole roll semi-finished product after cooling and shaping. Finally, the finished product is cut into individual pieces. The equipment cuts the composite roll of semi-finished products to length and trims the edges, ultimately separating them into individual finished products and completing the entire processing flow.

[0024] Please see Figures 1-2As shown, in response to some technical defects in the above-mentioned whole process processing, a processing control system for a blister sheet roll bonding workstation is proposed, including a processing control platform, which is communicatively connected to an environmental sensing unit, a tension feedback collaborative control unit, and a sticking roller critical state identification unit. The processing control platform generates environmental sensing signals and sends them to the environmental sensing unit; After receiving the environmental sensing signal, the environmental sensing unit monitors the environment of the blister sheet roll bonding workstation and performs processing control based on the environmental sensing during the real-time bonding stage to ensure that the bonding efficiency meets the set requirements. Environmental parameters are extracted from the bonding production line within the workstation, namely, the ambient temperature and humidity. Based on the set extraction time and the historical operating cycle of the bonding production line, the range of ambient temperature and humidity for each stage are obtained. The static electricity data of the production line area is collected through a distributed point-to-point acquisition method. The static charge is monitored and collected within various ranges of ambient temperature and humidity. At the same time, the generation and dissipation times of static charge are extracted according to the time period of the corresponding range to establish the static electricity life cycle. The amount of environmental electrostatic charge at the moment of completion of defect-free bonding of the bonding production line is obtained based on the historical operating cycle of the bonding production line, and the range of electrostatic charge is obtained based on the amount of environmental electrostatic charge at the extraction time. During the current period, under the fluctuating progress of environmental parameters around the bonding production line, the amount of environmental electrostatic charge at each moment during the bonding stage is obtained. If the amount of environmental electrostatic charge at the corresponding moment during the bonding stage is not within the range of electrostatic charge, it is inferred that the amount of electrostatic charge at the current moment is abnormal, and numerical analysis of the amount of electrostatic charge is performed. If the electrostatic charge exceeds the peak value of the electrostatic charge range, a sheet peeling risk signal is generated and sent to the processing control platform. After receiving the sheet peeling risk signal, the processing control platform controls the environmental parameters and monitors the sheet bonding progress in real time to avoid an increase in the amount of scrapped blister sheets. If the amount of electrostatic charge does not exceed the valley value of the electrostatic charge range, a sheet adsorption risk signal is generated and sent to the processing control platform. After receiving the sheet adsorption risk signal, the processing control platform controls the environmental parameters and monitors the sheet bonding progress in real time to avoid the blister sheet not being effectively adsorbed and causing unnecessary sheet deformation. If the amount of static charge in the environment is within the static charge range at the corresponding moment during the bonding stage, it is inferred that the amount of static charge at the current moment is normal, and a trend analysis of the amount of static charge is performed. Mark the time when the electrostatic charge is normal as the time when it is charged, and determine the location of the charge at the time when it is charged within the range of electrostatic charge. Collect the electrostatic charge values ​​at adjacent times, and establish the electrostatic fluctuation trend of the current bonding production line area based on the collection of electrostatic charge value fluctuations. The trend monitoring period covers the current time when it is charged. Analysis of electrostatic fluctuation trends: Starting from the normal charging moment, if the static charge value within the static fluctuation trend continues to fluctuate far from the static charge range, or if the static charge in the bonding production line area within the static fluctuation trend is not within the static charge range and continues to increase at intervals when it recovers to the range, it is inferred that the static fluctuation trend of the current bonding production line area is abnormal, a trend risk signal is generated and sent to the processing control platform. If the static charge value within the static fluctuation trend does not fluctuate continuously away from the static charge range, and the static charge in the bonding production line area within the static fluctuation trend is within the static charge range or recovers to the range at intervals lower than the interval threshold, then it is inferred that the current static fluctuation trend in the bonding production line area is normal, a trend stabilization signal is generated and sent to the processing control platform. After receiving the trend risk signal, the processing control platform obtains the trend of environmental parameter values ​​within the range of historical operating cycles, and establishes the mapping relationship between the trend of environmental parameter values ​​and the trend of electrostatic charge based on the trend of electrostatic charge in the production line area collected at the same time. When a trend risk signal is generated, the trend of environmental parameter value fluctuation is obtained based on the current electrostatic charge trend mapping relationship, and used as the direction of environmental control. The cause data of the environmental parameter value fluctuation trend is set as the environmental control object. After completing environmental perception, the processing control platform generates a tension feedback collaborative control signal and sends it to the tension feedback collaborative control unit; After receiving the tension feedback collaborative control signal, the tension feedback collaborative control unit performs tension feedback collaborative control on the thermoforming sheet roll bonding process to avoid sudden tension changes that could cause the sheet to be unexpectedly stretched (plastic deformation) or vibrate at the transition point, affecting the uniformity of subsequent heating and the forming accuracy. A high-precision encoder is installed on the unwinding motor to monitor the unwinding angular velocity in real time; Install a tension sensor on the feed guide roller to monitor the real-time tension; The difference between the "unwinding speed" (calculated from the angular velocity and the winding diameter) and the main traction speed is calculated as the "speed feedforward signal," while the current tension is received as the "feedback signal." Establish a correlation model between "velocity difference" and "tension change rate"; Based on the speed difference-tension change rate correlation model, the floating span of the tension change rate when the speed difference exceeds the set range is obtained. During real-time lamination, if the speed difference fluctuates and exceeds the set range, the actual floating span of the tension change rate is extracted, and the change trajectory of the blister pack roll is determined based on the tension change rate. If the change trajectory exceeds the correction area set by the lamination equipment, the lamination risk of the blister pack roll is inferred, a lamination deviation signal is generated and sent to the processing control platform. After receiving the signal, the processing control platform immediately controls the unwinding speed of the blister pack roll and performs correction on the current blister pack roll. If the correction cannot be completed, the current part of the blister pack roll is cut off to avoid increasing the waste of subsequent blister pack rolls. If the speed difference does not fluctuate and the actual fluctuation range of the tension change rate exceeds the set fluctuation range threshold, it is inferred that the blister pack has entered the boundary area of ​​different stress layers, generating a stress layer change signal and sending it to the processing control platform. After receiving the signal, the processing control platform re-controls the unwinding speed of the blister pack and makes corrections based on the real-time fluctuation of the tension change rate. The processing control platform generates a sticky roller critical state identification signal and sends it to the sticky roller critical state identification unit. After receiving the sticky roller critical state identification signal, the sticky roller critical state identification unit performs sticky roller critical state identification. Phase 1: Perform raw signal acquisition and dual-channel feature mining; Step 1.1: Acquire the original AE signal and the temperature change rate signal; Input: The raw voltage signal output by a broadband AE sensor (sampling rate ≥ 1MHz) mounted on the bearing housing of the hot press roller, and the temperature sequence output by a non-contact infrared temperature sensor attached near the roller surface; Processing: Perform first-order difference on the temperature sequence and calculate the temperature change rate dT / dt every 0.5 seconds. The physical meaning of this signal is the trend of heat flux density change of the hot press roller. It is not directly mechanically coupled with the adhesion process, but it is physically related (intensified adhesion will lead to frictional heat generation, causing dT / dt to drift in the positive direction). Output: Original AE waveform sequence S(t) and temperature change rate sequence .

