A temperature monitoring method and system applied to a bottle preform injection mold

Abstract

The application provides a temperature monitoring method and system applied to a bottle embryo injection mold, and belongs to the technical field of intelligent monitoring of injection molds. The method comprises the following steps: detecting the electromagnetic interference intensity and frequency range by using an electromagnetic environment detection device, and identifying the interference source and interference frequency band; obtaining temperature data, performing photoelectric isolation and magnetic isolation processing on the temperature data to obtain initial temperature data; performing filtering processing on the initial temperature data based on the interference source and interference frequency band, and performing compensation correction on the filtered data based on the thermal inertia of the bottle embryo injection mold to obtain target temperature data; comparing the target temperature data with a preset standard temperature curve, and calculating a temperature deviation value and a deviation change rate; if the temperature deviation value exceeds a first interval and the duration is greater than a preset first time window, a temperature anomaly early warning signal is generated, and a target abnormal trend type is judged; and a temperature monitoring report is generated based on the temperature anomaly early warning signal and the target abnormal trend type.

Description

Technical Field

[0001] This application relates to the field of intelligent monitoring technology for injection molds, and in particular to a temperature monitoring method and system for preform injection molds. Background Technology

[0002] The temperature stability of preform injection molds determines the preform molding accuracy, wall thickness uniformity, and production yield, serving as a key control indicator in injection molding production. However, injection molding workshops contain various equipment such as injection molding machines, robotic arms, and mold temperature controllers. The operation of these different devices, individually or in combination, generates complex electromagnetic interference that can easily intrude into the temperature sensing signal transmission link, leading to distorted temperature data. Furthermore, traditional temperature acquisition relies on a single isolation method, lacking targeted anti-interference design and failing to effectively filter out interference in specific frequency bands. In addition, preform injection molds exhibit significant thermal inertia, causing temperature measurement data to be prone to lag. Traditional monitoring methods do not compensate for this lag based on the mold's thermal characteristics, resulting in insufficient temperature monitoring accuracy. Existing technologies often fail to accurately identify temperature anomalies due to data distortion and lag issues, easily leading to false or missed warnings and affecting production continuity.

[0003] Therefore, there is an urgent need for a temperature monitoring method and system for preform injection molds to meet the precise control requirements of modern injection molding workshops. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides a temperature monitoring method and system for preform injection molds.

[0005] A first aspect of this application provides a temperature monitoring method applied to a preform injection mold, comprising: Electromagnetic environment detection equipment was used to measure the electromagnetic interference intensity and frequency range of each piece of equipment in the injection molding workshop under different operating conditions, and the interference source and its corresponding interference frequency band were identified. Temperature data is acquired and transmitted to the data acquisition circuit after photoelectric and magnetic isolation to obtain initial temperature data. Based on the interference source and its corresponding interference frequency band, the initial temperature data is filtered, and the filtered data is compensated and corrected based on the thermal inertia of the preform injection mold to obtain the target temperature data. The target temperature data is compared with a preset standard temperature curve, and the temperature deviation value and the deviation change rate are calculated. If the temperature deviation value exceeds the first interval and the duration is greater than the preset first time window, a temperature anomaly warning signal is generated, and the temperature anomaly trend type is determined based on the deviation change rate to obtain the target anomaly trend type. Based on the temperature anomaly warning signal and the target anomaly trend type, a temperature monitoring report is generated and uploaded to the centralized monitoring platform of the injection molding workshop.

[0006] A second aspect of this application provides a temperature monitoring system for use in preform injection molds, comprising: The electromagnetic interference detection and identification module is used to measure the electromagnetic interference intensity and frequency range of each piece of equipment in the injection molding workshop under different operating conditions using electromagnetic environment detection equipment, and to identify the interference source and its corresponding interference frequency band. A temperature data isolation acquisition module is used to acquire temperature data and transmit the temperature data to the data acquisition circuit after photoelectric isolation and magnetic isolation to obtain initial temperature data. The data filtering compensation and correction module is used to filter the initial temperature data based on the interference source and its corresponding interference frequency band, and to compensate and correct the filtered data based on the thermal inertia of the preform injection mold to obtain the target temperature data. The temperature anomaly detection and early warning module is used to compare the target temperature data with a preset standard temperature curve and calculate the temperature deviation value and the deviation change rate; if the temperature deviation value exceeds the first interval and the duration is greater than the preset first time window, a temperature anomaly early warning signal is generated, and the temperature anomaly trend type is determined according to the deviation change rate to obtain the target anomaly trend type. The monitoring report generation and upload module is used to generate a temperature monitoring report based on the temperature anomaly warning signal and the target anomaly trend type, and upload it to the centralized monitoring platform of the injection molding workshop.

