Dynamic monitoring method and system for multi-section material temperature of sheet extrusion

Abstract

This invention discloses a dynamic monitoring method and system for multi-stage material temperature in sheet extrusion, relating to the field of extrusion molding technology. It aims to solve the technical problem of lacking a dedicated online dynamic monitoring scheme for the critical sub-stage of die lip sealing during planned shutdowns in existing sheet extrusion processes. This makes it impossible to quantitatively control the sealing process and predict related quality defects and irreversible die damage risks. The method includes the following steps: S1 Real-time acquisition of the extruder's overall operating status signals to accurately identify the start node of the critical sub-stage of die lip sealing, triggering the monitoring system to start; S2 Based on pre-recorded die structural attributes, thermal performance attributes of the processed material, and real-time environmental conditions, a dynamic monitoring benchmark model adapted to the current shutdown conditions is generated. This invention has the advantages of achieving full-process quantitative monitoring of critical sub-stages, accurate risk warning, and standardized control, fundamentally avoiding die life loss and product quality risks.

Description

Technical Field

[0001] This invention relates to the field of extrusion molding technology, and more specifically, to a method and system for dynamic monitoring of multi-stage material temperature in sheet extrusion. Background Technology

[0002] Sheet extrusion is one of the core molding processes in the field of polymer material processing, widely used in high-end manufacturing fields such as food and pharmaceutical packaging, medical cleanroom consumables, optical functional films, and photovoltaic encapsulation films. As a core functional component of the sheet extrusion production line, the flat die head's structural integrity, operational stability, and service life directly determine the final sheet product's quality consistency, production yield, and overall manufacturing cost. Planned shutdown for cooling is an indispensable key process in the entire sheet extrusion production process. This process must be performed in a standardized manner at the end of each production cycle. The standardization and accuracy of this process directly affect the long-term performance of the die head, the product debugging cycle after subsequent startup, and the yield rate.

[0003] In existing technologies, temperature control and monitoring in sheet extrusion processes are mostly focused on closed-loop temperature control of the barrel and die head during steady-state production, as well as emergency cooling protection in case of sudden failures. The industry has long formed a fixed technical understanding, generally attributing common industry pain points such as irreversible scratches and deformation of the die head lip, die head leakage, persistent black spot and crystal point defects during startup, and significantly shortened die head fatigue life to factors such as improper control of process parameters during steady-state production, inadequate daily disassembly and maintenance of the die head, and non-standard operation during startup and heating. Long-term production practice has revealed that the core root cause of the aforementioned industry pain points lies in the fact that existing technologies completely ignore the extremely short, critical sub-stage of die lip sealing during planned shutdown and cooling. This stage is the key node that determines the microscopic deformation of the die lip, the effectiveness of the melt sealing layer, and the risk of residual melt oxidation. Existing technologies lack a dedicated, systematic online dynamic monitoring solution for this critical sub-stage, relying solely on the subjective judgment of sealing effect by on-site operators based on visual experience. This makes it impossible to achieve online quantitative monitoring of melt solidification behavior, die axial temperature field distribution, and sealing layer integrity during this stage, and even more impossible to predict and avoid irreversible structural damage to the die and subsequent product quality risks that may occur during this stage. In view of this, we propose a dynamic monitoring method and system based on multi-stage material temperature of sheet extrusion. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic monitoring method and system for multi-stage material temperature in sheet extrusion, in order to solve the technical problem that existing sheet extrusion processes lack a dedicated online dynamic monitoring scheme for the critical sub-stage of die lip sealing and finishing, making it impossible to quantitatively control the sealing process and predict related quality defects and the risk of irreversible damage to the die.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a dynamic monitoring method for multi-stage material temperature in sheet extrusion, comprising the following steps:

[0006] S1 collects real-time operating status signals of the extruder and accurately identifies the critical sub-stage start-up node of the die lip sealing and finishing, triggering the start of the monitoring system.

[0007] S2 generates a dynamic monitoring benchmark model adapted to the current shutdown condition based on the pre-entered mold head structural attributes, processing material thermal performance attributes, and real-time environmental conditions.

[0008] S3 performs fully synchronous acquisition of the material temperature of all sections along the axial direction of the die lip, and obtains a real-time material temperature dataset with timestamps that corresponds one-to-one with the section position;

[0009] S4, based on the collected multi-segment material temperature dataset, calculates the multi-dimensional monitoring feature quantities of the critical sub-stage by coupling the correlation between location and time.

[0010] S5 performs adaptive dynamic adjustment of the judgment threshold of the monitoring benchmark model based on the real-time progress of melt solidification.

[0011] S6 compares the real-time calculated monitoring features with the dynamically adjusted judgment thresholds in real time to perform hierarchical early warning and risk zoning and source tracing.

[0012] Based on the coupled calculation results of multi-dimensional monitoring features, S7 completes the quantitative confirmation of the edge sealing status and the archiving of monitoring data throughout the entire process.

[0013] Preferably, the critical sub-stage of the die lip sealing and finishing has a time boundary defined by the following start times: the extruder screw stops feeding, the barrel metering section is emptied, the screw drops to a preset idle speed, and no new material is continuously extruded from the die lip. The end times are defined by the formation of a stable and continuous sealing layer of melt at the lip, reaching a preset solidification critical state, and the screw officially stopping. The critical sub-stage is an independent monitoring cycle separate from the steady-state production stage and the subsequent overall cooling stage. It enters only when all state conditions at the start time are met and terminates when all state conditions at the end time are met. The start and end states are precisely locked through a joint determination method of multi-condition logical product.

[0014] The start and termination states of the critical sub-stage are quantitatively determined using the following algorithm formula:

[0015] ;

[0016] ;

[0017] ;

[0018] in, This is the comprehensive judgment value for the critical sub-stage start-up time. These are the individual judgment values ​​for each state condition corresponding to the startup time. This is the comprehensive judgment value at the end of the critical sub-stage. These are the individual judgment values ​​for each state condition corresponding to the end time. This serves as a trigger for the monitoring system.

[0019] Preferably, the dynamic monitoring benchmark model includes a melt solidification critical state judgment benchmark, a lip axial temperature difference allowable range benchmark, a solidification front advancement synchronicity benchmark, a thermal stress safety judgment benchmark, and a sealing integrity judgment benchmark; the benchmarks are interconnected and matched, and are synchronously generated based on the operating parameters of the current shutdown condition to form a unified monitoring and judgment system, which is different from the fixed monitoring benchmarks of steady-state production conditions that do not change with the production process. The threshold values ​​of each benchmark are generated through the mapping function of the corresponding operating parameters.

