A system for analyzing the entire process of die-cutting, waste removal, and stacking based on industrial big data.

Abstract

This invention relates to the field of industrial big data and intelligent manufacturing technology, specifically to a full-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data. The system includes: a data acquisition module for collecting multi-source heterogeneous operation data and flexible material status data from adjacent processes; a coupling degree calculation module for time-series comparison of upstream and downstream material flow throughput rates to extract phase differences and calculate cross-process rhythmic coupling degree; a dissipation rate calculation module for obtaining the dynamic buffer margin dissipation rate based on the integral of the difference between upstream output and downstream consumption and its derivative; a risk assessment module for weighted summation of coupling degree, dissipation rate, and processing delay time to obtain a cascade blockage risk index; and a status scheduling module for generating shutdown warnings, collaborative scheduling strategies, or maintenance operation marker data based on risk thresholds. This invention can achieve early identification of cascade blockages, closed-loop intervention, and pre-emptive suppression of full-line shutdowns.

Description

Technical Field

[0001] This invention relates to the field of industrial big data and intelligent manufacturing technology, specifically to a system for analyzing the entire process of die-cutting, waste removal, and stacking based on industrial big data. Background Technology

[0002] With the development of industrial big data, intelligent sensing and continuous manufacturing collaborative control technologies, cardboard packaging production lines consisting of adjacent processes such as die-cutting, waste removal and stacking have gradually evolved towards high speed, linkage and precision. In order to achieve stable operation of the whole line and early intervention for anomalies, the perception, analysis and scheduling of cross-process operation status in the production process has become particularly important. In continuous production scenarios involving die-cutting, waste removal, and stacking, the differences in operating mechanisms, cycle time characteristics, instantaneous loads, and material receiving methods among the die-cutting, waste removal, and stacking processes can easily lead to issues such as asynchronous upstream and downstream cycles, continuous depletion of buffer zone capacity, and gradual spread of localized delays. This is especially true for flexible materials like cardboard, where factors such as cardboard humidity, thickness, and die wear further affect waste removal resistance, material conveying stability, and end-stack continuity. This can result in physical mismatches in material flow that are not directly detected by sensors, even when the equipment control layer displays normal operation. Under current methods, relying solely on single equipment cycle time, local alarm signals, or static production information makes it difficult to accurately identify the true collaborative state between heterogeneous processes. Furthermore, it is challenging to combine buffer zone capacity, dynamic buffer zone changes, and processing delays to promptly assess whether localized anomalies will escalate into cascading blockages or even complete line shutdowns. Therefore, it is crucial to comprehensively collect and analyze multi-source heterogeneous operational data and flexible material status data throughout the entire process of die-cutting, waste removal, and stacking to obtain operational status results that reflect the rhythmic coupling relationship across processes, the dynamic buffer margin dissipation status, and the risk of cascading blockage. This will support the subsequent establishment of corresponding early warning and collaborative scheduling strategies by combining buffer capacity constraints and equipment control logic, and is essential for ensuring the safe and stable operation of continuous packaging production lines, suppressing the propagation of speed fluctuations, and improving overall production efficiency. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a system for analyzing the entire process of die-cutting, waste removal, and stacking based on industrial big data. Specifically, the technical solution of this invention includes: The data acquisition module is used to collect multi-source heterogeneous operation data and flexible material status data of heterogeneous processes; the heterogeneous processes are adjacent upstream and downstream processes in die-cutting, waste removal and stacking processes; the multi-source heterogeneous operation data includes upstream and downstream material flow throughput, upstream output, downstream consumption and processing delay time; The coupling degree calculation module is used to compare the throughput of upstream and downstream material flows to extract the time phase difference, and calculate the cross-process rhythm coupling degree between upstream and downstream processes based on the reciprocal of the time phase difference or a preset inverse proportional function value. The dissipation rate calculation module is used to calculate the dynamic buffer margin by integrating the difference between upstream output and downstream consumption within a specific time window, and to use the time derivative of the dynamic buffer margin as the dynamic buffer margin dissipation rate between upstream and downstream processes. The risk assessment module is used to calculate the cascade blockage risk index by weighted summation of the cross-process rhythm coupling degree, the dynamic buffer margin dissipation rate, and the processing delay time; The situational scheduling module is used to compare the cascading congestion risk index with preset minimum and maximum risk thresholds: if it is not less than the maximum risk threshold, a global shutdown warning data packet is generated; if it is not less than the minimum and less than the maximum risk threshold, an adaptive production cycle collaborative scheduling strategy data packet is generated and sent to the corresponding control node of the heterogeneous process that is connected to the system for execution within a preset situational collaborative response time window; if it is less than the minimum risk threshold, marker data to maintain the current operating state is generated.

[0004] Optionally, the multi-source heterogeneous operating data includes programmable logic controller data of the die-cutting machine, visual monitoring data of the waste removal area, and sensor data of the stacking area; wherein, the sensor data of the stacking area includes gravity sensor data and photoelectric sensor data.

[0005] Optionally, the flexible material status data includes paperboard humidity data, paperboard thickness data, and die-cutting die wear data.

[0006] Optionally, the system further includes a drag coefficient update module; The resistance coefficient update module is used to calculate the real-time waste removal resistance coefficient based on the paperboard humidity data, the paperboard thickness data, and the die-cutting die wear data; and to correct the downstream material flow throughput rate by multiplying the real-time waste removal resistance coefficient by the downstream material flow throughput rate.

[0007] Optionally, the coupling degree calculation module is used to quantitatively evaluate the coordination status of the operation frequency of upstream and downstream processes within a specific time window; the cross-process rhythm coupling degree is inversely proportional to the timing phase difference; if the absolute value of the difference between the cross-process rhythm coupling degree and the preset target value is not greater than the preset synchronization threshold, then marker data representing that the rhythms between processes have reached a synchronized state is generated; if the absolute value of the difference is greater than the preset synchronization threshold, then marker data representing that the rhythms between processes are not synchronized is generated.

[0008] Optionally, the data acquisition module is also used to acquire data on the remaining capacity of the buffer band; The risk assessment module combines the cross-process rhythm coupling degree, the dynamic buffer margin dissipation rate, the processing delay time, and the buffer band remaining capacity data to predict the probability and time threshold of a local delay evolving into a full-line shutdown event; the product of the probability and the time threshold is output as the cascade blockage risk index.

[0009] Optionally, the adaptive production cycle time collaborative scheduling strategy data package includes die-cutting machine speed adjustment instructions and waste removal belt speed adjustment instructions; The die-cutting machine is equipped with an independent die-cutting control node, and the actuator corresponding to the waste removal process is equipped with an independent waste removal control node; the situation scheduling module is used to send the die-cutting machine speed adjustment command to the control node corresponding to the die-cutting process for execution; and to send the waste removal belt speed adjustment command to the control node corresponding to the waste removal process for execution.

[0010] Optionally, the situational coordination response time window is the time period required from the time the system extracts the rhythm disorder features from the multi-source heterogeneous operating data to the time required to generate and distribute the adaptive production rhythm coordination scheduling strategy data packet.