[0007] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the temperature monitoring method applied to a preform injection mold described above.

[0008] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the temperature monitoring method applied to a preform injection mold described above.

[0009] The beneficial effects of the temperature monitoring method and system for preform injection molds provided in this application are as follows: This application identifies interference sources and corresponding frequency bands of various equipment in the workshop under different operating conditions through pre-electromagnetic interference detection and targeted anti-interference design. Based on photoelectric and magnetic dual isolation and directional filtering, the influence of electromagnetic interference on the temperature signal is effectively filtered out, further improving the purity of the initial temperature data. Simultaneously, based on the thermal inertia characteristics of the preform injection mold, the filtered data is compensated and corrected, improving the monitoring accuracy and timeliness of the target temperature data and reducing misjudgments caused by lag deviations. Furthermore, by comparing the target temperature data with a standard curve and realizing anomaly judgment and trend identification based on the deviation value and rate of change, and triggering an alarm only when the standard is exceeded during unplanned periods, the probability of false alarms and missed alarms is reduced, providing a reliable basis for mold temperature control. The final monitoring report is uploaded to the centralized monitoring platform in real time, helping staff to promptly grasp the mold temperature status, quickly respond to anomalies, ensure production continuity, and improve the preform molding accuracy, wall thickness uniformity, and production qualification rate. Attached Figure Description

[0010] Figure 1 A schematic flowchart illustrating a temperature monitoring method for preform injection molds provided in an embodiment of this application; Figure 2 This is a structural block diagram of a temperature monitoring system for preform injection molds provided in one embodiment of this application; Figure 3 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0011] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0012] To make the purpose, technical solution, and advantages of this application clearer, the following will be described in conjunction with the appendix. Figure 1 - Appendix Figure 3 The following is an explanation using specific examples.

[0013] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a temperature monitoring method for preform injection molds provided in an embodiment of this application. The method includes: S101: Using electromagnetic environment detection equipment, measure the electromagnetic interference intensity and frequency range of each piece of equipment in the injection molding workshop under different operating conditions, and identify the interference source and its corresponding interference frequency band.

[0014] In this embodiment, according to a preset measurement cycle, each piece of equipment in the injection molding workshop is controlled to operate individually and in combination sequentially, resulting in multiple operating states. In each operating state, multiple electromagnetic environment detection nodes are arranged in the injection molding workshop using a grid-based layout, and electromagnetic interference signals from each node are collected. Time-frequency analysis is performed on the electromagnetic interference signals from each node, and the electromagnetic interference intensity and frequency range of different equipment in each operating state are extracted. High-interference areas are located based on the electromagnetic interference intensity distribution heatmap, and interference sources are identified based on the step response discrimination method for startable and stopable equipment and the transfer entropy causal analysis of equipment that cannot be started or stopped at any time. The spectral features of the identified interference sources are extracted, and the center frequency and frequency bandwidth corresponding to each interference source are obtained through frequency domain peak search and bandwidth analysis, which are then summarized to form interference frequency bands.

[0015] S102: Acquire temperature data and transmit the temperature data to the data acquisition circuit after photoelectric isolation and magnetic isolation to obtain initial temperature data.

[0016] In this embodiment, acquiring temperature data is a fundamental step in ensuring the accuracy of temperature monitoring during the preform injection mold process. Therefore, a scientifically designed data acquisition scheme must be deployed, taking into account the mold's structural characteristics and the injection molding production conditions. Specifically, temperature sensors are installed in temperature-sensitive areas such as the mold cavity, core, and runner. Simultaneously, an appropriate data acquisition frequency is set based on process parameters such as the injection cycle and the mold's heating rate.