[0020] The benchmark thresholds of the dynamic monitoring benchmark model are generated using the following algorithm formula:

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] ;

[0026] in, It serves as a reference for the critical solidification temperature of the melt. This serves as a benchmark for the allowable range of axial temperature difference at the lip. To solidify the benchmark for frontier synchronization, As a benchmark for determining thermal stress safety, As a criterion for determining the integrity of the seal, For the thermal properties of the processed materials, For mold head structural properties, For real-time environmental status, , , , , Parameter mapping functions corresponding to various monitoring benchmarks.

[0027] Preferably, the simultaneous acquisition of multi-segment material temperature across all axial zones of the die lip is specifically achieved by: targeting the dynamic changes in the solidification state of the melt within the critical sub-stage, controlling the temperature acquisition units built into each zone of the die lip with a unified clock trigger signal to perform synchronous data acquisition at the same frequency and without time difference, synchronously recording the acquisition time and the unique location number of the corresponding zone, and forming a two-dimensional structured real-time material temperature dataset according to the acquisition sequence and zone location, and performing effective data acquisition only when the monitoring system is effectively activated;

[0028] The synchronously acquired real-time material temperature dataset is constructed using the following algorithm formula:

[0029] ;

[0030] ;

[0031] in, The unique location number is assigned to the partitions that are axially equidistantly distributed at intervals along the die lip. To ensure a unified clock-triggered synchronous acquisition time, For the number The partition at the time of collection Synchronously collected material temperature data; This represents the total number of axial partitions on the die head lip. This refers to the total number of synchronous data collections within a single monitoring cycle. This is a structured real-time material temperature dataset with location and timestamps. This serves as a trigger for the monitoring system.

[0032] Preferably, the multi-dimensional monitoring features of the critical sub-stage include four core features: the synchronous deviation of the melt solidification front, the characteristic value of the axial gradient change of the lip temperature field, the dynamic attenuation coefficient of the sealing integrity of the edge, and the cumulative value of the thermal stress fatigue of the lip. All features are obtained through a coupled calculation method of multi-data cross-operation based on the positional and temporal correlation of multi-segment material temperature datasets. Single-point material temperature monitoring data cannot establish a full-axial temperature field linkage relationship and cannot complete the independent calculation of the corresponding features.

[0033] The coupled calculation of the multi-dimensional monitoring feature quantities is implemented through the following general algorithm framework:

[0034] ;

[0035] in, The output value is calculated by coupling the multi-dimensional monitoring feature quantities. For coupled computation functions, For the number The partition at the time of collection Material temperature data, To match the number Adjacent partitions are acquired at the same time. Material temperature data, For the number The partition contains the material temperature data at the adjacent previous acquisition time.

[0036] Preferably, the adaptive dynamic adjustment of the judgment threshold of the monitoring benchmark model is specifically as follows: according to the real-time progress of melt solidification, the critical sub-stage is divided into multiple continuous progressive monitoring stages, with each monitoring stage corresponding to the real-time progress of melt solidification. A specific judgment threshold is matched for each monitoring stage to be adapted to the solidification characteristics of that stage. At the same time, based on the fluctuations of the real-time environmental state and the initial state of the die head, the judgment threshold is synchronously and in real-time corrected to ensure that the judgment threshold is continuously adapted to the current working condition.

[0037] The adaptive dynamic adjustment of the judgment threshold is achieved through the following algorithm formula:

[0038] ;

[0039] ;

[0040] in, For the time of data collection The lip-edge all-axial average melt solidification process, This represents the total number of axial partitions on the die head lip. For the number The partition at the time of collection The process of melt solidification, For the time of data collection The adaptively adjusted dynamic threshold matrix This is the initial baseline threshold matrix corresponding to each monitored feature. This is a stage adjustment factor based on the average solidification process. These are correction factors based on real-time environmental conditions. This represents the real-time ambient temperature.

[0041] Preferably, the implementation of graded early warning and risk zoning and source location specifically involves: calculating the relative deviation between the monitored feature quantity and the dynamically adjusted judgment threshold, eliminating the dimensional differences between different monitored feature quantities, setting a three-level progressive early warning level based on the magnitude of the relative deviation, with different early warning levels corresponding to different information prompting methods; simultaneously, using the unique location number of the zone, directly and accurately locking the specific axial position of the die lip corresponding to the risk, and synchronously outputting the real-time material temperature data and the calculation results of the monitored feature quantity at that position, thus completing the accurate source location of the risk;

[0042] The hierarchical early warning and risk zoning source location are implemented through the following algorithm formula:

[0043] ;

[0044] ;

[0045] in, For the number The partition at the time of collection The degree of relative deviation between the monitored feature quantity and the dynamic judgment threshold. For the number The calculated values ​​of the monitoring characteristic quantities corresponding to the partitions. For the time of data collection The dynamic threshold matrix, , The preset deviation level classification threshold, It is at the warning level.

[0046] Preferably, the quantitative confirmation of the completed edge sealing status is specifically as follows: based on the weighted coupling operation of multi-dimensional monitoring feature quantities, the quantitative edge sealing completion degree with a value range of 0 to 1 is output, and the sum of the weighting coefficients corresponding to each monitoring feature quantity is 1; when the edge sealing completion degree reaches the preset standard and all monitoring feature quantities are within the safe threshold range, the confirmation signal of edge sealing completion is output, which serves as the sole basis for determining whether to enter the subsequent overall cooling process;

[0047] The quantitative edge banding completion degree is calculated using the following algorithm formula:

[0048] ;

[0049] in, For the time of data collection The degree of completion of the quantitative edge banding , , , The weighting coefficients are the corresponding values ​​for each monitored characteristic. For the time of data collection The synchronization deviation of the melt solidification front advance For the time of data collection The dynamic attenuation coefficient of the edge sealing integrity, For the time of data collection The average value of the characteristic values ​​of the all-axial temperature field gradient change. This represents the maximum permissible value of the characteristic value of the temperature field gradient change. For the time of data collection Cumulative value of thermal stress fatigue, This represents the maximum permissible value for the cumulative thermal stress fatigue.

[0050] Preferably, the full-process monitoring data archiving specifically involves: uniformly archiving and storing the full-process collected data, coupled calculation results, and hierarchical early warning information of this critical sub-stage to construct a full life-cycle management database for the mold head;

[0051] Archived data is used for subsequent dynamic monitoring of the benchmark model for iterative optimization, trend analysis of the die head's operating status, and prediction of the degree of die head life decay.

[0052] The dynamic monitoring system based on multi-stage material temperature of sheet extrusion includes a critical sub-stage identification and benchmark calibration module with sequentially connected signals, a multi-zone synchronous data acquisition module, a multi-variable coupled calculation core module, a dynamic threshold adaptive adjustment module, a graded early warning and source tracing module, and a completion confirmation and data archiving module.