[0025] Step 1.2: Extract the "zero crossover rate of the pulse repetition frequency" from the AE signal; Processing logic: First, bandpass filtering is applied to S(t) to retain the 15-40kHz frequency band (which has been experimentally proven to be the main energy region of adhesion-peeling events). Then, the pulse envelope is extracted using the short-time energy detection method, and an adaptive threshold (based on twice the noise floor of the previous second) is set to identify the arrival time t of each pulse. i Calculate the pulse interval sequence Δτ i =t {i+1} -t i ; Key feature: for Δτ i Performing a first-order difference, we obtain δ i =Δτ {i+1} -Δ τi Then, calculate δ within a unit time window (e.g., 1 second). i The number of zero crossovers is denoted as Z(t); Physical meaning: Z(t) characterizes the drastic change in the pulse interval. When adhesion points appear randomly and sparsely, Δτ i Large fluctuations and high Z(t) indicate that as adhesion points gradually become denser and tend towards periodic peeling, Δτ... i As it stabilizes, Z(t) decreases. When it enters the precursor to macroscopic adhesion, the peeling becomes extremely irregular, and Z(t) rises again. The minimum point of Z(t) corresponds to the "optimal intervention window of the critical state". Output: Zero crossover rate time series Z(t).

[0026] Step 1.3: Extract the "short-term fluctuation entropy of the energy envelope" from the AE signal; Processing logic: First, for the pulse envelope E(t) obtained in step 1.2 (with the silent segment removed), calculate the probability distribution of the energy value within a sliding time window (window length 0.2 seconds, step size 0.05 seconds). Then, divide the energy range within the window into 8 quantization levels, count the frequency p_k of each level, and then calculate the Shannon entropy. Finally, the rate of change of the entropy value between adjacent windows, dH / dt, is calculated, and its absolute value is taken as the short-time fluctuation entropy W(t). Physical meaning: A high W(t) indicates a drastic change in the energy envelope morphology (the adhesion-peeling process is unstable and in the critical state of evolution), while a low W(t) indicates a stable energy envelope morphology (either completely normal or completely stuck to the roller). Both the normal state and the failure state are represented by low values ​​in W(t). Therefore, W(t) alone cannot distinguish between the two ends and must be orthogonally constrained with Z(t). Output: Short-time fluctuation entropy sequence W(t).