[0017] The temperature data acquired at this time is subject to interference from the complex electromagnetic environment of the injection molding workshop. The operation of equipment such as injection molding machines, robotic arms, and mold temperature controllers generates strong electromagnetic radiation. Direct transmission of temperature signals can easily lead to signal distortion and noise intrusion. Therefore, dual processing of photoelectric and magnetic isolation is required for the temperature data. A photoelectric isolation module converts the electrical transmission of the temperature signal into optical transmission, utilizing the insulating properties of optical signals to cut off the transmission path of electromagnetic interference. Simultaneously, magnetic isolation further suppresses electromagnetic coupling interference.

[0018] The temperature signal, after being photoelectrically and magnetically isolated, will be transmitted to a dedicated data acquisition circuit. This circuit receives the isolated analog temperature signal and converts it into a digital signal via a built-in analog-to-digital converter. During the conversion process, the data acquisition circuit performs preliminary amplitude calibration and timestamp marking on the signal to eliminate sensor errors and minor offsets during transmission, ultimately yielding the initial temperature data.

[0019] S103: Based on the interference source and its corresponding interference frequency band, the initial temperature data is filtered, and the filtered data is compensated and corrected based on the thermal inertia of the preform injection mold to obtain the target temperature data.

[0020] In this embodiment, the initial temperature data is decomposed into wavelet packets using a preset second time window to obtain wavelet packet coefficients corresponding to multiple frequency bands. The target frequency band affected by the interference is determined based on the interference frequency band corresponding to the interference source. The coupling interference intensity within the target frequency band is estimated based on the preset coupling transfer function from each interference source to the temperature signal transmission line and the operating state of each interference source. A masking threshold is constructed based on the coupling interference intensity. Wavelet packet coefficients with moduli less than the masking threshold are zeroed, and wavelet packet coefficients with moduli greater than the masking threshold are subjected to soft threshold shrinkage processing to obtain the target wavelet packet coefficients. The target wavelet packet coefficients are reconstructed using inverse wavelet packets, and the reconstructed value at the end of the second time window is taken as the filtered temperature data. Based on the thermal inertia constant of the preform injection mold, hysteresis compensation is performed on the filtered temperature data to obtain the target temperature data.

[0021] S104: Compare the target temperature data with the preset standard temperature curve and calculate the temperature deviation value and the deviation change rate; if the temperature deviation value exceeds the first interval and the duration is greater than the preset first time window, generate a temperature anomaly warning signal, and determine the temperature anomaly trend type based on the deviation change rate to obtain the target anomaly trend type.

[0022] In this embodiment, the target temperature data is dynamically compared with a preset standard temperature curve. This process needs to be carried out based on the entire injection molding production cycle, so that the temperature data at each time point corresponds to the standard curve. The preset standard temperature curve is a reference data calibrated and stored based on the material characteristics of the preform (e.g., PET, PP, etc.), molding specifications, and injection molding process parameters (e.g., injection speed, holding pressure, cooling time, etc.). The standard temperature curve includes the ideal temperature range of areas such as the mold cavity, core, and runner at different production stages. During the comparison process, the difference between the target temperature data and the corresponding node of the standard temperature curve is calculated point by point to obtain the real-time temperature deviation value. At the same time, the change range of the deviation value within adjacent time periods is continuously calculated using a sliding window algorithm to generate the deviation change rate.

[0023] If the calculated temperature deviation exceeds a preset first interval and the duration of this temperature deviation exceeds a preset first time window, a temperature anomaly warning signal is triggered. Simultaneously with generating the warning signal, the temperature anomaly trend type is further determined based on the deviation change rate to obtain the target anomaly trend type. Specifically, when the deviation change rate is positive and its absolute value is greater than a preset first change rate threshold, it is determined to be a rapid heating anomaly; when the deviation change rate is negative and its absolute value is greater than the first change rate threshold, it is determined to be a rapid cooling anomaly; when the absolute value of the deviation change rate is less than or equal to the first change rate threshold, but the temperature deviation continuously exceeds the first interval and the cumulative duration is greater than a preset cumulative duration threshold, it is determined to be a slow deviation anomaly; when the deviation change rate alternates between positive and negative, and the deviation value exceeds the first interval each time it alternates, while the fluctuation frequency exceeds a set fluctuation threshold, it is determined to be a fluctuation anomaly; when the deviation change rate gradually approaches zero, but the temperature deviation value remains outside the first interval, and the standard deviation of the temperature distribution in key areas of the mold (cavity corners, runner intersections) is greater than a preset uniformity threshold, it is determined to be a local temperature difference anomaly.