[0011] Optionally, the situational scheduling module is further configured to compare the dynamic buffer margin dissipation rate with preset minimum and maximum dissipation thresholds: if it is not less than the maximum dissipation threshold, the speed adjustment intervention instruction is incorporated into the adaptive production cycle collaborative scheduling strategy data packet; if it is not less than the minimum and less than the maximum dissipation threshold, an early warning message is output; if it is less than the minimum dissipation threshold, marker data is generated to maintain the current buffer capacity state.

[0012] Optionally, the adaptive production cycle time collaborative scheduling strategy data package is used to dynamically smooth the speed fluctuation difference between upstream and downstream processes based on the cascaded congestion risk index, so as to suppress the step-by-step amplification of speed fluctuations during the propagation process along the production line.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. Existing technologies rely solely on single equipment cycle time or local alarm signals, making it difficult to accurately identify the true collaborative state between heterogeneous processes. This invention acquires multi-source heterogeneous operating data and flexible material status data through a data acquisition module, and calculates the real-time waste removal resistance coefficient based on the material status to correct the downstream material flow throughput. This mechanism effectively eliminates interference where the equipment control layer displays normal operation but the material flow on-site is already obstructed. Combined with cross-process rhythm coupling degree, dynamic buffer margin dissipation rate, and processing delay time, it can accurately quantify the cascade blockage risk index, transforming uncaptured physical mismatches into quantifiable and monitorable data features, and achieving a high-confidence perception of the entire process operation status. 2. To address the issue of local delays easily spreading and causing line-wide shutdowns in continuous production scenarios, this invention calculates dynamic buffer margins and their dissipation rates, and incorporates buffer band remaining capacity data to accurately predict the probability and time threshold of local delays evolving into line-wide shutdown events. Based on the risk index and dissipation threshold, the situational scheduling module generates and distributes adaptive production cycle time coordination scheduling strategy data packets within the situational coordination response time window. This closed-loop scheduling mechanism can intervene before the buffer capacity is exhausted and dynamically smooth the speed fluctuation differences between upstream and downstream processes, thereby effectively absorbing local shocks, suppressing the gradual amplification of speed fluctuations propagating along the production line, and improving the matching degree of physical cycle times between multiple processes on the production line and the stability of continuous material flow. Attached Figure Description

[0014] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0016] like Figure 1 As shown, the die-cutting, waste removal, and stacking full-process operation status analysis system based on industrial big data includes: The data acquisition module is used to collect multi-source heterogeneous operation data and flexible material status data of heterogeneous processes; heterogeneous processes are adjacent upstream and downstream processes in die-cutting, waste removal and stacking processes; multi-source heterogeneous operation data includes upstream and downstream material flow throughput, upstream output, downstream consumption and processing delay time; The coupling degree calculation module is used to compare the throughput of upstream and downstream material flows to extract the time phase difference, and calculate the cross-process rhythm coupling degree between upstream and downstream processes based on the reciprocal of the time phase difference or a preset inverse proportional function value. The dissipation rate calculation module is used to calculate the dynamic buffer margin by integrating the difference between upstream output and downstream consumption within a specific time window, and to use the time derivative of the dynamic buffer margin as the dynamic buffer margin dissipation rate between upstream and downstream processes. The risk assessment module is used to calculate the cascading blockage risk index by weighted summation of cross-process rhythm coupling degree, dynamic buffer margin dissipation rate and processing delay time. The situational scheduling module compares the cascading congestion risk index with preset minimum and maximum risk thresholds: if it is not less than the maximum risk threshold, a global shutdown warning data packet is generated; if it is not less than the minimum and less than the maximum risk threshold, an adaptive production cycle collaborative scheduling strategy data packet is generated and sent to the corresponding control nodes of heterogeneous processes connected to the system for execution within a preset situational collaborative response time window; if it is less than the minimum risk threshold, marker data to maintain the current operating state is generated.