[0027] Phase Two: Constructing the orthogonal constraint space and generating dynamic thresholds; Step 2.1: Establish logical constraints between physically unrelated data; Input: Z(t), W(t), (Rate of temperature change); Processing Logic: The three signals originate from different physical mechanisms (AE pulse timing, AE energy distribution, and thermodynamic processes) and are statistically independent. This scheme utilizes this characteristic to construct an orthogonal constraint discriminant: only when Z(t) is in the "low value region" (indicating that the pulse interval tends to be stable), W(t) is in the "high value region" (indicating that the energy form is still changing drastically), and... When the value is positive (temperature is rising, consistent with the thermal effect of increased adhesion), it is determined to have entered the "true critical state". If only Z(t) is low and W(t) is also low, it may be a normal steady state. If W(t) is high but Z(t) is also high, it may be due to external random vibration disturbance. Output: Boolean flag critical And the real-time numerical vectors of the three signals [Z, W, ... ].

[0028] Step 2.2: Dynamically modulate the energy threshold based on the historical trend of Z(t); Input: Z(t) is the historical sequence of the past 10 seconds, and the current energy envelope amplitude E(t); Processing logic: The reason why the traditional fixed threshold method fails is that the critical AE energy baseline drifts with the material batch. However, this invention uses the evolution law of Z(t) to reverse-calibrate the energy threshold and calculates the average value Z(t) over the most recent 3 seconds. avg Its first derivative dZ / dt, and based on the experimentally calibrated relationship, dynamically generate the energy threshold Th at the current moment. E(t) : ; Among them, Th {base} Z is the baseline threshold (obtained through offline calibration), α and β are adjustable coefficients, and Z {min} and Z {max} This represents the statistical extreme value of Z(t) under normal operating conditions; Physical meaning: When Z_avg decreases (the pulse interval tends to stabilize, and the critical state is approached), Th E(t) Automatic reduction makes the system more sensitive to energy. When dZ / dt is negative and its absolute value is large (Z decreases rapidly), the exponential term further lowers the threshold, achieving forward-looking sensitivity. Output: Dynamic energy threshold Th E(t) .

[0029] Phase 3: Intermediate state mapping and risk index generation; Step 3.1: Map the feature vectors to three-dimensional intermediate state coordinates; Input: E(t) (current energy envelope amplitude), Z(t), W(t); Processing logic: Do not directly compare E(t) with Th E(t) Instead, it constructs a three-dimensional intermediate state vector: ; Z {min} W represents the theoretically achievable minimum value of Z(t) (corresponding to the optimal intervention window). {max} ε represents the upper limit of normal fluctuation entropy, ε is a small amount to prevent division by zero, the first dimension represents the normalized energy overshoot, the second dimension represents the degree to which the pulse repetition frequency approaches the ideal intervention window (the larger the value, the closer), and the third dimension represents the relative intensity of energy fluctuation entropy. Output: A three-dimensional vector V(t).

[0030] Step 3.2: Calculate the "Adhesion Risk Index"; Processing logic: The centroids of the two clusters are obtained in advance through offline experiments: the centroid C of the normal cluster. normal =(0.2,0.3,0.2) and the centroid C of the failure cluster failure =(1.2,0.1,0.8), and calculate the Euclidean distance d from the current V(t) to the two centroids. n and d f Then define the risk index: When R(t) is close to 0, it indicates that the cluster is close to failure (sticky roller), and when R(t) is close to 1, it indicates that the cluster is close to normal. The intermediate value (0.3~0.7) corresponds to the critical state. Output: Continuous risk index R(t), with a value range of [0,1].

[0031] Phase Four: Tiered Intervention and Closed-Loop Feedback Modulation; Step 4.1: Hierarchical decision-making based on risk index and orthogonal indicators; Input: R(t), Flag_critical; Processing logic: If R(t) < 0.2: it is determined that roller sticking has occurred, and the machine is stopped and an alarm is triggered immediately; If 0.2 ≤ R(t) < 0.4 and Flag critical ==True: Determined to be in "deep critical state", instantaneous idle intervention is performed: within 0.1 seconds, the hot press roller is axially lifted by 0.5mm by the servo motor and immediately reset, using inertial shear force to tear off the microscopic adhesion points; If 0.4 ≤ R(t) < 0.7 and Flag critical ==True: Determined to be in a "shallow critical state", perform micro-dose release agent spraying: Calculate the spraying time (T) linearly based on the R(t) value. spray =T max *(0.7-R(t)) / 0.3); Other situations: No intervention.