[0024] The standard temperature curve for injection molding of preforms of corresponding materials and specifications is used as the reference temperature for each node. This is combined with measured temperature data from qualified mass-produced products and mold temperature control process tolerances. Specifically, the normal deviation range between the actual temperature and the standard temperature at each time point within the entire normal production cycle is first statistically analyzed. Then, upper and lower limits are defined based on preform molding quality thresholds (no shrinkage marks, no warping, uniform wall thickness, and smooth demolding). The maximum allowable positive and negative deviation range under qualified production conditions is defined as the first interval. Exceeding this first interval indicates a temperature deviation from the qualified process range. The preset first time window is determined based on the duration of each stage of the complete preform injection molding process, including cooling, holding pressure, and injection, as well as the mold's thermal conduction response speed and the temperature control response delay of the mold temperature controller. Specifically, shorter process cycles result in a shorter first time window, while molds with large wall thickness and large volume have higher thermal inertia, resulting in a longer first time window. Simultaneously, based on historical abnormal data from the field, instantaneous temperature fluctuations are eliminated, and the first time window is selected to filter out occasional temperature jumps and promptly capture persistent temperature anomalies. The preset first rate of change threshold is based on the natural fluctuation rate of temperature deviation throughout the entire process cycle under normal production conditions, serving as a benchmark sample. The maximum extreme value of the deviation change rate under normal operating conditions is statistically analyzed. A safety margin is reserved based on this maximum extreme value and set upwards. Simultaneously, positive and negative thresholds are set separately for heating and cooling, matching the normal temperature change rates of different process stages in injection molding heating and cooling. Exceeding this first rate of change threshold indicates that the temperature rise / fall rate exceeds the reasonable range of the process. The preset cumulative duration threshold is set based on the preform molding temperature control duration and the long-term steady-state operating parameters of the equipment. It represents the longest duration during which product quality can be maintained even under slow temperature deviation. The safety margin is adjusted downwards based on this longest duration as a reference, with the cumulative duration threshold intentionally narrowed to proactively provide a buffer safety range based on the allowable limit of the critical deviation value during normal production. The fluctuation threshold is determined by using the frequency of natural fluctuations in temperature deviation during normal production as a base value. It is then further assessed by considering the frequency of normal temperature disturbances caused by equipment start-up and shutdown, water circuit opening and closing, and oil temperature fluctuations in the injection molding workshop, along with environmental disturbances in the production environment. For example, if the number of alternating fluctuations in the first interval exceeds the fluctuation threshold within a unit of time, it is considered a non-process-normal, disorderly temperature fluctuation. The uniformity threshold is set based on the standard deviation of the measured temperature distribution in areas with temperature differences, such as mold cavity corners and runner intersections, during the production of qualified preforms. It is then set by adding the inherent temperature difference margin of the structure to the base value, taking into account the preform wall thickness consistency requirements and differences in mold structure heat dissipation. The uniformity threshold is used to define whether the temperature difference in local areas of the mold exceeds the allowable molding range.

[0025] S105: Based on the temperature anomaly warning signal and the target anomaly trend type, generate a temperature monitoring report and upload it to the centralized monitoring platform of the injection molding workshop.

[0026] In this embodiment, the temperature monitoring report is a multi-dimensional comprehensive document formed based on the process background, equipment operating status, and abnormal impact analysis of the entire injection molding production process. The temperature monitoring report first determines the trigger time, trigger node (e.g., mold cavity, core, or runner area), peak temperature deviation value, and duration of the marked temperature abnormality warning signal. At the same time, it records the criteria for determining the type of target abnormal trend, including the specific value of the deviation change rate and the threshold comparison results.