[0017] This embodiment provides a mechanism for analyzing the operational status of the entire process of die-cutting, waste removal, and stacking based on industrial big data. Specifically, the application scenario is set as the continuous production of a batch of pharmaceutical outer packaging cardboard on the same paper box packaging production line. The production line includes a die-cutting machine, a waste removal unit, and a stacking unit in sequence along the material conveying direction, and a buffer zone of limited length is set between adjacent equipment. Since the die-cutting process is a periodic punching action, the waste removal process is a parallel action of waste edge peeling and conveying, and the stacking process is an action of cumulative receiving by sheet, the three have inherent differences in mechanical inertia, action frequency, and instantaneous load, which can easily lead to a physical mismatch state where the upstream material is released instantaneously but the downstream process's receiving capacity lags behind. Specifically, the data acquisition module continuously collects the operational status of adjacent processes. For the pair of adjacent processes from die-cutting to waste removal, it collects the output cycle time on the die-cutting side, the receiving cycle time on the waste removal side, the cumulative output on the die-cutting side, the cumulative consumption on the waste removal side, and the processing delay time on either side during conveying, waiting, and paper jam handling. For the pair of adjacent processes from waste removal to stacking, the acquisition method is the same. The throughput in the multi-source heterogeneous operating data does not only represent the nominal transmission rate, but in industrial physics, it reflects the release and acceptance capacity of the material flow within a certain time slice. The difference between output and consumption reflects whether the buffer is being filled or emptied. The processing delay time reflects whether the equipment has begun to experience local lag. The coupling degree calculation module performs a time-series comparison of the upstream and downstream throughput. Its purpose is not to require the two values ​​to be completely equal, but to identify whether the two beat waveforms change in the same direction, whether they reach the peak synchronously, and whether there is a lag between the peaks that exceeds the preset allowable time difference. If the upstream throughput has increased in advance while the downstream throughput is still lagging behind and remains low, it means that the sheet material released from the upstream is accumulating in the buffer zone. At this time, the extracted phase difference increases and the cross-process rhythm coupling degree decreases. Conversely, if the upstream and downstream cycle changes are basically in phase, it indicates that the entire line is in a stable and coordinated state. The inverse or inverse proportional function value of the phase difference is used as the coupling degree because the greater the phase deviation, the more it represents the beginning of rhythm mismatch in the physical actions of heterogeneous equipment, and the more it should be identified as an instability precursor. The preset inverse proportional function takes the timing phase difference as the independent variable and is configured with a scaling constant based on the nominal cycle time of the equipment under historical stable operating conditions. This is used to uniformly map the throughput fluctuations of different dimensions to the normalized interval that characterizes the coordinated state. The dissipation rate calculation module observes the difference between the cumulative output of the upstream and the cumulative consumption of the downstream within a fixed time window. The length of the fixed time window is preset to be greater than the cycle of a single die-cutting mechanical action and less than the passive emptying time of the buffer band under the maximum designed throughput. In the discrete sampling scenario of industrial data, the difference integration operation is specifically manifested in calculating the difference between the upstream output and the downstream consumption for each discrete sampling cycle within a specific time window and summing them up to reflect the dynamic buffer margin of the physical buffer band's newly added or reduced accumulation during that time period. The time derivative is further obtained by calculating the difference between the dynamic buffer margin of the current time window and the dynamic buffer margin of the previous adjacent time window, and dividing it by the time window step size, i.e., by performing a finite difference approximation calculation. If the derivative is positive and continues to expand, it indicates that the available physical capacity within the buffer zone is being occupied at a rate exceeding the preset dissipation threshold. If the derivative gradually shrinks or turns negative, it indicates that the downstream has compensated for the previous cycle delay and the buffer pressure is being released. The dynamic buffer margin dissipation rate reflects the dynamic consumption rate of the buffer safety boundary, rather than the static inventory value, so it is more suitable for early warning. For short-buffer continuous production lines such as die-cutting, waste removal, and stacking, the core focus of the system is not the small absolute value of the buffer margin at a certain static moment, but the process in which the margin is approaching the lower limit of the physical capacity at a high dissipation rate. The risk assessment module performs a weighted summation of cross-process rhythm coupling, dynamic buffer margin dissipation rate, and processing delay time to conduct a comprehensive assessment and calculate a cascading blockage risk index. In the specific weighting logic, since lower coupling indicates more severe mismatch, while higher dissipation rate and delay time indicate faster accumulation, a negative weight is assigned to cross-process rhythm coupling, and a positive weight is assigned to dynamic buffer margin dissipation rate and processing delay time. This index quantitatively represents the probability of a local cycle fluctuation exceeding the physical carrying capacity limit of the buffer zone, leading to a complete line stoppage. When coupling decreases but the buffer is still sufficient, the system tends to consider it a recoverable disturbance. When coupling decreases, the buffer dissipates rapidly, and the delay time of a certain device continues to lengthen, the weighted summation result exceeds the preset alarm threshold, and the system considers that the blockage chain has the conditions to spread to the upstream and downstream. The situational control module executes different responses based on risk levels. If the risk reaches the highest threshold, a global shutdown warning data packet is generated for the central control or field control nodes to organize a safe shutdown, preventing continuous compression of cardboard in the die-cutting area, waste removal area, and stacking area. If the risk is in the middle range, an adaptive production cycle collaborative scheduling strategy data packet is generated and sent to the corresponding control nodes within the situational collaborative response time window, so that the upstream appropriately slows down the release, the downstream appropriately increases the acceptance, or both are fine-tuned simultaneously. If the risk is below the lowest threshold, a marker data to maintain the current operating state is generated to avoid unnecessary intervention in stable operating conditions. In the anomaly handling mechanism, if a sensor source is briefly interrupted within a certain time window, such as a temporary lack of throughput data on the waste removal side, the system can temporarily maintain a short-term judgment based on the trend value of adjacent valid time slices or the equipment's most recent stable state value. However, it will not directly trigger the highest-level shutdown action. Instead, it will mark the current risk output as a low-confidence result and wait for the next complete collection cycle for verification. If data is incomplete for several consecutive cycles, automatic scheduling will be suspended, and only early warning prompts and manual verification entry points will be retained to avoid miscontrol due to data gaps. If it is detected that the upstream has stopped while the downstream is idling, the coupling degree assessment will automatically switch to the shutdown state and will no longer interpret the phase difference according to the normal production cycle. During the continuous production of the same batch of cardboard boxes, the die-cutting machine resumed high-speed operation after a blade change, while the waste removal unit experienced intermittent pauses due to poor waste edge separation. At this time, the throughput of the die-cutting side remained high in the first few time segments, but the throughput of the waste removal side showed a periodic collapse, with the phase difference between the two peak values ​​exceeding the preset allowable deviation threshold. Simultaneously, the cumulative die-cutting output began to consistently exceed the cumulative waste removal consumption, indicating that the buffer zone between die-cutting and waste removal was rapidly filled. If this situation continued to propagate, uneven material feeding would first occur on the stacking side, followed by a complete line stoppage. Based on this, the system first output a medium-level risk and issued a coordinated cycle time command to the die-cutting machine and the waste removal unit. If the waste removal delay continued to increase and the buffer zone approached full capacity, the system upgraded to a global shutdown warning. The purpose of this step is to integrate the local cycle time information, buffer status information and delay information that were originally scattered on each equipment side into a unified situation indicator that can be used for the entire process judgment. This enables early identification of cascading blockages, closed-loop intervention of cycle time misalignment, and proactive suppression of full-line downtime events, effectively improving the overall stability and yield of the production line. In this embodiment, the multi-source heterogeneous operating data includes programmable logic controller data of the die-cutting machine, visual monitoring data of the waste removal area, and sensor data of the stacking area; wherein, the sensor data of the stacking area includes gravity sensor data and photoelectric sensor data.

[0018] This embodiment provides a refined acquisition mechanism for multi-source heterogeneous operating data. Specifically, in the above-mentioned continuous production line scenario, relying solely on the cycle time information output by a single equipment controller can reflect the operating frequency at the equipment instruction level, but it cannot fully restore the actual passage state of materials in the waste removal area and stacking area. Therefore, the programmable logic controller data of the die-cutting machine, the visual monitoring data of the waste removal area, and the gravity and photoelectric sensor data of the stacking area are further introduced to form a cross-process, multi-view data base. Specifically, the programmable logic controller (PLC) data of the die-cutting machine mainly reflects the punching cycle, feeding cycle time, pause signals, and internal alarm status. This type of data can describe the upstream release capacity relatively stably, but its limitation is that the output cardboard determined by the controller system does not necessarily mean that it has entered the downstream effective processing state smoothly. To make up for this gap, a visual monitoring unit is set up in the waste removal area to observe waste edge peeling, cardboard posture deviation, local curling, and short-term congestion. The role of visual data here is not to perform high-precision image recognition, but to provide on-site physical evidence of whether the material is actually flowing in the waste removal area, whether it has accumulated, and whether there is any interruption through the cycle time. The stacking area further employs a combination of two types of sensor data for judgment: gravity sensor data reflects the increasing trend of the load on the stacking platform and can be used to identify whether the finished cardboard is continuously arriving; photoelectric sensor data is used to record the cardboard arrival pulse, on / off cycle, and material blockage status; after the two are combined, it is possible to distinguish between the different situations where the cardboard has arrived but has not yet formed a weight change that reaches the trigger threshold of the gravity sensor and the situation where the weight increases but the photoelectric cycle is abnormal, thereby more reliably determining whether the stacking unit is actually receiving material, short-term empty material, or an accumulation blockage. To facilitate the explanation of the data flow verification process, a specific scenario is used for illustration. Assume that within a certain short period of time, the die-cutting controller outputs three consecutive output pulses M1, M2, and M3; the waste removal vision side only recognizes two stable passing states C1 and C2, while the third image shows waste edge dragging; the corresponding photoelectric side of the stacking area only shows two arrival pulses P1 and P2, and the gravity side load only shows two step-like increases; at this time, the system determines that within this time slice, the three slices are not flowing continuously as normal, but rather the third slice is obstructed in the waste removal area; this joint acquisition method can quantitatively identify the hidden situation of normal equipment control layer commands but abnormal actual material flow on site through multi-source data fusion; In the fault-tolerance mechanism for data acquisition anomalies, if visual monitoring is affected by dust obstruction or sudden changes in lighting, the system can temporarily reduce the weight of visual data in the current time window and instead use the die-cutting controller's cycle time and the dual sensor signals of the stacking area for cross-verification; if the gravity sensor experiences zero-point drift, consistency can be verified by comparing the photoelectric positioning pulse with the historical range of the unit product's calibrated weight; if both are abnormal at the same time, the current window is marked as insufficient stacking data, and automatic speed adjustment is not directly triggered based on this. After the medicine box packaging line entered the night shift for continuous production, the die-cutting controller still showed normal cycle, and the operator did not find any obvious shutdown signal from the equipment alarm record. However, the visual screen of the waste removal area detected that the waste edge was causing local cardboard tail shaking, the photoelectric beat in the stacking area began to be lost, and the increase in gravity sensor data also slowed down significantly. Through the joint analysis of the three types of data, the system was able to confirm that the problem was not insufficient material supply on the die-cutting side, but abnormal physical flow in the waste removal area. The purpose of this step is to establish a multi-source data observation system that takes into account the equipment command layer, the field logistics layer, and the end-of-line receiving layer, so as to accurately locate the cause of cycle mismatch and avoid the true blockage process being obscured by data from a single source. In this embodiment, the flexible material state data includes paperboard humidity data, paperboard thickness data, and die-cutting die wear data.