[0053] The critical sub-stage identification and benchmark calibration module is used to collect extruder operating status signals, identify critical sub-stage start-up nodes, and generate a dynamic monitoring benchmark model adapted to the working conditions.

[0054] The multi-zone synchronous data acquisition module is used to perform full synchronous acquisition of material temperature in multiple zones of the die lip, and generate a structured material temperature dataset with location and timestamp.

[0055] The multivariable coupled calculation core module is used to coupled and calculate multi-dimensional monitoring feature quantities of critical sub-stages based on multi-segment material temperature datasets.

[0056] The dynamic threshold adaptive adjustment module is used to adaptively and dynamically adjust the judgment threshold of the monitoring benchmark based on the real-time process of melt solidification.

[0057] The hierarchical early warning and source tracing module is used to compare the monitoring feature quantity with the dynamic threshold and perform hierarchical early warning and risk zoning source tracing.

[0058] The completion confirmation and data archiving module is used to output the edge sealing completion confirmation result and complete the archiving, storage and full lifecycle management of the monitoring data throughout the entire process.

[0059] Compared with the prior art, the beneficial effects of the present invention are:

[0060] 1. This invention establishes a closed-loop, multi-segment dynamic material temperature monitoring method by precisely identifying the critical sub-stage of die lip sealing during planned shutdown and cooling. It is the first to treat the critical sub-stage of die lip sealing as an independent monitoring cycle for technical control. Through multi-condition joint judgment logic, the start and end boundaries of this stage are precisely locked, completely avoiding the subjectivity, lag, and operational inconsistencies caused by manual experience-based judgment. Simultaneously, based on pre-recorded die structural attributes, thermal properties of the processed material, and real-time environmental conditions, a dynamic monitoring benchmark model fully adapted to the current shutdown conditions is generated. Then, through synchronous acquisition and coupled calculation of multi-segment material temperature across the entire axial region of the die lip, online quantitative monitoring of the melt solidification process, die temperature field distribution, and sealing layer integrity within this critical sub-stage is achieved. This allows for real-time capture of micron-level die flexure deformation risks and micro-sealing gap hazards, providing core technical support for the standardized, precise, and intelligent management of planned shutdown processes in sheet extrusion.

[0061] 2. This invention also achieves refined and effective all-time control over the entire edge sealing process by using an adaptive dynamic adjustment mechanism for the judgment threshold based on the real-time progress of melt solidification and a coupled calculation system of multi-dimensional monitoring features. In the critical sub-stage of edge sealing at the die lip, the solidification state of the melt continuously changes dynamically over time. The risk characteristics and reasonable judgment criteria of different solidification stages differ significantly. This invention progressively divides the entire monitoring cycle into stages by averaging the melt solidification process along the entire lip axis. It matches specific judgment thresholds that are perfectly adapted to the solidification characteristics of each monitoring stage, and simultaneously corrects the thresholds based on fluctuations in the real-time environmental conditions, ensuring that the monitoring judgment thresholds continuously adapt to the current working conditions. Furthermore, through multi-segment material temperature coupled calculations with location and time-series dual-dimensional correlation, a multi-dimensional monitoring feature system is constructed, enabling accurate identification and quantitative characterization of different types of risks. This avoids the technical drawbacks of fixed thresholds and single-dimensional monitoring, significantly improving the anti-interference capability, reliability, and accuracy of risk prediction in the monitoring process.

[0062] 3. This invention achieves standardized quantitative confirmation of edge sealing completion through weighted coupling calculation of multi-dimensional monitoring features, providing a unified and objective basis for the automated and standardized operation of planned downtime processes. At the same time, through the unified archiving and storage of data collected throughout the critical sub-stage, coupled calculation results, and hierarchical early warning information, a full life cycle management database for the die head is constructed. The archived historical data can be directly used for iterative optimization of the dynamic monitoring benchmark model, trend analysis of the die head's operating status, and prediction of lifespan decay. This not only achieves precise monitoring and control of single downtime processes but also constructs a complete technical system that can be iteratively optimized over the long term. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0064] Figure 2 This is a schematic diagram of the system framework of the present invention;

[0065] Figure 3 This is a schematic diagram illustrating the triggering logic of the critical sub-stage start / stop determination and monitoring system of the present invention.

[0066] Figure 4 This is a schematic diagram illustrating the process of generating the dynamic monitoring benchmark model and adaptively adjusting the threshold in this invention.

[0067] Figure 5 This is a schematic diagram of the process for implementing the hierarchical early warning and risk zoning source location of the present invention. Detailed Implementation

[0068] To facilitate understanding of the technical solution of the present invention by those skilled in the art, the technical solution of the present invention will now be further described in conjunction with the accompanying drawings.

[0069] Example 1, such as Figures 1-5 As shown, this invention provides a dynamic monitoring method for multi-stage material temperature in sheet extrusion, comprising the following steps:

[0070] S1. Real-time acquisition of the extruder's overall operating status signals, accurate identification of the critical sub-stage start-up node for die lip sealing and finishing, and triggering the start-up of the monitoring system;

[0071] S2. Based on the pre-entered mold head structure attributes, processing material thermal performance attributes, and real-time environmental conditions, generate a dynamic monitoring benchmark model for this shutdown condition.

[0072] S3. Perform full synchronous acquisition of the material temperature of all sections in the axial direction of the die lip to obtain a real-time material temperature dataset that corresponds one-to-one with the section position.

[0073] S4. Based on the real-time acquisition of multi-segment material temperature datasets, coupled calculation of multi-dimensional monitoring feature quantities of critical sub-stages;

[0074] S5. Based on the real-time progress of melt solidification, the judgment threshold of the monitoring benchmark model is adaptively and dynamically adjusted.

[0075] S6. Compare the real-time calculated monitoring features with the dynamically adjusted judgment thresholds in real time, and perform hierarchical early warning and risk zoning source location.

[0076] S7. Based on the coupled calculation results of multi-dimensional monitoring features, complete the quantitative confirmation of the edge sealing status and the archiving of full-process monitoring data.

[0077] In an embodiment of the present invention, the critical sub-stage of sealing the die lip is defined by the following time boundary: the extruder screw stops feeding, the metering section of the barrel is emptied, the screw drops to a preset idle speed, and no new material is continuously extruded from the die lip. The end time is defined by the formation of a stable and continuous sealing layer of the lip melt, the attainment of a preset solidification critical state, and the formal stop of the screw.

[0078] The critical sub-stage is an independent monitoring cycle in the planned shutdown process of sheet extrusion, separate from the steady-state production stage and the subsequent overall cooling stage. It only enters when all the state conditions at the start time are met and terminates when all the state conditions at the end time are met.

