A mold state data acquisition method, system, device and medium

Abstract

This application provides a method, system, device, and medium for acquiring mold status data, relating to the field of equipment status monitoring technology. The method includes: acquiring first status data of a target mold; generating a first anomaly coefficient of the target mold based on the first status data; acquiring a correlation coefficient between the target mold and related molds, wherein the related molds are molds that have an upstream or downstream relationship with the target mold in the process flow; acquiring second status data of the related molds; generating a second anomaly coefficient of the related molds based on the second status data; adjusting a first anomaly index by combining the correlation coefficient and the second anomaly coefficient to generate a target anomaly coefficient of the target mold; generating anomaly information of the target mold when the target anomaly coefficient is greater than a preset threshold; and generating maintenance suggestions for the target mold based on the anomaly information. The technical effect of this application is to improve the accuracy of anomaly detection.

Description

Technical Field

[0001] This application relates to the field of equipment condition monitoring technology, specifically to a method, system, equipment, and medium for acquiring mold condition data. Background Technology

[0002] In modern manufacturing, the stamping production line, as a core process in metal forming, directly impacts product quality, production efficiency, and equipment safety due to the stability and reliability of its die condition. During long-term high-load operation, stamping dies are susceptible to various factors such as material fatigue, wear and aging, and stress concentration, which can easily lead to abnormal conditions such as dimensional deviations, surface damage, and structural deformation. Failure to identify and address these anomalies promptly and accurately can result in increased product scrap rates, frequent equipment failures, and even safety accidents. Therefore, establishing an effective die condition monitoring and anomaly detection system has become a key technological requirement for the intelligent management of stamping production lines.

[0003] In existing technologies, mold condition monitoring mainly employs a method based on independent analysis of individual molds. This involves installing monitoring devices such as displacement sensors and pressure sensors on the mold to collect its physical state parameters in real time. Then, based on preset threshold ranges or statistical analysis methods, anomaly detection and evaluation are performed on the condition data of individual molds. This method can identify direct abnormal states of the mold to a certain extent, providing basic data support for equipment maintenance. However, in actual stamping production lines, the molds do not operate independently but are closely interconnected through the process flow. An abnormal state of one mold often affects the working state of other molds through process transmission. The aforementioned method of independent analysis of individual molds cannot effectively identify and quantify this inter-mold correlation, resulting in insufficient accuracy in anomaly detection. Summary of the Invention

[0004] This application provides a method, system, device, and medium for acquiring mold status data, which can improve the accuracy of anomaly detection.

[0005] In a first aspect, this application provides a method for acquiring mold status data. The method includes: acquiring first status data of a target mold sent by a height sensor installed on a stamping equipment; generating a first anomaly coefficient of the target mold based on the first status data; acquiring a correlation coefficient between the target mold and an associated mold, wherein the associated mold is a mold that has an upstream or downstream relationship with the target mold in the process flow; acquiring second status data of the associated mold; generating a second anomaly coefficient of the associated mold based on the second status data; adjusting the first anomaly coefficient by combining the correlation coefficient and the second anomaly coefficient to generate a target anomaly coefficient of the target mold; generating anomaly information of the target mold when the target anomaly coefficient is greater than a preset threshold; and generating maintenance suggestions for the target mold based on the anomaly information.

[0006] By adopting the above technical solution and establishing a multi-mold collaborative status monitoring and anomaly detection mechanism, the accuracy of mold anomaly identification and the scientific nature of maintenance decisions are significantly improved. The system generates a first anomaly coefficient by acquiring the first state data of the target mold, which can accurately reflect the basic anomaly state of the target mold itself. By acquiring the correlation coefficient between the target mold and related molds, the system scientifically quantifies the influence intensity between upstream and downstream molds in the process flow. The first anomaly coefficient is adjusted by combining the correlation coefficient and the second anomaly coefficient of the related molds to generate the target anomaly coefficient. This allows the system to comprehensively consider the state of the target mold itself and the influence effect of related molds, accurately assess the true degree of anomaly of the target mold in a complex production environment, and effectively avoid the anomaly assessment bias caused by ignoring the upstream and downstream influence of the traditional single mold independent analysis method. When the target anomaly coefficient is greater than the preset threshold, anomaly information and maintenance suggestions are automatically generated, realizing a complete closed-loop management from anomaly detection to maintenance decision-making, and improving the accuracy of anomaly detection.

[0007] Secondly, this application provides a large-scale 3D scene reconstruction system, the system comprising: a first acquisition module, a second acquisition module, a third acquisition module, a combination module, and a generation module; wherein,

[0008] The first acquisition module is used to acquire first state data of the target mold sent by a height sensor installed on the stamping equipment, and generate a first anomaly coefficient of the target mold based on the first state data; the second acquisition module is used to acquire the correlation coefficient between the target mold and an associated mold, wherein the associated mold is a mold that has an upstream or downstream relationship with the target mold in the process flow; the third acquisition module is used to acquire second state data of the associated mold, and generate a second anomaly coefficient of the associated mold based on the second state data; the combination module is used to combine the correlation coefficient and the second anomaly coefficient to adjust the first anomaly index and generate a target anomaly coefficient of the target mold; the generation module is used to generate anomaly information of the target mold when the target anomaly coefficient is greater than a preset threshold, and generate maintenance suggestions for the target mold based on the anomaly information.

[0009] Thirdly, this application provides an electronic device that adopts the following technical solution: it includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes a computer program of any of the above-described mold state data acquisition methods.

[0010] Fourthly, this application provides a computer-readable storage medium, which adopts the following technical solution: storing a computer program that can be loaded by a processor and executed by any of the above-mentioned mold state data acquisition methods.

[0011] In summary, this application includes at least one of the following beneficial technical effects:

[0012] By establishing a multi-mold collaborative status monitoring and anomaly detection mechanism, the accuracy of mold anomaly identification and the scientific nature of maintenance decisions are significantly improved. The system generates a first anomaly coefficient by acquiring the first state data of the target mold, which can accurately reflect the basic anomaly state of the target mold itself. By acquiring the correlation coefficient between the target mold and related molds, the system scientifically quantifies the influence intensity between upstream and downstream molds in the process flow. The first anomaly coefficient is adjusted by combining the correlation coefficient and the second anomaly coefficient of the related molds to generate the target anomaly coefficient. This allows the system to comprehensively consider the state of the target mold itself and the influence effect of related molds, accurately assess the true degree of anomaly of the target mold in a complex production environment, and effectively avoid the anomaly assessment bias caused by the traditional single-mold independent analysis method due to the neglect of upstream and downstream influence. When the target anomaly coefficient is greater than the preset threshold, anomaly information and maintenance suggestions are automatically generated, realizing a complete closed-loop management from anomaly detection to maintenance decision-making, and improving the accuracy of anomaly detection. Attached Figure Description

[0013] Figure 1This is a flowchart illustrating a method for acquiring mold status data according to an embodiment of this application;

[0014] Figure 2 This is a schematic diagram of a mold status data acquisition and anomaly analysis system provided in an embodiment of this application;

[0015] Figure 3 This is a schematic diagram of the structure of a mold status data acquisition system provided in an embodiment of this application;

[0016] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0017] Explanation of reference numerals in the attached figures: 1000, electronic device; 1001, processor; 1002, communication bus; 1003, user interface; 1004, network interface; 1005, memory. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0019] In the description of the embodiments in this application, words such as "illustrative," "for example," or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "illustrative," "for example," or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of words such as "illustrative," "for example," or "for example" is intended to present the relevant concepts in a specific manner.

[0020] Figure 1 This is a flowchart illustrating a method for acquiring mold status data according to an embodiment of this application. Figure 1 As shown, the method includes S101-S105:

[0021] S101, acquire the first state data of the target mold sent by the height sensor installed on the stamping equipment, and generate the first anomaly coefficient of the target mold based on the first state data.