[0019] This embodiment provides a mechanism for supplementing the acquisition of flexible material status. Specifically, based on the aforementioned multi-source operating data, if only the equipment cycle time and logistics results are observed, although the mismatch that has occurred can be detected, the underlying physical causes of the mismatch are not sufficiently revealed. For continuous manufacturing processes of paperboard, paperboard humidity, paperboard thickness, and the wear of the die-cutting die will directly change the cutting quality, waste edge peeling resistance, and downstream material receiving stability. Therefore, it is necessary to further acquire the status data of these three types of flexible materials. Specifically, the humidity of the cardboard affects the fiber bonding state and local softness. When the humidity exceeds the upper limit of the preset standard range, the edges of the cardboard are more prone to fiber tearing, adhesion, and springback, making it difficult for waste edges to detach during waste removal. When the humidity is below the lower limit of the preset standard range, the cardboard becomes more brittle, increasing the probability of debris and edge defects at local cuts, which also interferes with stable conveying. The thickness of the cardboard directly affects the punching load and interlayer stiffness. Thicker boards have higher requirements for waste removal peeling and stacking neatness, and are more prone to local accumulation at the same speed. The wear of the die reflects the sharpness of the blade and the consistency of cutting. Although the die-cutting machine may still run at the set cycle after wear, the actual cutting quality decreases, and the material resistance of the waste removal process will increase significantly. In a specific simulation scenario, a batch of cardboard can be divided into three adjacent production sections: S1, S2, and S3. S1 corresponds to normal humidity, standard thickness, and a newly replaced die, and the system observes that waste removal is continuous and stable. In S2, humidity increases under rainy conditions, and waste edge trailing begins to appear on the visual side. S3 continues production until the die usage time is close to the maintenance cycle. Although the controller does not reduce the speed on the waste removal side, the stacking zone's arrival cycle begins to be intermittently lost. By associating the operational anomalies with the status data of these three types of materials, the system can identify that the anomalies are not random noise, but are related to the co-evolution of the material and die status. As a fault-tolerant mechanism, if the humidity sensor or thickness measuring device goes offline for a short period of time, the system can use the most recent stable detection result as a transitional reference, but at the same time reduce the confidence of subsequent resistance inference; if the wear degree of the die cannot be directly obtained by the online sensor, it can also be estimated by the cumulative number of punching times, the most recent die change record and the edge burr monitoring results; if the estimated result contradicts the on-site visual performance for a long time, manual maintenance will be prompted and automatic compensation based on the die status will be suspended. In the same shift of the aforementioned medicine box paperboard production line, the first half of the product operation was stable, while the second half was affected by the increase in environmental humidity due to the maintenance of the workshop dehumidification equipment, which caused the moisture content of the paperboard to change. At the same time, the die-cutting mold was close to the upper limit of continuous use. At this time, even if the die-cutting controller maintained the original set cycle, the actual resistance to passing through the waste removal area had increased significantly. After introducing flexible material state data, the system was able to extract and output the characteristics of this gradual deviation from steady state in advance. The purpose of this step is to incorporate hidden material factors that affect the smoothness of logistics into the situation analysis process, thereby moving from seeing abnormal results to understanding the causes of abnormalities, and providing a basis for subsequent throughput correction and risk assessment. In this embodiment, the system also includes a drag coefficient update module; The resistance coefficient update module is used to calculate the real-time waste removal resistance coefficient based on paperboard humidity data, paperboard thickness data, and die-cutting die wear data; and to correct the downstream material flow throughput rate by multiplying the real-time waste removal resistance coefficient by the downstream material flow throughput rate.

[0020] This embodiment provides a real-time waste removal resistance coefficient update mechanism. Specifically, in the aforementioned scheme, although the system already knows that cardboard humidity, thickness, and die wear affect the waste removal status, if the mechanical drive throughput rate reported by the waste removal equipment is used in the coupling degree analysis without correction, there is a probability of underestimating the actual cascading blockage risk. This is because the cycle time recorded by the equipment controller is often the motor drive cycle time, rather than the effective throughput capacity of the material flow without obstruction. Therefore, this embodiment introduces a resistance coefficient update module to correct the downstream material flow throughput rate. Specifically, the real-time waste removal resistance coefficient is used to characterize the reduction coefficient corresponding to the comprehensive resistance encountered by the cardboard when passing through the waste removal zone. In the specific quantitative calculation, the system pre-calibrates the baseline resistance coefficient under standard unobstructed material conditions, which is usually set to 1. During operation, the system extracts cardboard humidity data, cardboard thickness data, and die-cutting die wear data, and compares them with their respective preset ideal reference ranges to calculate the deviation rate. The above three deviation rates are multiplied by their respective process sensitivity weights. For example, if the cardboard material is more sensitive to humidity, a higher weight is set accordingly, and a weighted sum is performed. The process sensitivity weight is assigned a value based on the Pearson correlation coefficient between the material state parameters and the throughput rate decrease in historical abnormal production data. The summation result is converted into a negative bias term and superimposed on the baseline resistance coefficient to obtain the real-time waste removal resistance coefficient that reflects the real-time throughput efficiency. In terms of physical performance, as humidity increases, the difficulty of separating waste edges increases, and the corresponding deviation rate is positive, leading to a decrease in the calculated real-time waste removal resistance coefficient. As the thickness increases, the friction and bending loads in the peeling and guiding path increase, and the real-time waste removal resistance coefficient also decreases accordingly. When the die wear intensifies, the cut integrity decreases, and the boundary between the waste edge and the finished product becomes unclear, causing the real-time waste removal resistance coefficient to further decrease to a value less than 1. After the calculation is completed, the downstream material flow throughput rate is corrected by multiplying the real-time waste removal resistance coefficient by the downstream material flow throughput rate of the equipment's original output, so as to obtain the corrected throughput rate that characterizes the actual material receiving capacity. In this way, even if the waste removal motor continues to rotate at the original speed, the system will not mistakenly believe that the downstream processing capacity has not decreased. To illustrate this more clearly, specific data simulations can be performed. Assume that within three consecutive time slots T1, T2, and T3, the waste disposal controller reports similar operating cycles, but the visual stability decreases sequentially. The system detects that in T1, the cardboard is in normal condition, and the resistance coefficient is within the baseline range; therefore, the corrected throughput is close to the original throughput. In T2, humidity increases, accompanied by slight die wear, and the corrected throughput is moderately reduced. In T3, the thickness is too large, and die wear accumulates further, causing the corrected throughput to continue to decrease. Although the driving cycles of the three time slots appear similar, the actual downstream receiving capacity involved in the risk calculation has been accurately distinguished. In terms of fault tolerance for anomalies, if only some of the three types of status data are available, such as only high humidity being detected while thickness and die data are temporarily missing, the resistance coefficient can be updated in stages based on the existing effective factors, but the system will indicate in the result label that only some material statuses are involved in the correction; if visual monitoring shows smooth passage, but the resistance coefficient remains high for a long time, the system will first check whether the status sensor calibration is inaccurate to avoid excessively conservative adjustments to downstream capacity; if the throughput after correction is lower than the known physical lower limit of the equipment, the system will cut off the throughput at the safety lower limit to prevent abnormal data from causing distortion of the control strategy. When the same batch of cardboard for medicine boxes was produced to the later stage, the speed setting of the waste removal belt was not changed. However, due to the increase in cardboard moisture and the accumulation of die wear, waste edges were dragged in the waste removal area. If the speed is still reported directly by the controller, the system will misjudge that the downstream still has the original acceptance capacity. After adding the resistance coefficient, the effective throughput of the waste removal side was scientifically reduced, the risk assessment was raised in advance, and the scheduling module was prompted to adopt a safer cycle coordination strategy. The purpose of this mechanism is to clearly distinguish between the mechanical drive state of the equipment and the actual smooth flow of materials, thereby achieving a dynamic characterization of the downstream actual processing capacity and improving the confidence of subsequent coupling analysis and risk assessment. In this embodiment, the coupling degree calculation module is used to quantitatively evaluate the coordination status of the operation frequency of upstream and downstream processes within a specific time window; the cross-process rhythm coupling degree is inversely proportional to the timing phase difference; if the absolute value of the difference between the cross-process rhythm coupling degree and the preset target value is not greater than the preset synchronization threshold, then marker data representing that the rhythm between processes has reached a synchronized state is generated; if the absolute value of the difference is greater than the preset synchronization threshold, then marker data representing that the rhythm between processes is not synchronized is generated.