[0032] Step 4.2: Closed-loop feedback modulation of intervention parameters; Input: The trajectory of R(t) change within 5 seconds after the intervention, and the type of intervention; Processing logic: After each intervention, calculate the rate of decrease v of the risk index. R =(R {before} -R {after} ) / 5, if v R If the value is lower than the preset expected value (e.g., 0.05 / s), it indicates that the current intervention intensity is insufficient, and the system will automatically adjust the parameters for the next intervention of the same type. For spraying intervention: dynamically increase T max New value = old value × (1 + (0.1 - v) R () / 0.1), with an upper limit of 1.5 times the old value; For idling intervention: dynamically increase the lifting stroke, new stroke = old stroke × (1 + (0.15 - v)). R ) / 0.15), upper limit 1.0mm; At the same time, the v of this intervention R The Z(t) values ​​before intervention are packaged and stored in a historical database for periodic recalculation of the cluster centroid C. normal and C failure (Achieving long-term self-adaptation); Output: Updated intervention parameters, applied to the next round of critical state processing.

[0033] In summary, this invention achieves coordinated linkage and dynamic adaptation of the three major control links by integrating the environmental perception unit, tension feedback collaborative control unit, and sticky roller critical state identification unit through a processing control platform. It solves the technical bottlenecks of conventional control systems, such as low control accuracy, slow response, serious waste, and poor stability, from the three core dimensions of environment, tension, and sticky roller.

[0034] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A processing control system applied to a blister pack lamination workstation, characterized in that, Includes a machining control platform, wherein the machining control platform has the following communication connections: The environmental sensing unit monitors the environment of the blister pack laminating workstation and performs processing control based on the real-time environmental sensing during the lamination stage. The environmental sensing monitoring process within the environmental sensing unit is as follows: Extract ambient temperature and humidity, and based on the set extraction time and the historical operating cycle of the production line, obtain the ambient temperature range and ambient humidity range for each stage. Within the range of ambient temperature and humidity, electrostatic charge is monitored and collected. Based on the time period of the corresponding range, the generation and dissipation times of electrostatic charge are extracted to establish the electrostatic life cycle. The environmental electrostatic charge at the moment when the bonding production line completes defect-free bonding is obtained based on the historical operation cycle of the bonding production line, and the electrostatic charge range is obtained based on the environmental electrostatic charge at the extracted time. Within the current time period, under the fluctuating progress of environmental parameters around the bonding production line, the amount of environmental electrostatic charge at each moment during the bonding stage is obtained. When the amount of environmental electrostatic charge is not within the range of electrostatic charge, numerical analysis of the amount of electrostatic charge is performed. If the amount of electrostatic charge exceeds the peak value of the electrostatic charge range, a risk signal of sheet peeling will be generated. If the amount of electrostatic charge does not exceed the valley value of the electrostatic charge range, a risk signal of sheet adsorption will be generated; When the amount of static charge in the environment is within the static charge range at the corresponding moment during the bonding stage, a trend analysis of the static charge is performed. Mark the time when the electrostatic charge is normal as the time when it is charged, and establish the electrostatic fluctuation trend of the current bonding production line area based on the data collected on the fluctuation of the electrostatic charge value. Analysis of electrostatic fluctuation trends: Starting from the normal charging moment, if the static charge value within the static fluctuation trend continues to fluctuate far from the static charge range, or if the static charge in the production line area within the static fluctuation trend is not within the static charge range and continues to increase at intervals when it recovers to the range, a trend risk signal is generated and sent to the processing control platform. If the static charge value does not fluctuate continuously away from the static charge range within the static fluctuation trend, and the static charge in the production line area within the static fluctuation trend is within the static charge range or recovers to the range at intervals lower than the interval threshold, then a trend stabilization signal is generated and sent to the processing control platform. The tension feedback collaborative control unit performs tension feedback collaborative control on the thermoforming sheet roll bonding process; the tension feedback collaborative control process in the tension feedback collaborative control unit is as follows: Real-time monitoring of unwinding angular velocity; real-time monitoring of tension; calculation of the difference between unwinding line speed and main traction line speed as a speed feedforward signal; simultaneous reception of current tension as a feedback signal; establishment of a speed difference-tension change rate correlation model; Based on the velocity difference-tension change rate correlation model; When the speed difference exceeds the set range, the floating span of the tension change rate is obtained. During real-time bonding, if the speed difference fluctuates and exceeds the set range, the actual floating span of the tension change rate is extracted, and the change trajectory of the blister roll is determined based on the tension change rate. If the change trajectory exceeds the correction area set by the bonding equipment, a bonding risk of the blister roll is inferred, a bonding deviation signal is generated, and it is sent to the processing control platform. If the speed difference does not fluctuate, and the actual floating span of the tension change rate exceeds the set floating span threshold, it is inferred that the blister roll has entered the boundary area of ​​different stress layers, a stress layer change signal is generated, and it is sent to the processing control platform. The sticking roller critical state identification unit performs sticking roller critical state identification. The process of the sticking roller critical state identification unit is as follows: Acquire the raw waveform signal collected by the acoustic emission sensor and the temperature sequence collected by the temperature sensor; The first feature mining is performed on the original waveform signal to obtain a zero-crossing rate time series that characterizes the degree of drastic change in pulse interval; The original waveform signal is subjected to a second feature mining to obtain a short-time fluctuation entropy sequence that characterizes the degree of drastic change in the energy envelope shape; The temperature sequence is differentially processed to obtain the temperature change rate sequence; Based on the logic and constraints between the zero crossover rate time series, the short-time fluctuation entropy series, and the temperature change rate series, a true critical state indicator is generated; based on the historical trend of the zero crossover rate time series, the energy detection threshold for the current moment is dynamically generated. The current energy envelope amplitude, the current value of the zero crossover rate time series, the current value of the short-time fluctuation entropy series, and the energy detection threshold are mapped into a three-dimensional intermediate state coordinate vector; Based on the distance relationship between the three-dimensional intermediate state coordinate vector and the preset centroids of normal and failed clusters, a continuous risk index is calculated; based on the risk index and the true critical state indicator, a graded intervention instruction is generated.