[0022] In the stamping production process, the height change of the die is an important indicator reflecting its working status and wear degree. Traditional manual inspection methods suffer from low efficiency, poor accuracy, and inability to monitor in real time. Therefore, an automated status data acquisition and anomaly detection mechanism is needed. This invention installs height sensors at key locations on the stamping equipment to monitor the height changes of the target die in real time during operation. The target die refers to the specific die currently being monitored, distinct from other dies on the production line. The height sensor uses a laser displacement sensor or a contact displacement sensor to continuously collect distance data from the die surface to the sensor reference surface at a millisecond frequency, forming first status data including timestamps, position coordinates, and height values.

[0023] After preprocessing the initial state data, the system first extracts the baseline state data of the target mold from the database. The baseline state data refers to the standard height value of the mold calibrated under brand-new or standard working conditions. This data is typically obtained through precision measuring equipment and stored in the system when the mold is first put into use. The system calculates the deviation between the initial state data and the baseline state data. The deviation reflects the degree to which the mold's current state deviates from its ideal state. A positive deviation indicates an increase in mold height, possibly due to material accumulation or deformation. A negative deviation indicates a decrease in mold height, usually caused by wear or mechanical deformation.

[0024] Based on the magnitude and distribution characteristics of the deviation values, the system employs a multi-level evaluation mechanism to determine the deviation level, which is divided into four levels: slight deviation, moderate deviation, severe deviation, and extremely severe deviation. Slight deviation corresponds to a deviation value within ±2% of the baseline value; moderate deviation corresponds to ±2% to ±5%; severe deviation corresponds to ±5% to ±10%; and extremely severe deviation corresponds to deviations exceeding ±10%. The system also considers the trend of deviation value changes, identifying different patterns of rapid and slow changes by calculating the rate of change of deviation values ​​within a continuous time window. Rapid changes often indicate sudden failures, while slow changes reflect gradual wear.

[0025] Based on the deviation level and the preset anomaly coefficient calculation rules, the system generates a first anomaly coefficient for the target mold. This first anomaly coefficient is a value between 0 and 1, used to quantify the degree of anomaly in the current state of the mold; the closer the value is to 1, the higher the degree of anomaly. The anomaly coefficient calculation rule uses a piecewise linear function combined with weighting coefficients. For minor deviations, the anomaly coefficient is set between 0.1 and 0.3; for moderate deviations, it is set between 0.3 and 0.6; for severe deviations, it is set between 0.6 and 0.9; and for extremely severe deviations, it is set between 0.9 and 1.0. Simultaneously, the system introduces a time weighting factor, giving higher weight to recent anomaly states, while the weight of historical anomaly states decays over time, ensuring that the anomaly coefficient can promptly reflect the latest state changes of the mold.

[0026] Based on the above embodiments, as an optional implementation, in S101, generating the first anomaly coefficient of the target mold according to the first state data specifically includes S11-S13:

[0027] S11, Obtain the reference state data of the target mold. The reference state data is the standard height value of the target mold under normal working conditions.

[0028] The system first retrieves the baseline state data of the target mold from the database. This baseline state data refers to the standard height value of the target mold under normal operating conditions. This data is typically obtained during the initial installation and commissioning phase of the mold using precision measuring equipment, or calculated by analyzing stable state data during normal operation. The baseline state data includes not only a single numerical value but also the allowable normal fluctuation range, because even under normal operating conditions, the mold height will experience slight fluctuations due to factors such as temperature changes and mechanical vibrations. The system stores the baseline state data in a data structure containing standard values, upper limits, and lower limits, providing a complete reference standard for subsequent deviation calculations.

[0029] S12, calculate the deviation between the first state data and the reference state data, and determine the deviation level based on the deviation value.

[0030] Based on the acquired baseline state data, the system calculates the deviation between the first state data and the baseline state data. This deviation is a core indicator reflecting the degree to which the current state of the mold deviates from the normal standard. The system uses a relative deviation calculation method, where the deviation equals the measured value minus the baseline value, then divided by the baseline value. This approach eliminates the influence of dimensional differences between different molds, making anomaly detection more universal. The system also calculates the statistical characteristics of the deviation, including the mean deviation, standard deviation, and rate of change of deviation. These statistical indicators can more comprehensively reflect the changing characteristics of the mold's state.

[0031] Based on the calculated deviation value, the system determines the corresponding deviation level, which is a qualitative classification of the degree of mold abnormality. The system divides the deviation level into four levels: Level 1 deviation (within ±2%) indicates a slight deviation from normal; Level 2 deviation (within ±2% to ±5%) indicates a moderate deviation; Level 3 deviation (within ±5% to ±10%) indicates a more serious deviation; and Level 4 deviation (exceeding ±10%) indicates a severe abnormality. The deviation level classification criteria are based on extensive production practice data and equipment maintenance experience, effectively distinguishing between different degrees of abnormality.

[0032] S13, generate the first anomaly coefficient of the target mold according to the deviation level and the preset anomaly coefficient calculation rules.

[0033] Based on the determined deviation level and the preset anomaly coefficient calculation rules, the system generates the first anomaly coefficient for the target mold. The anomaly coefficient calculation rules use a piecewise linear mapping method to convert the deviation level into a value between 0 and 1. For level one deviation, the anomaly coefficient is set in the range of 0.1 to 0.3; for level two deviation, it is set in the range of 0.3 to 0.6; for level three deviation, it is set in the range of 0.6 to 0.9; and for level four deviation, it is set in the range of 0.9 to 1.0. The system also considers the specific position of the deviation value within each level, and determines the precise anomaly coefficient value through linear interpolation to ensure that the anomaly coefficient can continuously reflect subtle changes in the degree of anomaly.

[0034] S102, obtain the correlation coefficient between the target mold and the associated mold. The associated mold is the mold that has an upstream or downstream relationship with the target mold in the process flow.

[0035] In a stamping production line, the dies do not operate independently but rather form a close upstream-downstream relationship through the process flow. Changes in the working state of the upstream die directly affect the workload and wear of the downstream die, and similarly, anomalies in the downstream die can also have a reverse impact on the state of the upstream die. Therefore, simply monitoring the state data of the target die itself cannot fully reflect its true working condition and potential risks. To achieve more accurate anomaly detection and predictive maintenance, the system needs to establish a quantification mechanism for the influence relationship between the target die and its related dies, using correlation coefficients to describe the strength of this mutual influence.

[0036] Associated dies refer to dies that have upstream or downstream relationships with the target die in the process flow. These include directly adjacent preceding and subsequent dies, as well as dies indirectly connected through multiple process steps. The system first uses the process flow configuration information of the stamping production line. This configuration information includes detailed data such as the die number, processing sequence, and material transfer path for each station. By parsing this configuration information, the system can automatically identify all associated dies that are adjacent to or indirectly connected to the target die in the process flow. During the identification process, the system constructs a process flow graph, treating each die as a node and the material transfer relationship as directed edges, forming a complete process network topology.

[0037] Based on the process network topology, the system determines the process transfer paths between each associated mold and the target mold. The process transfer path refers to the sequence of process steps and material flow paths from the associated mold to the target mold. The system uses the shortest path algorithm from graph theory to calculate the process distance parameters between the target mold and each associated mold. These process distance parameters characterize the positional interval between molds in the process flow; the smaller the value, the stronger the correlation between the two molds in the process flow. For directly adjacent molds, the process distance parameter is set to 1; for molds passing through an intermediate station, the distance parameter is 2, and so on. Considering the complexity of material transfer, a path complexity correction factor is introduced for cases involving complex transfer paths such as branching and merging.