[0021] This embodiment provides a process synchronization status marking mechanism. Specifically, in the basic scheme, the coupling degree can be used to identify the risk of cycle mismatch. However, in actual production, it is also necessary to clearly distinguish between two types of states: those with slight fluctuations but normal overall coordination and those that have deviated to the point of needing intervention. Otherwise, the system may overreact to short-term normal elastic fluctuations. Therefore, this embodiment further introduces a preset target value and a synchronization threshold to output the marking data of the rhythm synchronization status between processes. Specifically, the preset target value can be understood as the typical coupling level corresponding to the stable coordination of adjacent processes under a certain product specification, a certain cardboard material, and a certain equipment configuration. Since die-cutting, waste removal, and stacking are not strictly synchronous equipment, the target value does not have to be equal to the theoretical absolute synchronization, but rather a steady-state coordination zone that conforms to the actual mechanical characteristics of this type of production line. If the difference between the real-time coupling degree and the target value is within the preset allowable deviation range, it means that although there are slight elastic differences between the upstream release rhythm and the downstream receiving rhythm, they still fall within the self-stabilizing range. The system can then generate a synchronization status flag as one of the bases for not needing scheduling. To facilitate understanding, a brief simulation scenario can be provided: For a certain product specification, the die-cutting to waste removal process typically exhibits a coupled waveform that remains close to the target range during stable mass production. If, in multiple consecutive time slices, the phase peak offset is only a slight sequential misalignment in a single time slice, and subsequently returns naturally, the system determines that it is still in a synchronized state and outputs a synchronization flag. If, in consecutive time slices, the peak misalignment widens and cannot return to the target range, then the asynchrony flag is established, and the process transitions to the risk assessment reinforcement branch. In the anomaly handling mechanism, if a certain time window is too short, resulting in the upstream and downstream beat waveforms not being able to reflect a stable trend, the system will not directly output the synchronization status, but will extend the observation window before making a judgment; if the equipment is in the start-stop switching, plate replacement, or manual spot check stage, the beat itself has planned disturbances, the system can temporarily freeze the synchronization judgment or use the dedicated target interval for this working condition to avoid misidentifying manually controlled fluctuations as abnormal rhythm disorders; During the initial set time period after the batch change is completed on the medicine box packaging line, the die-cutting machine and the waste removal unit gradually increase from low speed to target capacity. During this stage, although their cycle times are not completely overlapped, as the feeding stabilizes and the waste edge is smoothly removed, the coupling degree gradually approaches the preset target value and remains stable within the synchronization threshold range. The system then generates a process synchronization mark and allows the entire line to enter a normal automatic operation state. Conversely, if the subsequent increase in humidity causes the waste removal to lag, the synchronization mark will be canceled. The purpose of this step is to provide a clear criterion for stable coordination of the production line, thereby enabling tolerance to normal elastic fluctuations and differentiation of abnormal rhythmic drifts, and reducing unnecessary control actions. In this embodiment, the data acquisition module is also used to acquire data on the remaining capacity of the buffer band; The risk assessment module combines cross-process rhythm coupling degree, dynamic buffer margin dissipation rate, processing delay time and buffer band remaining capacity data to predict the probability and time threshold of local delay evolving into a full-line shutdown event; the product of probability and time threshold is output as the cascade blockage risk index.