[0079] The start and termination states of the critical sub-stage are quantitatively determined using the following algorithm formula:

[0080] ;

[0081] ;

[0082] ;

[0083] in, The comprehensive judgment value is the critical sub-stage start time. It is the logical product of all individual judgment values ​​of the start conditions. The comprehensive judgment value is 1 only when all individual judgment values ​​are 1, which means that all start conditions are met.

[0084] This is the individual judgment value for each state condition corresponding to the start time. The value is 1 when the corresponding individual start condition is met, and 0 when it is not met.

[0085] The comprehensive judgment value at the end of the critical sub-stage is the logical product of the individual judgment values ​​of all termination conditions. The comprehensive judgment value is 1 only when all individual judgment values ​​are 1, which means that all termination conditions are met.

[0086] This is the individual judgment value for each state condition corresponding to the end time. The value is 1 when the corresponding individual end condition is met, and 0 when it is not met.

[0087] This is the trigger flag for the monitoring system, and it is the output value of the piecewise function. A value of 1 indicates that the critical sub-stage has been effectively started and the monitoring system is officially running. A value of 0 indicates that the monitoring system has not been started or has been terminated.

[0088] This set of formulas achieves a multi-condition "AND gate" joint determination through logical multiplication. First, it calculates the comprehensive determination value of the critical sub-stage start and termination states separately, and then determines the trigger flag of the monitoring system through a piecewise function. Only when all start conditions are met simultaneously and the termination condition is not triggered is the monitoring system determined to be effectively started; otherwise, the monitoring system remains in the closed state, thus logically achieving precise boundary locking of the critical sub-stage.

[0089] In an embodiment of the present invention, the dynamic monitoring benchmark model includes a melt solidification critical state judgment benchmark, a lip axial temperature difference allowable range benchmark, a solidification front advancement synchronicity benchmark, a thermal stress safety judgment benchmark, and a sealing integrity judgment benchmark.

[0090] The various benchmarks are interconnected and matched, and the operating parameters based on the current shutdown condition are generated synchronously to form a unified monitoring and judgment system. This is different from the fixed monitoring benchmark for steady-state production conditions, which is a constant threshold that does not change with the production process.

[0091] The benchmark thresholds of the dynamic monitoring benchmark model can be generated using the following algorithm formula:

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] in, It serves as the critical temperature reference for melt solidification and as the core temperature threshold for determining the solidification state of melt.

[0098] It serves as the benchmark for the allowable range of axial temperature difference at the lip, and as the threshold for determining the uniformity of the temperature field at the die lip.

[0099] To serve as a synchronous benchmark for the solidification front and a preset benchmark curve for the melt solidification process;

[0100] It serves as the benchmark for judging thermal stress safety and as the upper limit threshold for the thermal stress of the die head.

[0101] This serves as the benchmark for determining the integrity of the seal, and as the benchmark threshold for the integrity of the sealing layer at the edge.

[0102] For the thermal properties of processed materials, it is a comprehensive set of parameters for the thermal characteristics of materials;

[0103] For the structural properties of the die head, it is a comprehensive parameter set of the die head structure and material properties;

[0104] This represents the real-time environmental status and the ambient temperature parameter during the monitoring process.

[0105] , , , , The parameter mapping functions correspond to various monitoring benchmarks and are used to generate corresponding benchmark thresholds based on input operating condition parameters.

[0106] This set of formulas enables customized generation of monitoring benchmarks based on operating conditions, which differs from the generalized setting of traditional fixed thresholds. It ensures that all benchmark thresholds are fully compatible with the material, mold head, and environmental conditions of the current shutdown, while the synchronous generation of multiple benchmarks ensures the logical consistency of each monitoring dimension.

[0107] In an embodiment of the present invention, the multi-segment material temperature of all axial partitions of the die lip is collected synchronously. Specifically, in view of the dynamic change characteristics of the melt solidification state within the critical sub-stage, the temperature acquisition units built into each partition of the die lip are collected synchronously at the same frequency and without time difference. The acquisition time and the unique position number of the corresponding partition are recorded synchronously to form a real-time material temperature dataset with position and timestamp.

[0108] Synchronous data acquisition with the same frequency and no time difference is achieved by controlling the acquisition action of each zone temperature acquisition unit through a unified clock trigger signal, ensuring that the time base of all acquired data is consistent. The real-time material temperature dataset with location and timestamp is arranged into structured data according to the acquisition sequence and zone location.

[0109] The real-time material temperature dataset collected synchronously is constructed using the following algorithm formula:

[0110] ;

[0111] ;

[0112] in, A unique position number is assigned to each axially equidistantly distributed partition of the die lip to identify the axial position of the material temperature data.

[0113] To provide a unified time reference for all material temperature data by unifying the timing of synchronous acquisition triggered by a unified clock;

[0114] For the number The partition at the time of collection Synchronously collected material temperature data;

[0115] The total number of axial partitions of the mold lip provides a total quantity benchmark for calculating the temperature field across the entire axis.

[0116] The total number of synchronous data collections within a single monitoring cycle provides a total duration benchmark for time series dimension calculation;

[0117] The dataset is a structured real-time material temperature dataset with location and timestamp, and is a two-dimensional matrix structure, with rows corresponding to partition locations and columns corresponding to acquisition times.

[0118] This serves as a trigger flag for the monitoring system, used to control the start and stop of the material temperature acquisition process;

[0119] This formula group first uses a piecewise function to link the trigger state of the monitoring system with the material temperature acquisition action, and only performs valid material temperature data acquisition when the monitoring system is effectively started. Then, it constructs a structured material temperature dataset with location and timestamp through a two-dimensional matrix structure, realizing the temporal and location-based association and binding of material temperature data, and providing a standardized data source for subsequent coupled calculations.

[0120] In an embodiment of the present invention, the multi-dimensional monitoring features of the critical sub-stage are obtained by linkage and coupling calculation based on the multi-segment material temperature dataset;

[0121] Linked and coupled calculations are used to establish a temperature field linkage relationship across the entire axis based on the location and temporal correlation of material temperature data in each zone, and to obtain monitoring characteristic quantities through cross-operation of multiple data.

[0122] The interconnected calculation of the multi-segment material temperature dataset is achieved through the following algorithm formula:

[0123] ;

[0124] in, The output value is the coupled calculation result of the multi-dimensional monitoring feature quantity, and the feature quantity calculation result is the corresponding monitoring dimension.

[0125] This is a coupled computation function, a general function framework for multi-dimensional feature calculation, with different function implementation logic corresponding to different monitoring dimensions;

[0126] For the number The partition at the time of collection Material temperature data;

[0127] To match the number Adjacent partitions are acquired at the same time. The material temperature data is used to perform axial position correlation calculations at the same time.