[0038] To quantify the actual impact of associated molds on the target mold, the system acquires historical state data records of the target mold and each associated mold within a preset time period. The preset time period typically selects data from the most recent 30 days or the most recent 100 production cycles to ensure sufficient sample size and reflect the current production status. The system employs time series analysis to calculate the correlation coefficient between the state changes of each associated mold and the state changes of the target mold. This coefficient is obtained through Pearson correlation analysis or Granger causality test, and its value ranges from -1 to 1. Positive values ​​indicate a positive impact, and negative values ​​indicate a negative impact; the absolute value reflects the intensity of the impact.

[0039] The system inputs the process distance parameter and the influence correlation coefficient into the correlation coefficient calculation function. The function uses a weighted average method, taking the inverse transformation of the process distance parameter as the location weight and the absolute value of the influence correlation coefficient as the influence intensity weight. The final correlation coefficient is generated through normalization. The correlation coefficient quantifies the overall influence of the associated mold on the target mold; a higher value indicates a more significant influence. Molds with shorter process distances and stronger historical influence correlations will have relatively higher correlation coefficients, while those with longer process distances or weaker influence correlations will have relatively lower correlation coefficients.

[0040] Based on the above embodiments, as an optional implementation, in S102, obtaining the correlation coefficient between the target mold and the associated mold specifically includes S21-S24:

[0041] S21. Based on the process flow configuration information of the stamping production line, identify the associated molds that are adjacent or indirectly connected to the target mold in the process flow, and determine the process transfer path between each associated mold and the target mold.

[0042] The system first reads the process flow configuration information of the stamping production line. This configuration information details the mold layout, process sequence, and connection relationships of the entire production line. Based on this configuration information, the system uses graph theory algorithms to identify associated molds that are adjacent or indirectly connected to the target mold in the process flow. Adjacent positions refer to upstream or downstream molds that are directly connected to the target mold in the process flow, while indirect connection positions refer to molds that form a process transfer relationship with the target mold through one or more intermediate molds. The system constructs a directed graph model of the process flow, where nodes represent each mold and edges represent the process transfer direction. A graph traversal algorithm is used to determine the process transfer path between each associated mold and the target mold. The process transfer path describes the specific path and intermediate links through which the process is transferred from the associated mold to the target mold.

[0043] S22, Calculate the process distance parameters between the target mold and each associated mold according to the process transfer path. The process distance parameters characterize the degree of positional interval between molds in the process flow.

[0044] After determining the process transfer path, the system calculates the process distance parameters between the target mold and each associated mold. These process distance parameters are crucial indicators for quantifying the positional intervals between molds in the process flow. The system sets the process distance between directly adjacent molds to 1. For each additional intermediate mold, the process distance increases by 1; for example, the process distance between the target mold and the second upstream mold is 2, and the process distance to the third upstream mold is 3. Considering the attenuation characteristics of the influence intensity during process transfer, the system also introduces an attenuation coefficient. For molds with greater distances, the influence intensity is adjusted according to an exponential attenuation law. The process distance parameters not only reflect spatial relationships but also embody the physical mechanism of process transfer, providing a theoretical basis for quantifying the correlation intensity.

[0045] S23, obtain the historical status data change records of the target mold and each associated mold within a preset time period, and calculate the correlation coefficient of the influence of the status change of each associated mold on the status change of the target mold.

[0046] To accurately assess the actual impact of each associated mold on the target mold, the system acquires historical state data records of the target mold and each associated mold within a preset time period. This preset time period typically selects data from the most recent 1-2 months to ensure sufficient data volume to support the reliability of statistical analysis. The system uses a sliding window method to segment the historical data and calculates the correlation coefficient between the state changes of each associated mold and the state changes of the target mold. This correlation coefficient reflects the actual degree of correlation between the state changes of the associated molds and the target mold. The system employs multiple statistical methods, including Pearson correlation coefficient and Granger causality test, for comprehensive analysis. Considering potential time delay effects, the system also calculates correlation coefficients at different time lags, selecting the time lag with the strongest correlation as the final impact correlation coefficient.

[0047] S24. Input the process distance parameter and the influence correlation coefficient into the correlation coefficient calculation function to generate the correlation coefficient between the target mold and each associated mold. The correlation coefficient is used to quantify the degree of influence of the associated mold on the target mold.

[0048] Based on the calculated process distance parameters and influence correlation coefficients, the system inputs these two key parameters into a preset correlation coefficient calculation function to generate correlation coefficients between the target mold and each associated mold. The correlation coefficient calculation function uses a weighted fusion approach to reasonably combine the process distance parameters and influence correlation coefficients. The specific calculation formula is: Correlation coefficient Cij = ρij × e^(-λ × dij) × αij, where ρij represents the influence correlation coefficient of associated mold i on target mold j, dij represents the process distance parameter between associated mold i and target mold j, λ is the distance attenuation factor (usually between 0.5 and 0.8), and αij is the process direction adjustment coefficient. When the associated mold is upstream of the target mold, αij is set to 1.2 to reflect the importance of the upstream influence; when the associated mold is downstream of the target mold, αij is set to 0.8 to reflect the relatively weak characteristics of the downstream reverse influence. This design considers both the actual influence intensity and the attenuation effect of process distance and the difference in influence of process flow direction. The system also normalizes the calculation results by dividing by the maximum value of the correlation coefficients of all associated molds to ensure that all correlation coefficients are between 0 and 1, which facilitates subsequent numerical calculations and result interpretation.

[0049] S103, obtain the second state data of the associated mold, and generate the second anomaly coefficient of the associated mold based on the second state data.

[0050] Because abnormal conditions in associated molds can have a cascading effect on target molds through the process transmission path, knowing only the correlation strength without understanding the actual state of the associated molds makes it impossible to accurately predict and assess the potential impact on the target mold. For example, although an upstream mold may have a high correlation coefficient with the target mold, if the upstream mold is operating normally, its impact on the target mold will be relatively small. Conversely, if the upstream mold has a serious abnormality, even a moderate correlation coefficient may still have a significant impact on the target mold.

[0051] The system acquires second-state data for each associated mold using the same data acquisition mechanism as the target mold. This second-state data refers to the height changes of the associated molds within the same time period. The acquisition method is completely consistent with the first-state data of the target mold, ensuring data synchronization and comparability. Since stamping production lines typically operate in a synchronous mode, with each mold completing its processing actions within a similar timeframe, the system needs to ensure that the acquisition time window for the second-state data remains consistent with that of the first-state data to avoid data correlation deviations caused by time differences. For processes with time delays, the system will compensate and adjust the data acquisition time accordingly based on the material transfer time in the process path.

[0052] After acquiring the second state data, the system uses the same calculation method and evaluation criteria as the first anomaly coefficient of the target mold to generate a corresponding second anomaly coefficient for each associated mold. The system first extracts the baseline state data of each associated mold from the database. Each associated mold has its own unique baseline state data because different molds have different design specifications, processing accuracy requirements, and workloads. By calculating the deviation between the second state data and the corresponding baseline state data, the system determines the deviation level of each associated mold. The criteria for classifying the deviation levels are consistent with those of the target mold, ensuring the uniformity and comparability of anomaly assessments.

[0053] In generating the second anomaly coefficient, the system specifically considers the differences in the working characteristics of different types of associated dies. For example, drawing dies, due to the greater forming force they bear, exhibit significantly different wear patterns compared to blanking dies. Therefore, a die type correction factor is introduced into the anomaly coefficient calculation rules. For forming dies subjected to high stress, the system lowers the anomaly judgment threshold, increasing its sensitivity to minute changes. For blanking dies primarily subjected to shear forces, the system focuses more on the height change patterns caused by edge wear. This differentiated evaluation mechanism ensures that the second anomaly coefficient accurately reflects the actual degree of anomaly in different types of associated dies.

[0054] The system also established a propagation model for anomalies in associated molds. By analyzing the correlation between the abnormal states of associated molds and the subsequent changes in the state of the target mold in historical data, it identifies the time delay and intensity attenuation patterns of anomaly propagation. The impact of anomalies in some associated molds may only manifest on the target mold after several production cycles, while the impact of anomalies in others may be partially absorbed or amplified by intermediate process steps. Based on these propagation patterns, the system performs time-weighted processing on the second anomaly coefficient, giving higher weight to recent anomaly states, while also considering the propagation delay time of the anomaly impact.