[0022] This embodiment provides a risk assessment mechanism that incorporates the remaining capacity of the buffer zone. Specifically, in the aforementioned basic scheme, the system is already able to identify increased risk based on coupling degree, dissipation rate, and delay time. However, in a short buffer production line, simply knowing that the pressure is increasing is insufficient to determine how far away a complete line shutdown is. This is because whether a local delay will actually expand into a complete line blockage is objectively limited by how much physical carrying capacity the buffer zone has left. Therefore, this embodiment further collects data on the remaining capacity of the buffer zone. Specifically, the remaining capacity of the buffer strip can be characterized by the remaining length of the cardboard that the strip can accommodate, the number of stacked sheets allowed, or the remaining acceptance time calculated according to the process. When this data is combined with the dynamic buffer margin dissipation rate, the system can determine whether the current mismatch occurs within the preset safety tolerance range or the critical zone. If the dissipation rate is greater than the preset first dissipation rate threshold but the remaining capacity is higher than the preset capacity safety lower limit, the local delay may still be eliminated by the downstream. If the dissipation rate is greater than the preset first dissipation rate threshold and the remaining capacity is lower than the preset capacity warning lower limit, any new pause may immediately trigger material blockage. In the specific reasoning logic, the system performs a basic weighted summation of cross-process rhythm coupling degree, dynamic buffer margin dissipation rate, and processing delay time. Since the lower the coupling degree, the higher the risk, a negative weight is assigned to the cross-process rhythm coupling degree during weighting to represent the degree of mismatch. The weighted summation result is mapped to a value in the range of 0 to 1 to represent the probability of a local delay evolving into a full-line shutdown event. Meanwhile, the system uses the currently collected buffer remaining capacity data divided by the current dynamic buffer remaining capacity dissipation rate to calculate the theoretical remaining time before the buffer is fully occupied, and uses the reciprocal transformation value or inverse proportional normalized value corresponding to the theoretical remaining time as the time threshold; the smaller the remaining space and the faster the dissipation, the shorter the remaining safe response time, and the larger the calculated time threshold; the risk assessment module multiplies the probability with the time threshold, and uses the resulting product as the final cascade blocking risk index output after incorporating buffer space constraints; the time threshold here reflects the remaining response margin from now until the most likely triggering of the line stop, the shorter it is, the stronger the corresponding multiplicative amplification effect, indicating that the system should trigger a high-priority coordinated scheduling instruction; Two scenarios can be illustrated using a simplified simulation. In scenario A, the buffer zone between die-cutting and waste removal still has a margin exceeding the preset safety boundary. Although waste removal exhibits a short-term tail, the stack-end reception is stable, and the system judges it to be more likely a recoverable disturbance. In scenario B, the same degree of waste removal tail occurs when the buffer zone is nearly full. At this point, even if the added delay is not significant, it is easier to form upstream squeezing within several time slices. Therefore, the risk outputs of the two scenarios are significantly different. This difference does not come from the algorithm itself, but from the change in the physical buffer conditions on the cascading propagation capability. As a fault-tolerance mechanism, if the buffer strip remaining capacity sensor malfunctions for a short time, for example, due to cardboard warping causing counting deviation, the system can simultaneously refer to upstream output, downstream consumption, and photoelectric counting difference for cross-verification; if the difference among the three is too large, the current capacity value is marked as needing correction, and the risk assessment can revert to the basic degraded assessment mode without capacity factor, but the abnormal value is not directly used as the basis for shutdown triggering; if the buffer strip itself is mechanically locked in a special position due to maintenance, the system should read the maintenance status word and suspend the routine inference based on remaining capacity. During the continuous production process of this medicine box packaging line, the buffer zone between die-cutting and waste removal was originally able to accommodate a transition length of cardboard flow. After the shift operated in high-capacity mode for a long time, the buffer zone gradually approached its upper limit. At this time, waste edge dragging occurred again on the waste removal side. The system not only saw a decrease in coupling and an increase in delay, but also detected that the remaining capacity was close to the preset minimum safety capacity. Therefore, it was determined that the local lag was very likely to evolve into a line stoppage within the preset time threshold, and the risk level was raised in advance. The purpose of this mechanism is to advance risk assessment from whether something is abnormal to how long the abnormality can be tolerated and whether it will escalate into a line stoppage, thereby achieving more timely early warnings and scheduling decisions that are more in line with the actual working conditions. In this embodiment, the adaptive production cycle time collaborative scheduling strategy data package includes die-cutting machine speed adjustment instructions and waste removal belt speed adjustment instructions; The die-cutting machine is equipped with an independent die-cutting control node, and the actuator corresponding to the waste removal process is equipped with an independent waste removal control node; the situation scheduling module is used to send the die-cutting machine speed adjustment command to the control node corresponding to the die-cutting process for execution; and to send the waste removal belt speed adjustment command to the control node corresponding to the waste removal process for execution.

[0023] This embodiment provides a dual-process collaborative speed regulation execution mechanism. Specifically, in the aforementioned scheme, the system is already able to identify medium-level risks and generate collaborative scheduling strategies. However, if the scheduling strategy remains only at the abstract level and is not implemented in specific control objects and specific instructions, it is difficult to form an executable closed loop. Therefore, this embodiment concretizes the strategy data package into die-cutting machine speed adjustment instructions and waste removal belt speed adjustment instructions, and sends them to the corresponding control nodes for execution. Specifically, the die-cutting machine speed adjustment command mainly affects the upstream release speed, which is suitable for scenarios where the upstream supply is too fast and the downstream digestion is insufficient; the waste removal belt speed adjustment command mainly affects the conveying and peeling coordination rhythm on the waste removal side, which is suitable for improving the downstream receiving and exporting efficiency; the two types of commands are not always increased or decreased at the same time, but are combined in different ways according to the source of risk; if the problem is mainly due to a decrease in the actual waste removal capacity, the die-cutting machine can be slowed down by the first preset gradient, and the waste removal belt can be moderately accelerated in coordination; if the problem is mainly due to large fluctuations in the die-cutting cycle and sufficient waste removal capacity, the focus can be on smoothing the changes in the die-cutting machine speed to avoid the cardboard being transported downstream in a pulse manner; Using a simulation approach, we can define a time window with a medium-level risk level. The system generates instruction package D1. D1 contains a command to slightly reduce the die-cutting machine's current speed by one level, and a command to compensate the waste removal belt by half a level. In the next time window, if the visual system shows that the waste edge has passed and stabilized, and the stacking area's positioning pulses become continuous again, the system maintains this combination. If the waste removal side remains unstable, the subsequent strategy package D2 can continue to reduce the die-cutting release speed instead of unconditionally increasing the belt speed, to avoid cardboard instability if the root cause of resistance is not resolved. In the anomaly handling mechanism, if the die-cutting machine control node confirms that execution is possible, but the waste removal control node refuses to accelerate due to local protection logic (e.g., the current waste removal area detects an open safety door or excessive tension), the system should allow unilateral adjustment and feed back the limiting condition to the central control layer. If any control node has a communication network delay risk, the policy data packet is configured with a preset time tag when it is generated. When the reception time exceeds the effective duration limited by the preset time tag, the corresponding instruction will no longer be executed to prevent delayed instructions from being delivered at the wrong time. If the site is in manual adjustment mode, the automatically issued action can be switched to a suggested instruction instead of mandatory execution. In this medicine box packaging line, the system detected two consecutive instances of trailing due to high humidity in the waste removal zone, but this did not yet reach the shutdown level. The situational scheduling module then generated a set of collaborative control strategies: sending a slight speed reduction command to the die-cutting machine to slow down the density of new cardboard entering the waste removal zone; and simultaneously sending a collaborative speed increase command to the waste removal conveyor belt to help the cardboard that had entered the waste removal zone to quickly leave the high-resistance section. After execution, the gravity and photoelectric data in the stacking area returned to a uniform increase, indicating that the linkage control was effective. The purpose of this step is to directly transform the risk identification results into executable actions for the equipment control layer, thereby achieving a seamless connection from situation analysis to closed-loop adjustment of production cycle time. In this embodiment, the situational coordination response time window is the time period required from the time the system extracts rhythm disorder features from multi-source heterogeneous operating data to the time required to generate and issue adaptive production rhythm coordination scheduling strategy data packets.