[0128] For the number The temperature data of the partition at the adjacent previous acquisition time is used to realize the time-series correlation calculation at the same location;

[0129] This formula defines a general coupled calculation framework for multi-dimensional monitoring features. By integrating material temperature data from adjacent zones at the same acquisition time and material temperature data from adjacent acquisition times in the same zone, and combining the correlation between zone location and acquisition time, cross-operation is achieved. It clarifies that the monitoring features must be obtained based on the linkage calculation of multiple segments of material temperature data, and single-point material temperature data cannot meet the input requirements of the calculation.

[0130] In an embodiment of the present invention, the multi-dimensional monitoring feature includes the synchronization deviation of the melt solidification front advancement. Specifically, it is: based on the real-time material temperature data of each zone, the melt solidification process of the corresponding zone is calculated, the continuous advancement curve of the melt solidification front along the lip axial direction is fitted, and the synchronization deviation between the actual advancement curve and the preset benchmark curve in the monitoring benchmark is calculated to quantify the risk of the formation of sealing micro-gap.

[0131] The melt solidification process is calculated based on the material temperature data and material thermal properties of the corresponding zone. The continuous advancement curve is a continuous function along the lip axis, which can intuitively reflect the axial distribution of the melt solidification state. The synchronization deviation directly corresponds to the specific position of the lip axis.

[0132] The synchronization deviation of the melt solidification front advance is calculated using the following algorithm formula:

[0133] ;

[0134] ;

[0135] in, For the number The partition at the time of collection The solidification process of the melt, with a value ranging from 0 to 1. The larger the value, the higher the degree of melt solidification.

[0136] The critical temperature reference for melt solidification is derived from the dynamic monitoring reference model.

[0137] For the number The partition at the time of collection Material temperature data;

[0138] To monitor the ambient temperature during the process;

[0139] For the time of data collection The synchronous deviation of the melt solidification front advance is used to quantify the uniformity of melt solidification along the entire axis;

[0140] This represents the total number of axial partitions at the die head lip.

[0141] For the time of data collection The preset baseline solidification process;

[0142] This group converts the material temperature data of different zones into melt solidification process values ​​of a uniform scale. Then, by averaging the absolute values ​​of the deviations between the solidification process of the entire zone and the baseline process, the synchronous deviation of the solidification front advancement along the entire axis is calculated, thereby realizing the continuous quantification of the melt solidification state and the determination of the uniformity of the entire axis.

[0143] In an embodiment of the present invention, the multi-dimensional monitoring feature includes the axial gradient change feature value of the lip temperature field. Specifically, it is: based on the real-time material temperature data of multiple zones, the continuous temperature field distribution curve of the lip axis is fitted, and the rate of change and second-order change feature of the temperature field along the axial unit zone are calculated to quantify the risk of asymmetric flexural deformation of the die lip.

[0144] The continuous temperature field distribution curve is obtained by fitting the material temperature data collected discretely from each zone with equal axial spacing of the lip. The axial spacing between adjacent zones is a fixed value, which reflects the continuous temperature change in the lip axis. The second-order change characteristics of the temperature field gradient reflect the degree of asymmetric distribution of the temperature field and the corresponding thermal deformation trend of the die lip.

[0145] The characteristic value of the axial gradient variation of the temperature field at the lip can be calculated using the following algorithm:

[0146] ;

[0147] ;

[0148] in, The fixed spacing between adjacent axial sections of the die lip provides a length reference for gradient calculation;

[0149] For the number The partition at the time of collection The first-order axial gradient of the lip temperature field reflects the rate of temperature change per unit length along the lip axis.

[0150] , For two adjacent partitions to be acquired at the same time Material temperature data;

[0151] For the number The partition at the time of collection The characteristic value of the axial gradient change of the temperature field, namely the second-order axial gradient of the temperature field, reflects the degree of asymmetry in the temperature field distribution.

[0152] This set of formulas achieves refined calculation of temperature field gradient through second-order difference operation. First, the first-order gradient of the temperature field along the axial direction is calculated by the ratio of the temperature difference between adjacent zones to the fixed spacing, reflecting the rate of temperature change along the lip axis. Then, the second-order gradient of the temperature field along the axial direction is calculated by the ratio of the difference of the first-order gradient to the fixed spacing, which is the characteristic value of the temperature field gradient change, reflecting the degree of asymmetric distribution of the temperature field.

[0153] In an embodiment of the present invention, the multi-dimensional monitoring feature includes the dynamic attenuation coefficient of the sealing integrity of the edge sealing, specifically: based on the synchronous deviation of the melt solidification front advancement and the real-time solidification state of the melt in each zone, the dynamic integrity of the continuous sealing layer of the lip is coupled and calculated to quantify the risk of sealing failure.

[0154] Dynamic integrity is calculated continuously based on the melt solidification state of the lip along the entire axis, reflecting the continuity of the sealing layer in real time. The risk of sealing failure is related to the attenuation of dynamic integrity.

[0155] The dynamic attenuation coefficient of edge sealing integrity is calculated using the following algorithm formula:

[0156] ;

[0157] in, For the time of data collection The dynamic attenuation coefficient of the sealing integrity of the edge sealing layer ranges from 0 to 1. The closer the value is to 1, the higher the integrity of the sealing layer.

[0158] For the time of data collection The synchronization deviation of the melt solidification front advance;

[0159] For the time of data collection The preset baseline solidification process;

[0160] Using a preset benchmark solidification process as a reference, the attenuation of seal integrity is quantified by the ratio of solidification synchronization deviation to benchmark process. Then, by subtracting this attenuation from 1, the dynamic attenuation coefficient of seal integrity is obtained, thus realizing continuous and dynamic quantification of seal integrity.

[0161] In an embodiment of the present invention, the multi-dimensional monitoring feature includes the cumulative value of thermal stress fatigue at the lip, specifically: based on real-time temperature difference data of multiple zones, thermophysical properties of the die head material and historical shutdown monitoring data, the thermal stress concentration value and cumulative fatigue damage value of the current shutdown condition are coupled and calculated to quantify the degree of life decay of the die head throughout its entire life cycle.

[0162] The thermal stress concentration value is calculated based on the temperature difference distribution along the axial direction of the lip, and the cumulative fatigue damage value is calculated by combining the thermal stress data of this working condition with historical cumulative data, thus quantifying the fatigue damage process of the die lip.

[0163] The cumulative value of thermal stress fatigue at the lip is calculated using the following formula:

[0164] ;

[0165] ;

[0166] in, For the number The partition at the time of collection The thermal stress concentration value reflects the magnitude of the thermal stress at the corresponding location;

[0167] The elastic modulus of the die head material is derived from the die head structural property parameter set.

[0168] The coefficient of thermal expansion of the die material is derived from the die structure property parameter set.