[0055] S104, combining the correlation coefficient and the second anomaly coefficient, adjust the first anomaly index to generate the target anomaly coefficient of the target mold.

[0056] After obtaining the first anomaly coefficient of the target mold and the second anomaly coefficients of each associated mold, the system needs to comprehensively analyze this scattered anomaly information. This is because the actual degree of anomaly in the target mold depends not only on its own state changes but also on the cumulative effect of the anomalies in associated molds. Traditional single-mold monitoring methods often ignore this correlation effect, leading to biased anomaly detection results. This can result in situations where the target mold appears normal but is actually affected by upstream anomalies, or where the target mold shows a slight anomaly but the actual risk is amplified due to the chain reaction of downstream molds. Therefore, the system needs to establish a comprehensive adjustment mechanism, using the correlation coefficient and the second anomaly coefficient as correction factors to scientifically adjust the first anomaly coefficient of the target mold.

[0057] The system first calculates the actual impact strength of each associated mold on the target mold. This is achieved by arithmetically multiplying the association coefficient of each associated mold by its corresponding second anomaly coefficient to generate the influence coefficient for each associated mold. The influence coefficient is a comprehensive indicator that considers both the objective association strength between the associated mold and the target mold, and reflects the current actual degree of anomaly in the associated mold. A higher value indicates a more significant current impact of the associated mold on the target mold. For example, if an upstream mold has a association coefficient of 0.8 and a second anomaly coefficient of 0.6, its influence coefficient is 0.48, meaning that the abnormal state of this upstream mold will have a moderate negative impact on the target mold.

[0058] Based on the influence coefficients of each associated mold, the system employs a normalization method to determine the influence weight value of each associated mold. The influence coefficient and influence weight value are directly proportional; associated molds with larger influence coefficients receive greater weight in subsequent adjustments. Normalization ensures that the sum of the influence weight values ​​of all associated molds equals 1, avoiding imbalances in weight distribution. The system also introduces an influence attenuation factor. Considering the natural attenuation effect in the process transmission path, the influence weight value of associated molds farther from the target mold decreases accordingly. This approach better aligns with the physical laws of influence transmission in actual production.

[0059] The system performs a weighted summation of the influence weights of all associated molds to generate an association influence adjustment factor. This factor, a value between 0 and 1, quantifies the overall impact of abnormal states of all associated molds on the target mold. When all associated molds are in normal condition, the association influence adjustment factor is close to 0, indicating a small association impact. When multiple associated molds simultaneously experience abnormalities, the factor increases significantly, reflecting the combined risks faced by the target mold. The system also considers the directionality of the impact. Beneficial impacts that reduce the workload of the target mold are adjusted with a negative factor, while adverse impacts that increase the workload or accelerate wear are adjusted with a positive factor.

[0060] After generating the correlation impact adjustment factor, the system employs a tiered adjustment strategy to combine the first anomaly coefficient with the correlation impact adjustment factor. The system first determines whether the correlation impact adjustment factor is greater than a preset adjustment threshold, typically set at 0.3. This threshold is an empirical value determined based on statistical analysis of extensive historical data and effectively distinguishes between significant and minor impacts. When the correlation impact adjustment factor exceeds the adjustment threshold, it indicates that the abnormal state of the associated mold has a significant impact on the target mold. The system then multiplies the correlation impact adjustment factor by a preset impact amplification coefficient to obtain the adjustment increment. The impact amplification coefficient is customized based on different production line characteristics and mold types, typically ranging from 1.2 to 2.0.

[0061] The system adds the first anomaly coefficient of the target mold to the adjustment increment value to generate the target anomaly coefficient, which reflects the true degree of anomaly of the target mold after considering the correlation effects. When the correlation effect adjustment factor is less than or equal to the adjustment threshold, it indicates that the influence of the correlation mold is low and insufficient to significantly change the anomaly assessment result of the target mold. In this case, the system directly uses the first anomaly coefficient as the target anomaly coefficient to avoid over-adjustment due to minor influences. To ensure the rationality of the target anomaly coefficient, the system also sets an upper limit constraint mechanism. When the calculation result exceeds 1.0, it is automatically adjusted to 1.0 to maintain the standardized range of the anomaly coefficient.

[0062] Based on the above embodiments, as an optional implementation, in S104, adjusting the first anomaly index by combining the correlation coefficient and the second anomaly coefficient to generate the target anomaly coefficient of the target mold specifically includes S41-S43:

[0063] S41, multiply the correlation coefficient and the second anomaly coefficient of each associated mold by arithmetic to generate the influence coefficient of each associated mold. Based on the influence coefficient, determine the influence weight value of each associated mold. The influence coefficient and the influence weight value are directly proportional.

[0064] To accurately reflect the true abnormal state of the target mold in a multi-mold collaborative working environment, the system needs to incorporate the influence of related molds into the target mold's anomaly assessment process. This is because the first anomaly coefficient calculated solely based on the target mold's own state data often fails to fully reflect its actual degree of anomaly in a complex production environment. In actual production, the abnormal state of a mold may be significantly affected by upstream and downstream related molds. For example, slight wear on an upstream mold may cause a downstream mold to experience additional stress and exhibit abnormal characteristics. Ignoring this correlation may lead to misjudgment of the target mold's anomaly degree, affecting the accuracy of maintenance decisions.

[0065] The system first quantifies the actual impact strength of each associated mold. It then arithmetically multiplies the association coefficient of each associated mold with its corresponding second anomaly coefficient to generate an impact coefficient for each associated mold. The impact coefficient reflects the actual impact strength of the associated mold on the target mold under the current anomalous state. The calculation logic is that the association coefficient represents the potential for influence transmission, while the second anomaly coefficient represents the current degree of anomalousness of the associated mold. The multiplication of the two yields the impact coefficient, which represents the actual influence of the associated mold on the target mold under the current state. For example, even if an associated mold has a high association coefficient, if its second anomaly coefficient is low, it indicates that the mold is currently in a normal state, and its actual impact on the target mold is relatively small.

[0066] S42, the influence weight values ​​of all related molds are weighted and summed to generate the related influence adjustment factor.

[0067] Based on the calculated influence coefficients of each associated mold, the system determines the influence weight value of each associated mold. This influence weight value is used to reflect the differences in importance among different associated molds in subsequent weighted calculations. The system employs a normalization method, dividing the influence coefficient of a particular associated mold by the sum of the influence coefficients of all associated molds to obtain its influence weight value, ensuring that the sum of all influence weight values ​​equals 1. This method ensures a direct proportionality between the influence coefficient and the influence weight value; associated molds with larger influence coefficients receive higher influence weight values, playing a more important role in the final anomaly coefficient adjustment process. Simultaneously, the normalization process eliminates the influence of changes in the number of associated molds on the calculation results, guaranteeing the stability and consistency of the anomaly coefficient adjustment.

[0068] The system performs a weighted summation of the influence weights of all associated molds to generate an association influence adjustment factor. This factor is a core indicator for quantifying the overall impact of all associated molds on the target mold. The specific calculation process for the weighted summation involves multiplying the influence weight of each associated mold by its corresponding second anomaly coefficient, and then summing all the products to obtain the association influence adjustment factor. The association influence adjustment factor ranges from 0 to 1. A larger value indicates a more severe negative impact of the associated molds on the target mold, requiring a greater upward adjustment to the anomaly level of the target mold. Conversely, a smaller value indicates a relatively milder impact from the associated molds, requiring a smaller adjustment to the anomaly level of the target mold.

[0069] S43, combine the first anomaly coefficient with the correlation influence adjustment factor to generate the target anomaly coefficient of the target mold.