[0024] This embodiment provides a situational coordination response time window definition and control mechanism. Specifically, in a continuous manufacturing site, the effectiveness of scheduling depends not only on the correctness of the strategy, but also on whether the strategy is issued within the preset situational coordination response time window. If the system issues an instruction only after the rhythm misalignment has been transmitted to the buffer limit, even if the judgment is accurate, it is difficult to prevent the shutdown. Therefore, this embodiment clearly defines the situational coordination response time window as the time boundary from misalignment identification to strategy issuance. Specifically, the starting point of this time window is not the moment when the equipment actually experiences mechanical collision or blockage, but rather the moment when the system extracts rhythmic imbalance characteristics from multi-source heterogeneous operating data, such as the coupling waveform starting to deviate continuously, abnormal clearance visual pass status, and intermittent stacking zone arrival cycle; the ending point is the moment when the corresponding collaborative scheduling strategy data packet is generated and sent to the equipment control node; by using this time period as an independent controlled indicator, the system can assess whether it has enough time to intervene using the existing buffer margin; In engineering implementation, this time window can be divided into four consecutive sub-stages: feature recognition, risk confirmation, strategy generation, and communication distribution. If the time consumption of a certain sub-stage is abnormally high, such as visual data processing queuing or control network congestion, the system can trigger a simplified strategy branch. The so-called simplified strategy branch means that instead of waiting for more complex multi-factor refinement, a conservative speed adjustment command is issued based on the most critical imbalance evidence at present, thereby delaying the occupancy rate of the slow-pass belt and waiting to obtain complete feedback data for precise adjustment. A simplified simulation scenario can be illustrated. Suppose that in time slice A, the system first detects a continuous deviation between the peak values ​​of the die-cutting and waste removal cycles. In time slice B, the risk confirmation module determines that the deviation is consuming the buffer capacity. If the strategy has been delivered to the control node before time slice C, it can be considered that it is still within the effective response window. Conversely, if the instruction is not delivered until after time slice D, and the buffer is already close to full, then although the scheduling is completed, the best intervention opportunity has been missed. As a fault-tolerance mechanism, if data loss makes it difficult to accurately pinpoint the starting point, the system can use the earliest credible anomaly moment as the window starting point and record the conservative forward shift of the starting point in the log; if the control node confirmation is delayed or there is a network anomaly, the system should explicitly report the current status and, if necessary, directly upgrade it to an early warning instead of continuing to wait; if the site is in a planned shutdown, version change, or trial run state, this time window will not be included in the production response evaluation. Under the high humidity conditions of this medicine box packaging line, the visual monitoring in the waste removal area identifies the trailing waste edge. After a preset response time, the photoelectric pulses on the stacked side begin to sparse. If the system can complete the strategy packaging and issue the policy within this early window, the die-cutting machine can still slow down and the waste removal belt can still compensate, thus keeping the disturbance local. If the response is significantly delayed, the disturbance has been transmitted to the entire line, and subsequent handling can only rely on shutdown protection. The purpose of this mechanism is to establish clear real-time constraints for the entire situation analysis system, thereby enabling timely management of the early warning link, computing link, and control link, and ensuring the field availability of collaborative scheduling. In this embodiment, the situational scheduling module is also used to compare the dynamic buffer margin dissipation rate with the preset minimum and maximum dissipation thresholds: if it is not less than the maximum dissipation threshold, the speed adjustment intervention instruction is programmed into the adaptive production cycle collaborative scheduling strategy data packet; if it is not less than the minimum and less than the maximum dissipation threshold, an early warning message is output; if it is less than the minimum dissipation threshold, marker data to maintain the current buffer capacity state is generated.

[0025] This embodiment provides a hierarchical scheduling mechanism based on dynamic buffer margin dissipation rate. Specifically, in the aforementioned scheme, the risk index can comprehensively reflect the overall situation, but in actual control, the buffer zone is the most direct safety boundary before the blockage spreads. If its dissipation rate cannot be judged hierarchically, the system may miss the best time for light intervention before the risk has fully formed, or issue speed adjustment commands too early when there are slight fluctuations. Therefore, this embodiment adds dissipation threshold comparison logic. Specifically, the lowest dissipation threshold corresponds to the buffer zone being consumed but still within the observable zone; the highest dissipation threshold corresponds to the buffer zone rapidly losing its safety margin, requiring immediate speed adjustment intervention; when the dissipation rate is below the lowest threshold, it indicates that the slight imbalance between upstream and downstream is not enough to threaten buffer safety, and the system can output marker data to maintain the current buffer zone capacity status; when in the intermediate range, the system does not rush to adjust the speed, but first outputs warning information to remind the operator to pay attention to the waste removal screen, cardboard status, or die-cutting condition; only when the dissipation rate reaches or exceeds the highest threshold will the speed adjustment intervention command be incorporated into the collaborative scheduling strategy data packet in order to try to change the logistics rhythm before the buffer zone is exhausted; Using a simulated scenario, the buffer status of a certain shift can be divided into three stages: H1, H2, and H3. In stage H1, although there is a slight difference between upstream and downstream, the buffer margin changes slowly, and the system only marks it as stable. In stage H2, the increased resistance to waste removal causes the margin to begin to decrease more significantly, and the system issues a warning but does not force any action. In stage H3, the rate of decrease in the margin continues to accelerate, meaning that if speed adjustment is not implemented, the buffer zone will be filled within the calculated theoretical remaining time. Therefore, the system writes the speed adjustment command into the strategy package and sends it out immediately. This avoids triggering the shutdown mechanism for all anomalies. In the anomaly handling mechanism, if the dissipation rate fluctuates repeatedly near the threshold boundary, the system can introduce a short-term hold mechanism, which requires that the corresponding interval conditions be met for several consecutive effective time slices before upgrading the action, in order to prevent frequent speed adjustment triggered by a single noise. If the buffer capacity sensing is abnormal and the confidence of the dissipation rate decreases, the intermediate interval will prioritize the output of prompt information and will not directly execute automatic speed adjustment. If a higher level shutdown warning has been received on site, this hierarchical strategy will automatically give way to shutdown protection. In this medicine box packaging line, as the high-moisture cardboard enters the waste removal area, the buffer strip between die-cutting and waste removal initially decreases slowly, and the system only displays a warning on the central control interface; as the waste edge drag increases, the rate of decrease in the buffer strip accelerates significantly, reaching the high dissipation zone, and the system immediately writes and issues the die-cutting deceleration and waste removal coordination adjustment instructions into the strategy package; after a certain period of dissipation rate decline, the system returns to the warning or maintenance state; The purpose of this mechanism is to manage the buffer safety boundary as an independent control object in a hierarchical manner, so as to achieve gradual control with less intervention when there are slight fluctuations, early response when there is a deteriorating trend, and rapid speed adjustment when there is a critical state. In this embodiment, the adaptive production cycle time collaborative scheduling strategy data package is used to dynamically smooth the speed fluctuation difference between upstream and downstream processes based on the cascading congestion risk index, so as to suppress the phenomenon of speed fluctuation being amplified step by step during the propagation of the production line.