[0027] In addition, the report will supplement key related information: on the one hand, it integrates production condition data during the period of anomaly occurrence, including the injection molding machine operating parameters (injection speed, holding pressure, etc.), mold temperature controller setpoints and actual output values, equipment start-up and shutdown records, and process switching status; on the other hand, based on the typical fault cause library corresponding to the target anomaly trend type, it lists the priority investigation items that lead to this type of anomaly. For example, for rapid temperature rise anomalies, it is recommended to prioritize checking the cooling water flow, temperature control system PID parameters, and heating module operating status. It also notes the preform molding defects that this type of anomaly might cause if it continues and its potential impact on production progress. Preform molding defects include shrinkage marks and burning. The typical fault cause library is constructed based on historical preform injection mold temperature anomaly trend type data, summarizing the fault causes, parameter anomaly characteristics, molding defect correlations, and corresponding investigation priorities for various temperature anomalies, and then classifying and organizing them. The method for constructing a typical fault cause database includes: based on historical data of abnormal temperatures in preform injection molds accumulated over a long period, firstly, classifying and collecting various types of abnormal temperature trends such as rapid heating, rapid cooling, slow deviation, fluctuation, and local temperature difference; then, systematically analyzing the abnormal characteristics of equipment operating parameters and actual causes of failures on site for each type of temperature anomaly; determining the correlation between different abnormal temperature phenomena and preform molding defects such as shrinkage marks and scorching; and based on on-site operation and maintenance experience, determining the order of repair and troubleshooting priorities for various faults; finally, integrating all related information for systematic classification, organization, and summarization to obtain the typical fault cause database.

[0028] After generating the temperature monitoring report, this embodiment uploads the report to the centralized monitoring platform of the injection molding workshop in a standardized format via Modbus TCP. During the upload process, an upload log is generated to record the report upload time, transmission status, and received feedback results. After the temperature monitoring report is uploaded to the centralized monitoring platform of the injection molding workshop, the platform parses and classifies the temperature monitoring report data, synchronizes the temperature anomaly warning signal, target anomaly trend type, and temperature monitoring report to the platform's visual interface, and notifies relevant management personnel through pop-up reminders, audible and visual alarms, etc. At the same time, management personnel can query report details, export data files, and issue anomaly handling instructions through the platform.

[0029] As can be seen from the above, this application, through pre-processing electromagnetic interference detection and targeted anti-interference design, identifies the interference sources and corresponding frequency bands of various equipment in the workshop under different operating conditions. Based on photoelectric and magnetic dual isolation and directional filtering, it effectively filters out the influence of electromagnetic interference on the temperature signal, further improving the purity of the initial temperature data. Simultaneously, based on the thermal inertia characteristics of the preform injection mold, the filtered data is compensated and corrected, improving the monitoring accuracy and timeliness of the target temperature data and reducing misjudgments caused by lag deviations. Furthermore, by comparing the target temperature data with the standard curve and identifying anomalies and trends based on deviation values ​​and rates of change, and triggering warnings only when the standard is exceeded during unplanned periods, the probability of false and missed warnings is reduced, providing a reliable basis for mold temperature control. The final monitoring report is uploaded to the centralized monitoring platform in real time, helping staff to promptly grasp the mold temperature status, quickly respond to anomalies, ensure production continuity, and improve the preform molding accuracy, wall thickness uniformity, and production qualification rate.

[0030] In one embodiment of this application, the equipment in the injection molding workshop includes an injection molding machine, a robotic arm, a mold temperature controller, and a dryer. Electromagnetic environment detection equipment was used to measure the intensity and frequency range of electromagnetic interference in the injection molding workshop under different operating conditions, and the interference sources and their corresponding interference frequency bands were identified, including: According to the preset measurement cycle, the equipment in the injection molding workshop is controlled to run individually and in combination in sequence to obtain multiple operating states; In each operating state, multiple electromagnetic environment detection nodes are arranged in the injection molding workshop using a grid-based layout, and electromagnetic interference signals of each node are collected. Time-frequency analysis was performed on the electromagnetic interference signals of each node, and the electromagnetic interference intensity and frequency range of different devices under each operating state were extracted. Based on the electromagnetic interference intensity and frequency range of different nodes under different operating conditions, a heat map of electromagnetic interference intensity distribution for each device under its operating condition is drawn. High-interference areas are located based on electromagnetic interference intensity distribution heatmaps, and interference sources are identified based on the step response discrimination method for startable and stopable equipment and the causal analysis of the transfer entropy of equipment that cannot be started or stopped at any time. The spectral features of the identified interference sources are extracted, and the center frequency and frequency bandwidth of each interference source are obtained through frequency domain peak search and bandwidth analysis. The results are then summarized to form the interference frequency band.