[0070] After obtaining the correlation influence adjustment factor, the system combines the first anomaly coefficient with the correlation influence adjustment factor to generate the target anomaly coefficient of the target mold. The combination calculation uses a weighted average method, specifically calculated as: Target Anomaly Coefficient = α × First Anomaly Coefficient + β × Correlation Influence Adjustment Factor, where α and β are weighting coefficients. Typically, α is set to 0.7 and β to 0.3. This weighting allocation ensures both the dominant role of the target mold's own state and reasonably considers the influence of related molds. The system also sets upper and lower limits for the anomaly coefficient to ensure that the target anomaly coefficient remains within a reasonable range of 0 to 1. When the calculation result exceeds the upper limit, it is automatically set to 1; when the calculation result is below the lower limit, it is automatically set to 0, avoiding numerical anomalies from interfering with subsequent analysis.

[0071] Based on the above embodiments, as an optional implementation, in S43, the first anomaly coefficient and the correlation influence adjustment factor are combined to generate the target anomaly coefficient of the target mold, specifically including S431-S432:

[0072] S431, determine whether the correlation influence adjustment factor is greater than the preset adjustment threshold. When the correlation influence adjustment factor is greater than the adjustment threshold, multiply the correlation influence adjustment factor by the preset influence amplification coefficient to obtain the adjustment increment value; add the first anomaly coefficient to the adjustment increment value to generate the target anomaly coefficient of the target mold.

[0073] S432, when the correlation influence adjustment factor is less than or equal to the adjustment threshold, the first anomaly coefficient is used as the target anomaly coefficient of the target mold.

[0074] To ensure the scientific rigor and effectiveness of the correlation effect adjustment, the system needs to establish an intelligent adjustment judgment mechanism. This is because not all correlation effects require adjustment of the target mold's anomaly coefficient; only when the correlation effect reaches a certain level will it have a significant practical impact on the target mold. Adjusting even minor correlation effects may introduce unnecessary noise interference, reducing the accuracy of anomaly assessment. Conversely, ignoring significant correlation effects will lead to incomplete anomaly assessment. Therefore, the system employs a threshold judgment and hierarchical processing strategy to ensure that the corresponding anomaly coefficient adjustment is only performed when the correlation effect is sufficiently significant.

[0075] The system first assesses the magnitude of the correlation impact adjustment factor and compares it with a preset adjustment threshold. This threshold, typically set to 0.3, is a critical value determined based on statistical analysis of extensive historical data. It represents the dividing point where the correlation impact transitions from minor to significant. The determination of the adjustment threshold comprehensively considers the statistical distribution characteristics of the correlation impact, the sensitivity requirements for anomaly detection, and the fault tolerance needs of the actual production process. It aims to ensure the identification of truly meaningful correlation impacts while avoiding overreaction to normal, minor fluctuations.

[0076] When the correlation impact adjustment factor exceeds the adjustment threshold, it indicates that the correlated mold has a significant negative impact on the target mold, requiring an upward adjustment of the target mold's anomaly level. The system multiplies the correlation impact adjustment factor by a preset impact amplification coefficient to obtain the adjustment increment. The impact amplification coefficient is an important parameter for controlling the magnitude of the correlation impact adjustment, typically set between 1.5 and 2.0. This coefficient is set based on the transmission attenuation characteristics of the correlation impact and the statistical regularity of the actual impact effect. The impact amplification coefficient is introduced because the correlation impact adjustment factor itself has already undergone normalization, and its value is relatively small; appropriate amplification is needed to reflect the actual intensity of its impact on the target mold's anomaly level.

[0077] The system adds the first anomaly coefficient to the calculated adjustment increment value to generate the target anomaly coefficient for the target mold. The specific calculation formula is: Target Anomaly Coefficient = First Anomaly Coefficient + Adjustment Increment Value, where the adjustment increment value = Correlation Influence Adjustment Factor × Influence Amplification Coefficient. This additive calculation method ensures that the target mold's own anomaly state always serves as the basis, while correlation influences are only adjusted incrementally on this basis, reflecting the dominant role of the target mold's state and the auxiliary role of correlation influences. To prevent the adjusted target anomaly coefficient from exceeding a reasonable range, the system sets an upper limit constraint mechanism. When the calculation result exceeds 1, the target anomaly coefficient is automatically set to 1, ensuring the validity and interpretability of the value.

[0078] When the correlation influence adjustment factor is less than or equal to the adjustment threshold, it indicates that the influence of the correlated mold on the target mold is relatively minor and insufficient to significantly affect the anomaly assessment of the target mold. In this case, the system directly uses the first anomaly coefficient as the target anomaly coefficient of the target mold. This approach avoids oversensitivity to minor correlation influences, maintains the stability and reliability of anomaly assessment, simplifies the calculation process, and improves the system's operating efficiency. Through this hierarchical processing strategy, the system can ensure the completeness of correlation influence consideration while avoiding unnecessary increases in noise interference and computational complexity.

[0079] S105, when the target anomaly coefficient is greater than the preset threshold, generate anomaly information of the target mold, and generate maintenance suggestions for the target mold based on the anomaly information.

[0080] After confirming that the target anomaly coefficient exceeds a preset threshold, the system begins generating anomaly information for the target mold. This anomaly information provides a comprehensive description and analysis of the current abnormal state of the target mold, including key elements such as the time of occurrence, severity level, anomaly characteristic description, and potential influencing factors. The system determines the anomaly severity level based on the specific value of the target anomaly coefficient: 0.7 to 0.8 is classified as mild anomaly, 0.8 to 0.9 as moderate anomaly, and 0.9 to 1.0 as severe anomaly. Different levels correspond to different urgency levels and processing priorities. The anomaly characteristic description identifies specific anomaly manifestations by analyzing the change patterns of the first-state data. For example, a continuous decrease in height indicates increased wear, increased height fluctuations indicate decreased accuracy, and a sudden jump in height indicates a possible mechanical failure.

[0081] The system also identifies potential influencing factors of anomalies through backtracking analysis. Combining the second anomaly coefficient and correlation coefficient of related molds, it identifies whether there are upstream and downstream transmission factors. If a related mold has a high second anomaly coefficient and a large correlation coefficient with the target mold, the system identifies it as a major influencing factor, clearly indicating the specific source and transmission path of the related impact in the anomaly information. Simultaneously, the system analyzes the historical state change trends of the target mold itself to determine whether there are different anomaly patterns such as progressive wear, sudden failure, or periodic anomalies, providing important basis for subsequent maintenance strategy formulation.

[0082] Based on the generated anomaly information, the system further generates maintenance suggestions for the target mold. These suggestions are specific maintenance guidance plans provided by the system based on the anomaly type, severity, historical maintenance records, and best practice experience. The generation of maintenance suggestions employs a combination of expert systems and machine learning. The system has a built-in rich fault diagnosis knowledge base and maintenance experience database. Through pattern matching and similar case retrieval, it recommends the most suitable maintenance strategy for the current anomaly. For minor anomalies, the system may suggest preventative maintenance, such as cleaning, lubrication, and tightening checks. For moderate anomalies, the system may suggest localized repairs, such as edge sharpening and surface treatment. For severe anomalies, the system may suggest immediate shutdown for inspection or replacement of critical components.

[0083] The maintenance recommendations also include detailed operating instructions and resource requirements, such as the necessary maintenance tools, spare parts list, estimated maintenance time, and skill requirements, helping maintenance personnel to be fully prepared. Based on the target mold's model and specifications and historical maintenance records, the system provides personalized maintenance guidance, including key checkpoints, common failure modes, and maintenance precautions. Simultaneously, the system assesses the urgency and impact of the maintenance, providing decision support for production scheduling. For example, it may recommend performing maintenance after the current production task is completed to minimize production impact, or suggest immediate shutdown for maintenance to prevent more serious equipment damage.