2. The processing control system for a blister pack laminating workstation according to claim 1, characterized in that, After receiving the trend risk signal, the processing control platform obtains the trend of environmental parameter fluctuations within the range of historical operating cycles, and establishes the mapping relationship between the trend of environmental parameter fluctuations and the trend of electrostatic charge based on the trend of electrostatic charge in the production line area collected at the same time. When a trend risk signal is generated, the trend of environmental parameter value fluctuation is obtained based on the current electrostatic charge trend mapping relationship, and used as the direction of environmental control. The cause data of the environmental parameter value fluctuation trend is set as the environmental control object.

3. The processing control system for a blister pack laminating workstation according to claim 1, characterized in that, The specific steps for performing the first feature mining on the original waveform signal to obtain the zero crossover rate time series include: The original waveform signal is bandpass filtered to retain the frequency band from 15kHz to 40kHz, resulting in the filtered signal. The energy envelope of the filtered signal is extracted using the short-time energy detection method, and the arrival time of each pulse is identified based on the adaptive threshold of the background noise. The time interval between the arrival times of adjacent pulses is calculated to obtain the pulse interval sequence. The pulse interval sequence is first-order differencing is performed to obtain the difference sequence. The number of zero-crossings in the difference sequence is counted within a 1-second sliding time window, and the zero-crossing rate is obtained by dividing the number of zero-crossings by 1 second.

4. The processing control system for a blister pack laminating workstation according to claim 3, characterized in that, The steps for dynamically generating the energy detection threshold at the current moment based on the historical trend of the zero crossover rate time series specifically include: Calculate the mean of the zero crossover rate time series over the most recent 3 seconds and its first derivative over the most recent 0.6 seconds; calculate the first adjustment factor based on the relative position of the mean with respect to the minimum and maximum zero crossover rates under normal operating conditions. The first adjustment factor is negatively correlated with the mean. The second adjustment factor is calculated using an exponential function based on the negative part of the first derivative. The second adjustment factor has a positive exponential relationship with the magnitude of the negative value of the first derivative. The dynamic energy detection threshold is obtained by multiplying the baseline threshold by the product of the first and second adjustment factors.

5. The processing control system for a blister pack laminating workstation according to claim 4, characterized in that, The specific steps for generating tiered intervention instructions based on the risk index and the true critical state indicator include: When the risk index is below 0.2, a shutdown alarm command is generated; When the risk index is greater than or equal to 0.2 and less than 0.4, and the true critical state flag is true, an instantaneous idle intervention command with a lifting stroke of 0.5 mm is generated; When the risk index is greater than or equal to 0.4 and less than 0.7, and the true critical state flag is true, a release agent spraying intervention instruction is generated that is linearly negatively correlated with the spraying time and the risk index.

Images