[0026] This embodiment provides a dynamic smoothing mechanism to suppress the propagation of speed fluctuations along the production line. Specifically, in the aforementioned scheme, the system can already adjust the speed according to the risk level. However, if the speed adjustment itself takes abrupt steps, such as a sudden drop upstream and a sudden increase downstream, although it may seem to alleviate the current congestion locally, it may continue to transmit new fluctuations to subsequent processes, forming a chain reaction of alternating periods of slow speed at the beginning, excessive speed at the end, and material interruption or accumulation at the end. Therefore, this embodiment further specifies that an adaptive production cycle collaborative scheduling strategy data package is used to dynamically smooth the speed fluctuation difference between upstream and downstream processes. In detail, the smoothing here is not simply about pursuing uniform speed across all equipment, but rather about determining the adjustment range and rhythm based on the cascading blockage risk index. When the risk is low, only step-by-step corrections are made to maintain the original capacity as much as possible. When the risk increases, the adjustment is accelerated, but large, sudden jumps are still avoided. The physical basis for this is that the mechanical inertia of the die-cutting, waste removal, and stacking equipment are different, and cardboard is a flexible material. Sudden speed changes can easily cause instability in the sheet posture, increased waste edge pulling, or stacking misalignment. By dynamically smoothing the speed difference, local impacts can be broken down into several smaller beat corrections, allowing the buffer zone to absorb fluctuations instead of becoming a fluctuation amplifier. A brief simulation scenario can be provided; if the control system abruptly reduces the die-cutting machine's speed from its current speed to a preset extremely low threshold, a short-term material flow interruption will occur on the waste removal side, affecting the continuity of the stacking cycle; if the die-cutting machine's speed remains unchanged, while the control waste removal unit abruptly increases the transmission rate, the tailing may be further aggravated because the cardboard's posture has not yet recovered; in this embodiment, the strategy data packet can break down a large adjustment into several consecutive small steps, first reducing the difference, then observing whether the buffer margin and stacking reception status have stabilized, and then deciding whether to continue to the next step of correction; in this way, speed fluctuations will not propagate along the line in the form of spikes; As a fault-tolerance mechanism, if a device is unable to execute a target segment of the smoothing curve due to its own safety boundaries, such as a die-cutting machine approaching its minimum stable speed or a waste removal belt approaching its allowable upper limit, the system will automatically truncate that target segment and transfer the unfinished smoothing amount to compensation suggestions for the other side. If the risk index has reached the highest level, it means that smoothing control is insufficient to ensure safety. In this case, the system should prioritize generating a shutdown warning instead of continuing to attempt gradual speed adjustment. If physical blockage has occurred in the downstream stacking area, the conventional strategy based on speed smoothing will be stopped, and a fault clearing process will be implemented instead. In this medicine box packaging line, the waste removal zone experienced continuous slight trailing due to the rising moisture content of the cardboard. If the die-cutting machine continued to maintain high-frequency sheet output, the speed difference would continue to widen. Based on the current risk index, the system did not drastically reduce the die-cutting speed all at once. Instead, it first issued a series of continuous slight reductions in the adjustment sequence, while simultaneously allowing the waste removal belt to provide gentle compensation. After observing that the photoelectric pulses in the stacking area became uniform again, the system gradually restored the new balanced rhythm. In this way, the fluctuations in the first stage were not further amplified into alternating periods of empty material and stacking in the second stage. The purpose of this mechanism is to control the scheduling action itself as a low-impact, absorbable, and recoverable rhythm correction process, thereby weakening the speed fluctuation propagation chain and suppressing the phenomenon of fluctuation amplification at each stage in continuous manufacturing.

[0027] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A system for analyzing the entire process of die-cutting, waste removal, and stacking based on industrial big data, characterized in that: include: The data acquisition module is used to collect multi-source heterogeneous operation data and flexible material status data of heterogeneous processes; the heterogeneous processes are adjacent upstream and downstream processes in die-cutting, waste removal and stacking processes; The multi-source heterogeneous operation data includes upstream and downstream material flow throughput, upstream output, downstream consumption, and processing delay time; The coupling degree calculation module is used to compare the throughput of upstream and downstream material flows to extract the time phase difference, and calculate the cross-process rhythm coupling degree between upstream and downstream processes based on the reciprocal of the time phase difference or a preset inverse proportional function value. The dissipation rate calculation module is used to calculate the dynamic buffer margin by integrating the difference between upstream output and downstream consumption within a specific time window, and to use the time derivative of the dynamic buffer margin as the dynamic buffer margin dissipation rate between upstream and downstream processes. The risk assessment module is used to calculate the cascade blockage risk index by weighted summation of the cross-process rhythm coupling degree, the dynamic buffer margin dissipation rate, and the processing delay time; The situational scheduling module is used to compare the cascading blocking risk index with preset minimum and maximum risk thresholds: if it is not less than the maximum risk threshold, a global shutdown warning data packet is generated. If the risk threshold is not less than the minimum and is less than the maximum risk threshold, an adaptive production cycle collaborative scheduling strategy data packet is generated and sent to the corresponding control node of the heterogeneous process that is connected to the system for execution within a preset situational collaborative response time window; if the risk threshold is less than the minimum risk threshold, marker data to maintain the current operating state is generated.

2. The whole-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 1, characterized in that, The multi-source heterogeneous operating data includes programmable logic controller data of the die-cutting machine, visual monitoring data of the waste removal area, and sensor data of the stacking area; wherein, the sensor data of the stacking area includes gravity sensor data and photoelectric sensor data.

3. The whole-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 2, is characterized in that, The flexible material status data includes paperboard humidity data, paperboard thickness data, and die-cutting die wear data.

4. The whole-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 3, is characterized in that, The system also includes a drag coefficient update module; The resistance coefficient update module is used to calculate the real-time waste removal resistance coefficient based on the paperboard humidity data, the paperboard thickness data, and the die-cutting die wear data; and to correct the downstream material flow throughput rate by multiplying the real-time waste removal resistance coefficient by the downstream material flow throughput rate.

5. The whole-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 1, characterized in that, The coupling degree calculation module is used to quantitatively evaluate the coordination status of the operation frequency of upstream and downstream processes within a specific time window; the cross-process rhythm coupling degree is inversely proportional to the timing phase difference; if the absolute value of the difference between the cross-process rhythm coupling degree and the preset target value is not greater than the preset synchronization threshold, then marker data representing that the rhythms between processes have reached a synchronized state is generated; if the absolute value of the difference is greater than the preset synchronization threshold, then marker data representing that the rhythms between processes are not synchronized is generated.

6. The full-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 1, characterized in that, The data acquisition module is also used to acquire data on the remaining capacity of the buffer band. The risk assessment module combines the cross-process rhythm coupling degree, the dynamic buffer margin dissipation rate, the processing delay time, and the buffer band remaining capacity data to predict the probability and time threshold of a local delay evolving into a full-line shutdown event; the product of the probability and the time threshold is output as the cascade blockage risk index.

7. The whole-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 2, is characterized in that, The adaptive production cycle collaborative scheduling strategy data package includes die-cutting machine speed adjustment instructions and waste removal belt speed adjustment instructions; The die-cutting machine is equipped with an independent die-cutting control node, and the actuator corresponding to the waste removal process is equipped with an independent waste removal control node; the situation scheduling module is used to send the die-cutting machine speed adjustment command to the control node corresponding to the die-cutting process for execution; and to send the waste removal belt speed adjustment command to the control node corresponding to the waste removal process for execution.

8. The full-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 1, characterized in that, The situational coordination response time window is the time period required from the time the system extracts the rhythm disorder features from the multi-source heterogeneous operating data to the time required to generate and issue the adaptive production rhythm coordination scheduling strategy data packet.

9. The whole-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 1, characterized in that, The situational scheduling module is also used to compare the dynamic buffer margin dissipation rate with the preset minimum and maximum dissipation thresholds: if it is not less than the maximum dissipation threshold, the speed adjustment intervention instruction is programmed into the adaptive production cycle collaborative scheduling strategy data packet; if it is not less than the minimum and less than the maximum dissipation threshold, an early warning message is output. If the value is less than the minimum dissipation threshold, then marker data is generated to maintain the current buffer band capacity state.

10. The full-process operation status analysis system for die-cutting, waste removal, and stacking based on industrial big data as described in claim 1, characterized in that, The adaptive production cycle time collaborative scheduling strategy data package is used to dynamically smooth the speed fluctuation difference between upstream and downstream processes based on the cascaded congestion risk index, so as to suppress the phenomenon of speed fluctuation being amplified step by step during the propagation of the production line.

Images