[0084] Based on the above embodiments, the method further includes S106-S109:

[0085] S106, when the target anomaly coefficient is not greater than the preset threshold, acquire the first historical state data of the target mold and the second historical state data of the associated mold.

[0086] When the target anomaly coefficient is no greater than a preset threshold, it indicates that the target mold is currently in a relatively normal working state. However, this does not mean that its monitoring and management can be ignored, because mold anomalies are often a gradual process, evolving from initial minor changes to a significant abnormal state. If only static judgments based on the current state are relied upon, the optimal opportunity for preventative maintenance may be missed. Traditional passive maintenance methods often wait until obvious equipment failures occur before repairs are carried out. This approach not only causes unplanned production interruptions but may also lead to more serious equipment damage and higher repair costs. Therefore, the system needs to analyze historical data trends to achieve true predictive maintenance and identify potential anomaly trends in advance, even before the target mold reaches the anomaly threshold.

[0087] The system initiates a historical data backtracking analysis process to acquire the first historical state data of the target mold and the second historical state data of each associated mold. The time span of the historical data is typically set to the most recent 3 months or the most recent 500 production cycles to ensure sufficient data volume to support the statistical requirements of trend analysis, while also ensuring the timeliness of the data reflects the current production conditions and equipment status. The first historical state data contains all height change records of the target mold within the historical time period, arranged chronologically. Each data point includes a precise timestamp and corresponding measurement value, providing a foundation for subsequent time-series analysis. The second historical state data contains the state change records of each associated mold within the same time period. The system ensures the time synchronization of all historical data to avoid interference from time deviations in the correlation analysis.

[0088] S107, Based on the first historical state data, calculate the first historical anomaly coefficient change sequence of the target mold; based on the second historical state data, calculate the second historical anomaly coefficient change sequence of the associated mold.

[0089] Based on the acquired first historical state data, the system uses a sliding time window method to calculate the first historical anomaly coefficient change sequence of the target mold. The sliding time window refers to the system moving the calculation window across the historical data at fixed time intervals, calculating the corresponding anomaly coefficient for each time window. The size of the time window is typically set to the amount of data from one production shift or one day, and the step size is set to one-quarter of the window size to ensure sufficient overlap between adjacent time windows to capture continuous changes. The system calculates the anomaly coefficient for each time window using the same calculation method as the current target anomaly coefficient, ensuring comparability between historical and current anomaly coefficients, forming a first historical anomaly coefficient change sequence arranged chronologically.

[0090] The same processing method is applied to the second historical state data of each associated mold. The system calculates the second historical anomaly coefficient change sequence for each associated mold, reflecting the evolution of the associated mold's abnormal state over a historical period. Since different associated molds may have different operating cycles and maintenance periods, the system considers these differences during the calculation process, adjusting the time window size and calculation frequency accordingly to ensure accurate capture of the state change characteristics of each associated mold. The system also smooths the historical anomaly coefficient change sequence to eliminate data fluctuations caused by measurement noise or random factors, highlighting the true trend of change.

[0091] S108. Based on the correlation coefficient, the first historical anomaly coefficient change sequence and the second historical anomaly coefficient change sequence are correlated to determine the change pattern of the target anomaly coefficient.

[0092] Based on the established correlation coefficients, the system performs correlation analysis on the first historical anomaly coefficient change sequence and the second historical anomaly coefficient change sequences of each associated mold. The correlation analysis employs multivariate time series analysis methods, including advanced algorithms such as Granger causality test, vector autoregression model, and dynamic time warping. The system first identifies the time lag relationships between the sequences. Because the impact of changes in the state of associated molds on the target mold may have a certain time delay, the system determines the optimal lag time through cross-correlation analysis, typically between several hours and several production shifts. After determining the time lag relationships, the system calculates the influence weights of the historical anomaly coefficient changes of each associated mold on the historical anomaly coefficient changes of the target mold. These influence weights combine the correlation coefficients and actual historical correlations, more accurately reflecting the actual degree of influence of each associated mold on the target mold.

[0093] By comprehensively analyzing the first historical anomaly coefficient change sequence and the weighted second historical anomaly coefficient change sequence, the system identifies the changing patterns of the target anomaly coefficient. These patterns include different types such as periodic change patterns, trend change patterns, and abrupt change patterns. Periodic change patterns reflect regular fluctuations related to production cycle, maintenance cycle, or seasonal factors; trend change patterns reflect gradual changes such as mold wear and aging; and abrupt change patterns reflect sudden influencing factors such as equipment failure and process adjustments. The system uses frequency domain analysis to identify periodic components, trend decomposition to identify long-term trends, and anomaly detection algorithms to identify abrupt change points, thus forming a comprehensive understanding of the changing patterns of the target anomaly coefficient.

[0094] Based on the above embodiments, as an optional implementation, in S108, the correlation analysis of the first historical anomaly coefficient change sequence and the second historical anomaly coefficient change sequence based on the correlation coefficient to determine the change pattern of the target anomaly coefficient specifically includes S81-S84:

[0095] S81, calculate the slope and period of the first historical anomaly coefficient change sequence.

[0096] The system first conducts in-depth mathematical analysis of the historical anomaly coefficient change sequence, calculating the slope and period of the sequence. The slope reflects the rate and direction of change of the target mold's anomaly coefficient over time, and is a key indicator for judging the evolution trend of the mold's state. The system uses the least squares method to perform linear fitting on the historical anomaly coefficient change sequence, calculating the slope of the fitted line as the overall slope. Positive values ​​indicate a gradual increase in the degree of anomaly, while negative values ​​indicate a gradual decrease in the degree of anomaly; the magnitude of the value reflects the speed of change. To capture detailed features during the change process, the system also uses the moving average method to calculate the local slope. By comparing the local slopes at different time points, it identifies accelerations, decelerations, or turning points in the change process.

[0097] When calculating the variation period, the system uses frequency domain analysis to perform a Fourier transform on the first historical anomaly coefficient variation sequence to identify periodic components in the sequence. The variation period reflects the regularity of the target mold's anomaly coefficient changes, which is usually related to factors such as production cycle, maintenance cycle, and raw material batch changes. The system determines the dominant periodic frequency through power spectral density analysis and identifies multiple secondary periodic components, because mold state changes are often influenced by the combined effects of multiple periodic factors. For each identified periodic component, the system calculates its corresponding amplitude and phase, quantifying the contribution of the periodic component to the overall variation pattern, providing important periodic parameters for subsequent prediction models.

[0098] S82, the second historical anomaly coefficient change sequence and the correlation coefficient are weighted to obtain the correlation influence sequence.

[0099] Based on the second historical anomaly coefficient change sequence and the corresponding correlation coefficient of each associated mold, the system performs weighted processing to generate an associated influence sequence. The purpose of weighted processing is to reasonably combine the influence of different associated molds according to their actual importance. The system performs point-by-point multiplication of the second historical anomaly coefficient change sequence of each associated mold with its corresponding correlation coefficient to obtain the weighted influence sequence of that associated mold. The influence of molds with larger correlation coefficients is amplified accordingly, and the influence of molds with smaller correlation coefficients is reduced accordingly. Subsequently, the system arithmetically sums the weighted influence sequences of all associated molds to generate a comprehensive associated influence sequence, which reflects the change of the overall influence effect of all associated molds on the target mold over time.

[0100] S83, analyze the time delay correlation between the associated impact sequence and the first historical anomaly coefficient change sequence to determine the delay time of the associated impact.

[0101] To accurately describe the transmission characteristics of associated effects, the system analyzes the time-delay correlation between the associated effect sequence and the first historical anomaly coefficient change sequence, determining the delay time of the associated effects. The delay time refers to the time interval required for a change in the state of the associated mold to affect the target mold. The system employs cross-correlation function analysis to calculate the correlation coefficient between the two sequences under different time delays, and determines the optimal delay time by finding the peak position of the correlation coefficient. Considering that multiple transmission paths and different transmission mechanisms may exist in actual production, the system also identifies multiple local peaks corresponding to different delay times, reflecting complex multi-path transmission effects.