[0169] For the number The partition at the time of collection The temperature field has an axial second-order gradient;

[0170] The fixed spacing between adjacent sections along the axial direction of the die lip;

[0171] For the time of data collection The cumulative value of thermal stress fatigue at the lip reflects the degree of cumulative fatigue damage to the die head;

[0172] The fatigue damage values ​​accumulated from historical downtime conditions are derived from the die head's entire lifecycle database;

[0173] This refers to the total number of synchronous data collections within a single monitoring cycle.

[0174] This represents the total number of axial partitions at the die head lip.

[0175] The thermal stress safety assessment benchmark is derived from the aforementioned dynamic monitoring benchmark model;

[0176] First, the real-time thermal stress concentration value of a single zone is calculated by coupling the material properties of the die head, the second-order gradient of the temperature field, and the spacing parameters. Then, the real-time thermal stress fatigue cumulative value is calculated by accumulating the ratio of thermal stress to the safety benchmark across all zones and time series, combined with historical fatigue damage values, thereby quantifying the fatigue damage process of the die head throughout its entire life cycle.

[0177] In an embodiment of the present invention, the judgment threshold of the monitoring benchmark model is adaptively and dynamically adjusted. Specifically, the critical sub-stage is divided into multiple continuous progressive monitoring stages according to the real-time progress of melt solidification. A corresponding judgment threshold is matched for each monitoring stage. At the same time, the judgment threshold is corrected in real time based on the fluctuations of the real-time environmental state and the initial state of the die head.

[0178] Multiple continuous progressive monitoring stages correspond one-to-one with the real-time process of melt solidification. The judgment thresholds for different monitoring stages are matched and set according to the solidification characteristics of that stage. The real-time correction of the thresholds is performed synchronously based on the real-time changes in the environment and the state of the die head, ensuring that the judgment thresholds are adapted to the current working conditions.

[0179] The adaptive dynamic adjustment of the threshold for the monitoring benchmark model is achieved through the following algorithm formula:

[0180] ;

[0181] ;

[0182] in, For the time of data collection The average melt solidification process along the entire axial direction of the lip reflects the overall solidification progress.

[0183] This represents the total number of axial partitions at the die head lip.

[0184] For the number The partition at the time of collection The process of melt solidification;

[0185] For the time of data collection The adaptively adjusted dynamic threshold matrix contains the dynamic thresholds corresponding to each monitored feature.

[0186] The initial benchmark threshold matrix corresponding to each monitoring feature is derived from the aforementioned dynamic monitoring benchmark model;

[0187] This is a stage adjustment coefficient based on the average solidification process, with different coefficient values ​​corresponding to different solidification process intervals;

[0188] This is a correction factor based on the real-time environmental conditions;

[0189] Real-time ambient temperature;

[0190] First, the average melt solidification process along the entire lip axis is obtained by averaging the solidification process of the entire zone. Based on this, progressive monitoring stages are divided. Then, based on the initial benchmark threshold, combined with the stage adjustment coefficient and environmental state correction coefficient corresponding to the solidification process, a dynamic judgment threshold matrix is ​​calculated in real time to achieve adaptive adjustment of the threshold as the working condition and environment change.

[0191] In an embodiment of the present invention, the implementation of graded early warning and risk zoning and source location is as follows: based on the degree of deviation between the monitored feature quantity and the judgment threshold, a multi-level progressive early warning level is set, with different early warning levels corresponding to different prompting methods. At the same time, the module lip zoning number corresponding to the risk is accurately located, and the risk type and development trend are clarified.

[0192] The multi-level progressive early warning level is set step by step based on the degree of deviation. Different early warning levels correspond to different information output methods. The risk zone tracing and positioning directly corresponds to the specific zone position in the lip axis, and simultaneously outputs the real-time material temperature data and the calculation results of the monitoring characteristic quantity at that position.

[0193] Tiered early warning and risk zoning and source tracing are implemented through the following algorithm formula:

[0194] ;

[0195] ;

[0196] in, For the number The partition at the time of collection The relative deviation between the monitored characteristic quantity and the dynamic judgment threshold is a dimensionless value.

[0197] For the number The calculated values ​​of the monitoring features corresponding to the partitions cover the monitoring features of all the aforementioned dimensions;

[0198] For the time of data collection The dynamic threshold matrix for judgment;

[0199] , Preset threshold values ​​for the degree of deviation are used to classify different warning levels;

[0200] The warning level is represented by values ​​1, 2, and 3, which correspond to a progressive warning level of Level 1, Level 2, and Level 3, respectively.

[0201] First, the relative deviation of the monitored feature quantity and the dynamic threshold is calculated through relative deviation calculation to eliminate the dimensional differences of different monitored feature quantities. Then, a three-level progressive early warning level is divided according to the magnitude of the deviation through a piecewise function. At the same time, the specific axial position corresponding to the risk is directly locked through the partition number, so as to realize the graded early warning and precise source tracing of the risk.

[0202] In the embodiments of the present invention, the quantitative confirmation of the edge sealing status and the archiving of full-process monitoring data are completed as follows: based on the coupled calculation results of multi-dimensional monitoring feature quantities, the quantitative edge sealing completion result is output; when the edge sealing completion reaches the preset standard and all monitoring feature quantities are within the safe threshold range, the confirmation signal of edge sealing completion is output.

[0203] At the same time, the data collected throughout the critical sub-stage, the calculation results, and the early warning information will be archived and stored to form a database for the full life cycle management of the mold head.

[0204] The edge sealing completion result is obtained based on the weighted coupling calculation of multi-dimensional monitoring features. The preset standard is the judgment condition for edge sealing completion, which must simultaneously meet the completion requirements and the safety requirements of all monitoring features. The full life cycle management database can realize the unified management and retrieval of monitoring data from each time.

[0205] The quantification degree of edge sealing completion is calculated using the following algorithm formula:

[0206] ;

[0207] in, For the time of data collection The quantitative edge banding completion rate ranges from 0 to 1. The closer the value is to 1, the closer the edge banding is to completion.

[0208] , , , These are the weighting coefficients corresponding to each monitored characteristic quantity, and the sum of all weighting coefficients is 1;

[0209] For the time of data collection The synchronization deviation of the melt solidification front advance;

[0210] For the time of data collection The dynamic attenuation coefficient of the edge sealing integrity;

[0211] For the time of data collection The average value of the characteristic values ​​of the all-axial temperature field gradient change;

[0212] This represents the maximum permissible value of the characteristic value of the temperature field gradient change.

[0213] For the time of data collection The cumulative value of thermal stress fatigue;

[0214] This represents the maximum permissible value for the cumulative thermal stress fatigue.

[0215] By performing weighted coupling operations on multi-dimensional monitoring features, a comprehensive quantitative edge sealing completion degree is obtained. All input features are dimensionless, and the sum of the weighting coefficients of each feature is 1, ensuring that the completion degree value is between 0 and 1, thus achieving a multi-dimensional comprehensive quantitative judgment of the edge sealing status.