[0031] In this embodiment, the equipment in the injection molding workshop includes injection molding machines, robotic arms, mold temperature controllers, and dryers. During the production process, these devices generate complex electromagnetic radiation due to motor operation, circuit switching, high-frequency heating, and other operations, forming a superimposed electromagnetic environment that affects the transmission accuracy of temperature signals.

[0032] Electromagnetic environment detection equipment was used to measure the electromagnetic interference intensity and frequency range of equipment in the injection molding workshop under different operating conditions, and the interference sources and their corresponding interference frequency bands were identified. Specifically, firstly, according to a preset measurement cycle, the injection molding machine, robot arm, mold temperature controller, and dryer in the injection molding workshop were controlled to operate independently in sequence, while multiple sets of equipment joint operation modes were set (e.g., injection molding machine-mold temperature controller, injection molding machine-robot arm-dryer, etc.), forming multiple operating states including single equipment independent operation and multiple equipment collaborative operation. Based on the actual production and operation patterns of the injection molding workshop, the measurement cycle is set in a comprehensive manner, taking into account three factors: workshop production shifts, equipment daily maintenance cycles, and production process switching frequency. It also considers the needs of all equipment operating conditions. First, the duration of regular morning, noon, and evening production shifts, equipment downtime for periodic maintenance, and the switching intervals for different preform specifications and injection molding processes are statistically analyzed as basic references. Then, the measurement cycle is divided into operating stages such as equipment start-up under no-load, stable load operation, standby with reduced load, and shutdown to ensure that the cycle duration fully encompasses the complete workflow from equipment start-up and steady-state production to shutdown. Furthermore, the cycle level is flexibly fine-tuned by considering the start-up and shutdown sequences and operating rhythms of different equipment such as injection molding machines, robotic arms, mold temperature controllers, and dryers in the workshop. Finally, a standardized electromagnetic interference measurement cycle that can cover various production conditions and adapt to on-site maintenance and production rhythms is determined.

[0033] In each operating state, multiple electromagnetic environment detection nodes are deployed in the injection molding workshop using a grid-based layout. For example, the node deployment needs to be based on the workshop layout, equipment placement, and the routing of temperature signal transmission lines. The grid density is planned according to the principles of uniform coverage and focused density. The density of detection nodes is increased in electromagnetically sensitive areas such as around the injection molding machine, near the mold, and along the temperature sensor signal lines, while the density is appropriately reduced in areas with weaker interference, such as the workshop edges. This ensures that the detection nodes can capture electromagnetic interference signals from different areas. Electromagnetic interference signals are collected synchronously by each detection node, with a unified sampling frequency and data format during the acquisition process. The location coordinates and corresponding operating state of each node are also recorded. Time-frequency analysis is performed on the electromagnetic interference signals of each node. Fourier transform and wavelet transform algorithms are used to convert the time-domain signals into frequency-domain signals. This allows for the extraction of the electromagnetic interference intensity (quantified by electric or magnetic field strength) and frequency range corresponding to different equipment in each operating state, determining the basic interference level of a single device and the interference superposition effect when multiple devices operate together.

[0034] Subsequently, using the location coordinates of each electromagnetic environment detection node as a spatial dimension benchmark, the electromagnetic interference intensity and corresponding frequency range information collected under different equipment operating conditions, such as individual operation and joint operation, are integrated. Through data visualization technology, the interference intensity of different areas is transformed into differentiated color gradients, where the higher the interference intensity, the darker the corresponding color. This generates an electromagnetic interference intensity distribution heatmap that corresponds one-to-one with the operating conditions of each device. The electromagnetic interference intensity distribution heatmap can present the distribution pattern of electromagnetic interference strength at different spatial locations in the workshop under specific equipment operating conditions, as well as the correlation characteristics between this distribution and the interference frequency range, providing clear visualization support for quickly locating high-interference areas.