[0102] Based on the determined delay time, the system performs time offset processing on the correlated impact sequence to achieve optimal temporal alignment with the first historical anomaly coefficient change sequence. This processing eliminates the bias caused by the time delay in the correlation analysis and improves the accuracy of impact relationship identification. The system calculates the dynamic correlation between the two time-aligned sequences and identifies the pattern of correlation change over time, as the strength of the correlated impact may dynamically adjust with changes in production conditions, equipment status, and other factors. By analyzing the changing patterns of the dynamic correlation, the system can identify the stability and predictability of impact transmission.

[0103] S84, combining the slope of change, the period of change, and the delay time, determines the variation pattern of the target anomaly coefficient.

[0104] By combining key parameters such as the calculated slope, cycle, and delay time, the system constructs a comprehensive mathematical model of the changing patterns of the target anomaly coefficient. This model comprises three main components: a trend term, a cycle term, and a correlation influence term. The trend term describes the long-term development trend of the target anomaly coefficient based on the slope, the cycle term describes the regular fluctuation pattern based on the identified cycle, and the correlation influence term describes the transmission effect of external influences based on the delay time and the sequence of correlation influences. The system uses multinomial regression or machine learning algorithms to optimize and combine these three components, determine the weight coefficients and interaction relationships of each part, and form a complete mathematical expression that can accurately describe the historical changing patterns of the target anomaly coefficient.

[0105] S109, based on the changing pattern, predict the changing trend of the target anomaly coefficient within a preset time period in the future; based on the changing trend, generate maintenance suggestions for the target mold.

[0106] Based on the identified patterns of change, the system establishes a predictive model for the target anomaly coefficient. This model employs machine learning algorithms, such as Long Short-Term Memory networks, Support Vector Regression, or time series forecasting algorithms, combining historical data patterns with current status information to predict the trend of the target anomaly coefficient over a predetermined future time period. This predetermined time period is typically set to 1-4 weeks, providing sufficient preparation time for maintenance plans while ensuring the reliability of the prediction results. The prediction results include key information such as the expected value of the target anomaly coefficient, the rate of change, and the likely time when it will reach the anomaly threshold. The system also provides a prediction confidence interval to quantify the uncertainty of the prediction results.

[0107] Based on the predicted trends, the system generates maintenance recommendations for the target mold. These recommendations differ in nature from the repair recommendations in S105; the former focuses on preventative maintenance, while the latter emphasizes fault repair. When the prediction indicates that the target anomaly coefficient may exceed a preset threshold at some future point in time, the system recommends scheduling preventative maintenance before that point, such as early cleaning, precision adjustments, or replacement of some components, to prevent further development of the abnormal state. The maintenance recommendations also include recommended maintenance time windows. The system will combine production plans and equipment load conditions to recommend the most suitable maintenance time, ensuring maintenance effectiveness while minimizing the impact on production. For situations where the predicted anomaly coefficient is developing slowly, the system may suggest extending the monitoring frequency or adjusting the maintenance cycle to optimize the allocation of maintenance resources.

[0108] The application also integrates a complete data acquisition and communication support module, which establishes a real-time data connection with the height sensor on the stamping equipment through serial communication technology to obtain the raw measurement data of the mold height, providing a reliable data source foundation for the entire anomaly detection system.

[0109] The system has an intelligent data quality control mechanism that automatically performs format verification and validity checks on the received raw data. It can accurately identify and extract valid data containing device numbers and height values, while automatically filtering out invalid data with format errors, abnormal values, or damaged transmission, ensuring the data quality for subsequent analysis and processing.

[0110] To ensure data security and reliability, the system adopts a dual storage strategy, simultaneously writing valid data that has passed verification to a local CSV file and a remote database. This ensures both real-time data availability and data backup protection, effectively preventing the risk of data loss due to a single storage failure.

[0111] The system also has a robust communication stability guarantee mechanism. It monitors the status of the serial communication receive buffer in real time through a timer. When the data in the buffer exceeds a preset threshold, it automatically performs a clearing operation to prevent buffer overflow from affecting data reception. At the same time, it monitors the time interval for receiving valid data. When a communication interruption or abnormal data reception is detected, it can automatically execute a restart process to restore normal data acquisition function.

[0112] In addition, the system is equipped with an automatic reconnection mechanism. When serial communication is unexpectedly disconnected, the reconnection process can be automatically started. The system will continuously attempt to re-establish the connection at preset time intervals to ensure that normal data acquisition and storage operations can continue immediately after communication is restored, thus maximizing the continuity of data acquisition and the stability of system operation.

[0113] like Figure 2 As shown, Figure 2 This diagram illustrates a mold status data acquisition and anomaly analysis system provided in this application embodiment. The diagram systematically demonstrates the complete logical loop of a mold status data acquisition and anomaly analysis method. The overall architecture is divided into a physical device layer, a data transmission layer, and a data processing layer from bottom to top. In the bottom physical device layer, the "target mold" on the left is equipped with a height sensor to collect first-state data in real time. The "related mold" on the right, as a device with upstream and downstream relationships in the process flow, synchronously collects its operating parameters, i.e., second-state data. The two are clearly linked by a dotted line, indicating their process correlation and correlation coefficient. The two sets of data are converged to the top "data processing center" via the middle data transmission channel. The server first calculates the basic "first anomaly coefficient" of the target mold based on the first-state data. Then, it introduces crucial correction logic, using the "second anomaly coefficient" and "correlation coefficient" of the related mold to weight or adjust the preliminary results, thereby generating a "target anomaly coefficient" that better reflects the actual working conditions. Finally, the system compares the coefficient with a preset threshold. If the threshold is exceeded, as shown in the alarm module on the right, it outputs abnormal information and automatically generates corresponding maintenance suggestions, thus completing the entire process from the acquisition of raw sensor data to intelligent maintenance decision-making.

[0114] Based on the above method, this application also discloses a mold state data acquisition system, such as... Figure 3 As shown, Figure 3 This is a schematic diagram of a mold status data acquisition system provided in an embodiment of this application. The system includes: a first acquisition module, a second acquisition module, a third acquisition module, a combination module, and a generation module; wherein,

[0115] The first acquisition module is used to acquire the first state data of the target mold sent by the height sensor installed on the stamping equipment, and generate a first anomaly coefficient of the target mold based on the first state data; the second acquisition module is used to acquire the correlation coefficient between the target mold and related molds, where related molds are molds that have an upstream or downstream relationship with the target mold in the process flow; the third acquisition module is used to acquire the second state data of the related molds, and generate a second anomaly coefficient of the related molds based on the second state data; the combination module is used to combine the correlation coefficient and the second anomaly coefficient to adjust the first anomaly index and generate a target anomaly coefficient of the target mold; the generation module is used to generate anomaly information of the target mold when the target anomaly coefficient is greater than a preset threshold, and generate maintenance suggestions for the target mold based on the anomaly information.

[0116] It should be noted that the system provided in the above embodiments is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0117] Please see Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 1000 may include: at least one processor 1001, at least one network interface 1004, a user interface 1003, a memory 1005, and at least one communication bus 1002.

[0118] The communication bus 1002 is used to realize the connection and communication between these components.

[0119] The user interface 1003 may include a display screen and a camera. Optionally, the user interface 1003 may also include a standard wired interface and a wireless interface.

[0120] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0121] The processor 1001 may include one or more processing cores. The processor 1001 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 1005, and by calling data stored in the memory 1005. Optionally, the processor 1001 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 1001 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content to be displayed on the screen; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 1001 and may be implemented as a separate chip.

[0122] The memory 1005 may include random access memory (RAM) or read-only memory. Optionally, the memory 1005 may include a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 1005 may also be at least one storage device located remotely from the aforementioned processor 1001. Figure 4 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for a mold status data acquisition method.