[0216] In an embodiment of the present invention, the full-process monitoring data archiving specifically involves: uniformly archiving and storing the full-process collected data, coupled calculation results, and hierarchical early warning information of this critical sub-stage to construct a full life-cycle management database for the mold head; the archived data can be used for iterative optimization of the dynamic monitoring benchmark model, trend analysis of the mold head's operating status, and prediction of the degree of mold head life decay, thereby realizing the full-cycle reuse of monitoring data.

[0217] Example 2, as follows Figures 1-5 As shown, the present invention provides a dynamic monitoring system for multi-stage material temperature based on sheet extrusion, including: a critical sub-stage identification and benchmark calibration module, a multi-zone synchronous data acquisition module, a multi-variable coupled calculation core module, a dynamic threshold adaptive adjustment module, a graded early warning and traceability positioning module, and a completion confirmation and data archiving module;

[0218] The critical sub-stage identification and benchmark calibration module is used to collect extruder operating status signals, identify critical sub-stage start-up nodes, and generate a dynamic monitoring benchmark model.

[0219] The multi-zone synchronous data acquisition module is used to perform full synchronous acquisition of material temperature in multiple zones of the die lip, and generate a real-time material temperature dataset with location and timestamp.

[0220] The multivariable coupled calculation core module is used to coupled and calculate multi-dimensional monitoring feature quantities of critical sub-stages based on multi-segment material temperature datasets.

[0221] The dynamic threshold adaptive adjustment module is used to adaptively and dynamically adjust the judgment threshold of the monitoring benchmark based on the real-time process of melt solidification.

[0222] The graded early warning and source tracing module is used to compare monitoring feature quantities with dynamic thresholds and perform graded early warning and risk zoning source tracing.

[0223] The completion confirmation and data archiving module is used to output the edge sealing completion confirmation results and complete the archiving, storage and lifespan management of the monitoring data throughout the entire process.

[0224] In an embodiment of the present invention, the multi-zone synchronous data acquisition module is signal-connected to the temperature acquisition unit built into each zone of the mold lip, which can realize multi-channel synchronous data acquisition without phase difference, and the acquisition frequency and acquisition timing can be adaptively adjusted according to the progress of the critical sub-stage.

[0225] In the embodiments of the present invention, the multivariable coupled calculation core module has a built-in preset coupled calculation model, which can complete the real-time parallel calculation of multi-dimensional monitoring characteristic quantities based on the linkage data of multiple material temperatures, and the calculation results can be output to other modules in real time.

[0226] In embodiments of the present invention, the graded early warning and source tracing module has built-in multi-level early warning rules, which can match the corresponding output method according to the risk level, and can accurately mark the location of the lip section of the model corresponding to the risk and the risk details in the visual interface.

[0227] In an embodiment of the present invention, the completion confirmation and data archiving module has a built-in mold head full life cycle management database, which can store full process monitoring data of each downtime condition and support historical data backtracking, trend analysis and lifespan decay prediction.

[0228] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A dynamic monitoring method for multi-stage material temperature in sheet extrusion, characterized in that, Includes the following steps: S1 collects real-time operating status signals of the extruder and accurately identifies the critical sub-stage start-up node of the die lip sealing and finishing, triggering the start of the monitoring system. S2 generates a dynamic monitoring benchmark model adapted to the current shutdown condition based on the pre-entered mold head structural attributes, processing material thermal performance attributes, and real-time environmental conditions. S3 performs fully synchronous acquisition of the material temperature of all sections along the axial direction of the die lip, and obtains a real-time material temperature dataset with timestamps that corresponds one-to-one with the section position; S4, based on the collected multi-segment material temperature dataset, calculates the multi-dimensional monitoring feature quantities of the critical sub-stage by coupling the correlation between location and time. S5 performs adaptive dynamic adjustment of the judgment threshold of the monitoring benchmark model based on the real-time progress of melt solidification. S6 compares the real-time calculated monitoring features with the dynamically adjusted judgment thresholds in real time to perform hierarchical early warning and risk zoning and source tracing. Based on the coupled calculation results of multi-dimensional monitoring features, S7 completes the quantitative confirmation of the edge sealing status and the archiving of monitoring data throughout the entire process.

2. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The critical sub-stage for sealing the die lip is defined by the following time boundaries: the extruder screw stops feeding, the metering section of the barrel is emptied, the screw drops to a preset idle speed, and no new material is continuously extruded from the die lip. The critical sub-stage is defined by the following time boundaries: the melt at the lip forms a stable and continuous sealing layer, a preset solidification critical state is reached, and the screw officially stops rotating. The critical sub-stage is an independent monitoring cycle separate from the steady-state production stage and the subsequent overall cooling stage. It is entered only when all the conditions at the start time are met, and terminated when all the conditions at the end time are met. The start and end states are precisely locked through a joint determination method of multiplying multiple conditions. The start and termination states of the critical sub-stage are quantitatively determined using the following algorithm formula: ； ； ； in, This is the comprehensive judgment value for the critical sub-stage start-up time. These are the individual judgment values ​​for each state condition corresponding to the startup time. This is the comprehensive judgment value at the end of the critical sub-stage. These are the individual judgment values ​​for each state condition corresponding to the end time. This serves as a trigger for the monitoring system.

3. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The dynamic monitoring benchmark model includes the melt solidification critical state judgment benchmark, the lip axial temperature difference allowable range benchmark, the solidification front advancement synchronicity benchmark, the thermal stress safety judgment benchmark, and the sealing integrity judgment benchmark. The benchmarks are interconnected and matched, and the operating parameters based on the current shutdown condition are generated synchronously to form a unified monitoring and judgment system. This is different from the fixed monitoring benchmarks of steady-state production conditions that do not change with the production process. The thresholds of each benchmark are generated through the mapping function of the corresponding operating parameters. The benchmark thresholds of the dynamic monitoring benchmark model are generated using the following algorithm formula: ； ； ； ； ； in, It serves as a reference for the critical solidification temperature of the melt. This serves as a benchmark for the allowable range of axial temperature difference at the lip. To solidify the benchmark for advancing synchronization at the forefront, As a benchmark for determining thermal stress safety, As a criterion for determining the integrity of the seal, For the thermal properties of the processed materials, For mold head structural properties, For real-time environmental status, , , , , Parameter mapping functions corresponding to various monitoring benchmarks.

4. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The process of synchronously acquiring the multi-segment material temperature of all axial zones of the die lip involves: targeting the dynamic changes in the solidification state of the melt within the critical sub-stage, controlling the temperature acquisition units built into each zone of the die lip with a unified clock trigger signal to perform synchronous data acquisition at the same frequency and without time difference, synchronously recording the acquisition time and the unique location number of the corresponding zone, and forming a two-dimensional structured real-time material temperature dataset according to the acquisition sequence and zone location, and performing effective data acquisition only when the monitoring system is effectively started; The synchronously acquired real-time material temperature dataset is constructed using the following algorithm formula: ； ； in, The unique location number is assigned to the partitions that are axially equidistantly distributed at intervals along the die lip. To ensure a unified clock-triggered synchronous acquisition time, For the number The partition at the time of collection Synchronously collected material temperature data; This represents the total number of axial partitions on the die head lip. This refers to the total number of synchronous data collections within a single monitoring cycle. This is a structured real-time material temperature dataset with location and timestamps. This serves as a trigger for the monitoring system.

5. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The multi-dimensional monitoring features of the critical sub-stage include four core features: the synchronous deviation of the melt solidification front, the characteristic value of the axial gradient change of the lip temperature field, the dynamic attenuation coefficient of the sealing integrity, and the cumulative value of the thermal stress fatigue of the lip. All features are obtained through a coupled calculation method of multi-data cross-operation based on the positional and temporal correlation of multi-segment material temperature datasets. Single-point material temperature monitoring data cannot establish a full-axial temperature field linkage relationship and cannot complete the independent calculation of the corresponding feature. The coupled calculation of the multi-dimensional monitoring feature quantities is implemented through the following general algorithm framework: ； in, The output value is calculated by coupling the multi-dimensional monitoring feature quantities. For coupled computation functions, For the number The partition at the time of collection Material temperature data, To match the number Adjacent partitions are acquired at the same time. Material temperature data, For the number The partition contains the material temperature data at the adjacent previous acquisition time.

6. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The adaptive dynamic adjustment of the judgment threshold of the monitoring benchmark model is specifically as follows: based on the real-time progress of melt solidification, the critical sub-stage is divided into multiple continuous progressive monitoring stages, with each monitoring stage corresponding to the real-time melt solidification process. A specific judgment threshold is matched for each monitoring stage to be adapted to the solidification characteristics of that stage. At the same time, based on the fluctuations of the real-time environmental state and the initial state of the die head, the judgment threshold is synchronously and in real-time corrected to ensure that the judgment threshold is continuously adapted to the current working condition. The adaptive dynamic adjustment of the judgment threshold is achieved through the following algorithm formula: ； ； in, For the time of data collection The lip-edge all-axial average melt solidification process, This represents the total number of axial partitions on the die head lip. For the number The partition at the time of collection The process of melt solidification, For the time of data collection The adaptively adjusted dynamic threshold matrix This is the initial baseline threshold matrix corresponding to each monitored feature. This is a stage adjustment factor based on the average solidification process. These are correction factors based on real-time environmental conditions. This represents the real-time ambient temperature.

7. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The implementation of graded early warning and risk zoning and source location is as follows: calculate the relative deviation between the monitored feature quantity and the dynamically adjusted judgment threshold, eliminate the dimensional differences between different monitored feature quantities, set a three-level progressive early warning level according to the magnitude of the relative deviation, and different information prompting methods are corresponding to different early warning levels; at the same time, through the unique location number of the zone, the specific axial position of the die lip corresponding to the risk is directly and accurately located, and the real-time material temperature data and the calculation results of the monitored feature quantity at that position are output simultaneously to complete the accurate source location of the risk. The hierarchical early warning and risk zoning source location are implemented through the following algorithm formula: ； ； in, For the number The partition at the time of collection The degree of relative deviation between the monitored feature quantity and the dynamic judgment threshold. For the number The calculated values ​​of the monitoring characteristic quantities corresponding to the partitions. For the time of data collection The dynamic threshold matrix, , The preset deviation level classification threshold, It is at the warning level.

8. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The quantitative confirmation of the completed edge sealing status is specifically as follows: based on the weighted coupling operation of multi-dimensional monitoring features, the output is a quantitative edge sealing completion degree with a value range of 0 to 1, and the sum of the weighting coefficients corresponding to each monitoring feature is 1; when the edge sealing completion degree reaches the preset standard and all monitoring features are within the safe threshold range, the confirmation signal of edge sealing completion is output, which serves as the sole basis for determining whether to enter the subsequent overall cooling process. The quantitative edge banding completion degree is calculated using the following algorithm formula: ； in, For the time of data collection The degree of completion of the quantitative edge banding , , , The weighting coefficients are the corresponding values ​​for each monitored characteristic. For the time of data collection The synchronization deviation of the melt solidification front advance For the time of data collection The dynamic attenuation coefficient of the edge sealing integrity, For the time of data collection The average value of the characteristic values ​​of the all-axial temperature field gradient change. This represents the maximum permissible value of the characteristic value of the temperature field gradient change. For the time of data collection Cumulative value of thermal stress fatigue, This represents the maximum permissible value for the cumulative thermal stress fatigue.

9. The dynamic monitoring method for multi-stage material temperature in sheet extrusion according to claim 1, characterized in that, The full-process monitoring data archiving specifically involves: uniformly archiving and storing the full-process data collected during this critical sub-stage, coupled calculation results, and hierarchical early warning information to construct a full life-cycle management database for the mold head; Archived data is used for subsequent dynamic monitoring of the benchmark model for iterative optimization, trend analysis of the die head's operating status, and prediction of the degree of die head life decay.

10. A dynamic monitoring system for multi-stage material temperature in sheet extrusion, characterized in that, It includes a critical sub-stage identification and benchmark calibration module with sequentially connected signals, a multi-zone synchronous data acquisition module, a multi-variable coupled calculation core module, a dynamic threshold adaptive adjustment module, a hierarchical early warning and source tracing module, and a completion confirmation and data archiving module; The critical sub-stage identification and benchmark calibration module is used to collect extruder operating status signals, identify critical sub-stage start-up nodes, and generate a dynamic monitoring benchmark model adapted to the working conditions. The multi-zone synchronous data acquisition module is used to perform full synchronous acquisition of material temperature in multiple zones of the die lip, and generate a structured material temperature dataset with location and timestamp. The multivariable coupled calculation core module is used to coupled and calculate multi-dimensional monitoring feature quantities of critical sub-stages based on multi-segment material temperature datasets. The dynamic threshold adaptive adjustment module is used to adaptively and dynamically adjust the judgment threshold of the monitoring benchmark based on the real-time process of melt solidification. The hierarchical early warning and source tracing module is used to compare the monitoring feature quantity with the dynamic threshold and perform hierarchical early warning and risk zoning source tracing. The completion confirmation and data archiving module is used to output the edge sealing completion confirmation result and complete the archiving, storage and full lifecycle management of the monitoring data throughout the entire process.

Images