[0035] From the above, it can be concluded that this embodiment improves the comprehensiveness of the actual production operation scenario by controlling the individual and combined operation of equipment such as injection molding machines, robotic arms, mold temperature controllers, and dryers through setting measurement cycles; the use of grid-based detection nodes to collect interference signals enables the collection of electromagnetic data across the entire workshop; time-frequency analysis to extract the interference intensity and frequency parameters of various equipment and the creation of electromagnetic interference heatmaps can improve the speed of identifying high-interference areas; the step response discrimination method and transfer entropy causal analysis to classify and identify interference sources effectively improve the accuracy of interference source identification; and the extraction of spectral features to determine the center frequency and bandwidth and integrate them to form interference frequency bands can provide a reliable electromagnetic interference basis for subsequent temperature data filtering, ensuring the anti-interference capability of temperature measurement data from the source and further effectively improving the overall accuracy of mold temperature monitoring.

[0036] In one embodiment of this application, a high-interference area is located based on a heat map of electromagnetic interference intensity distribution, and the interference source is identified according to the step response discrimination method for startable / stoppable devices and the causal analysis of the transfer entropy of devices that cannot be started / stopped at any time, including: Based on the electromagnetic interference intensity distribution heat map, the area where the electromagnetic interference intensity is greater than the preset high interference threshold is located as the interference area. For devices that can be independently started and stopped within the interference area, start-stop control is performed. Devices whose electromagnetic interference intensity in the interference area increases stepwise after starting and decreases stepwise after stopping are identified as the first interference source. For devices that cannot be started and stopped at any time, a state space is constructed based on the time series of the operating state parameters of the devices that cannot be started and stopped at any time and the corresponding electromagnetic interference intensity, and the causal contribution is calculated using transfer entropy to obtain the second interference source. The interference source includes the first interference source and the second interference source. The spectral features of the interference sources are extracted to determine the center frequency and frequency bandwidth of each interference source, thus obtaining the interference frequency band.

[0037] In this embodiment, based on a pre-established one-to-one mapping relationship between color levels and electromagnetic interference intensity, the colors are arranged in order from light to dark, sequentially matching the measured electromagnetic interference value ranges from small to large. Simultaneously, linear fitting conversions are performed between the full spectrum color values, grayscale values, and standard interference detection values. Specifically, grayscale sample values ​​corresponding to different color levels are collected along with measured standard electromagnetic interference intensity data at the same location. The least squares method is used to perform linear fitting calculations between color values, grayscale values, and measured interference values, constructing a linear conversion formula. This linear conversion formula enables rapid conversion from heatmap-visualized color data to accurately quantified electromagnetic interference values. The mapping calibration relationship construction method includes: firstly, collecting multiple sets of measured standard electromagnetic interference intensity data under different locations and equipment operating conditions in the workshop, and simultaneously collecting color values ​​and grayscale value samples of different shades of color in the corresponding point heat map; then, using the least squares method to perform linear fitting calculations on the color values, grayscale values ​​and measured electromagnetic interference intensity values, and establishing a fixed matching rule based on the calculation results that the color arrangement order from light to dark corresponds to the electromagnetic interference value range from small to large, completing the linear association between the full spectrum color values, grayscale values ​​and standard interference detection values, and finally forming a mapping calibration relationship that can convert the visual color parameters of the heat map into quantitative electromagnetic interference values.

[0038] For each pixel area and the detection node area corresponding to the grid points in the heat map, the corresponding color parameters are extracted one by one. Based on the linear conversion formula, the intuitive color depth visual features are uniformly converted into quantitative interference values ​​that can participate in numerical comparison and logical judgment, thus completing the transformation of the entire electromagnetic interference intensity distribution heat map from a color visualization graphic to a full-domain digital interference value matrix.