[0123] exist Figure 4In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and to acquire the user input data; while the processor 1001 can be used to call an application program stored in the memory 1005 for a mold state data acquisition method. When executed by one or more processors, the electronic device performs one or more of the methods described in the above embodiments.

[0124] An electronic device readable storage medium stores instructions that, when executed by one or more processors, cause the electronic device to perform one or more of the methods described in the above embodiments.

[0125] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0126] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0127] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings or direct couplings or communication connections may be through some service interfaces; indirect couplings or communication connections between devices or units may be electrical or other forms.

[0128] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0129] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0130] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0131] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Other embodiments of this disclosure will be readily apparent to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method of collecting mold state data, characterized by, The method comprises: obtaining first state data of a target die sent by a height sensor installed on a stamping device, and generating a first abnormality coefficient of the target die according to the first state data; obtaining a correlation coefficient of the target die and associated dies, comprising: identifying associated dies which are in adjacent or indirectly connected positions in a process flow with the target die according to process flow configuration information of a stamping production line, and determining a process transmission path between each of the associated dies and the target die; calculating a process distance parameter between the target die and each of the associated dies according to the process transmission path, the process distance parameter representing a position interval degree between dies in the process flow; obtaining a historical state data change record of the target die and each of the associated dies within a preset time period, and calculating an influence correlation coefficient of a state change of each of the associated dies on a state change of the target die; inputting the process distance parameter and the influence correlation coefficient into a correlation coefficient calculation function to generate the correlation coefficient of the target die and each of the associated dies, the correlation coefficient being used to quantify the influence degree of the associated dies on the target die; the correlation coefficient calculation function is Cij = ρij × e^(-λ × dij) × αij, wherein Cij represents the correlation coefficient, ρij represents the influence correlation coefficient of the associated die i on the target die j, dij represents the process distance parameter between the associated die i and the target die j, λ is a distance attenuation factor, and αij is a process direction adjustment coefficient; the associated dies are dies which have an upstream-downstream relationship with the target die in the process flow; obtaining second state data of the associated dies, and generating a second abnormality coefficient of the associated dies according to the second state data; adjusting the first abnormality index in combination with the correlation coefficient and the second abnormality coefficient to generate a target abnormality coefficient of the target die; when the target abnormality coefficient is greater than a preset threshold, generating abnormal information of the target die, and generating a maintenance suggestion for the target die according to the abnormal information.

2. The mold state data acquisition method according to claim 1, characterized by, The method further comprises: when the target abnormality coefficient is not greater than the preset threshold, obtaining first historical state data of the target die and second historical state data of the associated dies; calculating a first historical abnormality coefficient change sequence of the target die according to the first historical state data, and calculating a second historical abnormality coefficient change sequence of the associated dies according to the second historical state data; based on the correlation coefficient, performing correlation analysis on the first historical abnormality coefficient change sequence and the second historical abnormality coefficient change sequence to determine a change law of the target abnormality coefficient; according to the change law, predicting a change trend of the target abnormality coefficient within a future preset time period; according to the change trend, generating a maintenance suggestion for the target die.

3. The mold state data acquisition method according to claim 2, characterized by, The correlation analysis on the first historical abnormality coefficient change sequence and the second historical abnormality coefficient change sequence to determine the change law of the target abnormality coefficient based on the correlation coefficient comprises: Calculate a change slope and a change period of the first historical abnormal coefficient change sequence; Weight the second historical abnormal coefficient change sequence and the correlation coefficient to obtain a correlation influence sequence; Analyze time delay correlation between the correlation influence sequence and the first historical abnormal coefficient change sequence to determine a delay time of correlation influence; Determine a change rule of the target abnormal coefficient in combination with the change slope, the change period and the delay time.

4. The mold state data acquisition method according to claim 1, characterized by, The first abnormal coefficient of the target mold is generated according to the first state data, including: Obtaining reference state data of the target mold, the reference state data being a standard height value of the target mold in a normal working state; Calculating a deviation value between the first state data and the reference state data, and determining a deviation level according to the deviation value; Generating the first abnormal coefficient of the target mold according to the deviation level and a preset abnormal coefficient calculation rule.

5. The mold state data acquisition method according to claim 1, characterized by, The target abnormal coefficient of the target mold is generated by adjusting the first abnormal index in combination with the correlation coefficient and the second abnormal coefficient, including: Arithmetic multiplication of the correlation coefficient and the second abnormal coefficient of each correlation mold generates an influence coefficient of each correlation mold, and the influence weight value of each correlation mold is determined according to the influence coefficient, the influence coefficient being directly proportional to the influence weight value; Weighted summation of the influence weight values of all correlation molds generates a correlation influence adjustment factor; Combination operation of the first abnormal coefficient and the correlation influence adjustment factor generates the target abnormal coefficient of the target mold.

6. The mold state data acquisition method according to claim 5, wherein The target abnormal coefficient of the target mold is generated by combination operation of the first abnormal coefficient and the correlation influence adjustment factor, including: Judging whether the correlation influence adjustment factor is greater than a preset adjustment threshold value; When the correlation influence adjustment factor is greater than the adjustment threshold value, multiplying the correlation influence adjustment factor by a preset influence amplification coefficient to obtain an adjustment increment value, and performing addition operation of the first abnormal coefficient and the adjustment increment value to generate the target abnormal coefficient of the target mold; When the correlation influence adjustment factor is less than or equal to the adjustment threshold value, the first abnormal coefficient is taken as the target abnormal coefficient of the target mold.

7. A mold condition data acquisition system characterized by comprising: The system includes a first acquisition module, a second acquisition module, a third acquisition module, a combination module and a generation module, wherein The first acquisition module is configured to acquire first state data of a target mold sent by a height sensor installed on a stamping device, and generate a first abnormal coefficient of the target mold according to the first state data. The second obtaining module is configured to obtain the correlation index of the target mold and the associated molds, including: identifying the associated molds which are in adjacent or indirectly connected positions with the target mold in a process flow according to process flow configuration information of a stamping production line, and determining a process transmission path between each of the associated molds and the target mold; calculating a process distance parameter between the target mold and each of the associated molds according to the process transmission path, the process distance parameter representing a position interval degree between molds in the process flow; obtaining a historical state data change record of the target mold and each of the associated molds in a preset time period, and calculating an influence correlation coefficient of a state change of each of the associated molds on a state change of the target mold; inputting the process distance parameter and the influence correlation coefficient into a correlation index calculation function to generate the correlation index of the target mold and each of the associated molds, the correlation index being used to quantify the influence degree of the associated molds on the target mold; the correlation index calculation function is: Cij=ρij×e^(-λ×dij)×αij, where Cij represents the correlation index, ρij represents the influence correlation coefficient of the associated mold i on the target mold j, dij represents the process distance parameter between the associated mold i and the target mold j, λ is a distance attenuation factor, and αij is a process direction adjustment coefficient; the associated mold is a mold which has an upstream and downstream relationship with the target mold in the process flow; The third obtaining module is configured to obtain second state data of the associated molds, and generate a second abnormality coefficient of the associated molds according to the second state data. The combination module is configured to combine the correlation index and the second abnormality coefficient to adjust the first abnormality index, and generate a target abnormality coefficient of the target mold. The generation module is configured to generate abnormality information of the target mold when the target abnormality coefficient is greater than a preset threshold, and generate a maintenance suggestion of the target mold according to the abnormality information.

8. An electronic device, comprising: An electronic device includes a processor, a memory, a user interface, and a network interface. The memory is configured to store instructions. The user interface and the network interface are configured to communicate with other devices. The processor is configured to execute the instructions stored in the memory to cause the electronic device to perform the method of any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, A computer program is stored in a memory and can be loaded and executed by a processor to perform the method of any one of claims 1-6.

Images