[0039] After converting the heatmap colors into quantized interference values, areas with electromagnetic interference intensity exceeding a preset high interference threshold are located based on the electromagnetic interference intensity distribution heatmap, designated as interference zones. Subsequently, devices within the interference zone that can be independently started and stopped are controlled, with real-time quantized electromagnetic interference values ​​collected synchronously throughout the process. The fluctuation patterns of these values ​​are continuously observed. Devices within the interference zone whose electromagnetic interference intensity exhibits a step increase after startup and a step decrease after shutdown are identified as primary interference sources. The interference characteristics of these primary sources are intuitive, allowing for direct identification of the source based on operating condition switching. For equipment in the injection molding workshop that cannot be stopped and debugged at will, and which cannot be started and stopped at any time during production, the working condition test method is abandoned. Instead, analysis is carried out based on actual production operation data. Based on the time series of the operating status parameters of the equipment that cannot be started and stopped at will and the corresponding electromagnetic interference intensity, a complete data association state space is built. Then, the causal contribution between the two is accurately calculated by using transfer entropy. The target equipment is screened by quantifying the magnitude of causal influence, and finally the second interference source is obtained. The second interference source refers to the electromagnetic interference source determined by the transfer entropy causal analysis of the equipment in the injection molding workshop that cannot be started and stopped at will. Together with the first interference source determined by the equipment that can be started and stopped, they constitute the interference source. That is, the interference source includes two categories: the first interference source and the second interference source. The preset high interference threshold is used as a criterion for determining whether the electromagnetic environment in the injection molding workshop belongs to a strong interference area. The setting of the high interference threshold is based on the following: First, the average value and maximum normal fluctuation value of the electromagnetic interference intensity collected by the electromagnetic environment detection nodes in the entire area are statistically analyzed under various working conditions such as normal no-load production and multi-equipment coordinated steady-state production in the injection molding workshop. This is used as the basic benchmark value. Then, the critical electromagnetic interference limit for normal transmission of workshop temperature monitoring signals without interference is combined with the interference intensity critical value corresponding to the anti-interference capability parameter of temperature sensor signal lines and the actual measured signal distortion. Within a reasonable safe buffer range between the maximum interference fluctuation value and the critical value of signal interference distortion during normal production, short-term high-intensity interference values ​​that occur instantaneously and accidentally are eliminated, and finally, the high interference threshold is determined.

[0040] After all interference sources were identified, the spectral features of the interference sources were extracted in detail. The frequency-related parameters of the electromagnetic signals radiated by each type of interference source were sorted out, the center frequency and frequency bandwidth of each interference source were determined, and the frequency-related data of all interference sources were integrated and classified in a unified manner. Finally, a complete interference frequency band that fits the actual working conditions of the workshop was obtained.

[0041] From the above, it can be concluded that this embodiment, by delineating the interference area based on the electromagnetic interference heat map and threshold, can improve the speed and accuracy of locking the high interference range in the workshop. For startable / stoppable equipment, a step response discrimination method is used to quickly identify the interference source based on the step change in interference intensity before and after start-up and shutdown, resulting in a simple and efficient identification logic. Simultaneously, for non-startable / stoppable equipment, a state space is constructed using time series analysis, and the causal contribution is calculated using transfer entropy to determine the interference source, adapting to the equipment's operating characteristics and avoiding blind spots in identification. Classification is completed by distinguishing between two types of interference sources, making the identification of interference sources more comprehensive and thorough. Finally, the spectral features of the interference source are extracted to determine the center frequency and bandwidth, identifying the corresponding interference frequency band, providing parameter support for subsequent targeted filtering and noise reduction of temperature data.

[0042] In one embodiment of this application, for a device that cannot be started or stopped at any time, a state space is constructed based on the time series of the operating state parameters of the device and the corresponding electromagnetic interference intensity, and the causal contribution is calculated using transfer entropy to obtain a second interference source, including: Collect the sequence of operating status parameters of equipment that cannot be started or stopped at any time in the injection molding workshop, and retrieve the electromagnetic interference intensity of the corresponding electromagnetic monitoring node; The electromagnetic interference intensity of the corresponding electromagnetic monitoring node is discretized and timestamped according to the time sequence to obtain the electromagnetic interference intensity time series. A time-state space is constructed based on the sequence of operating state parameters and the time sequence of electromagnetic interference intensity. In the state space, an asymmetric causal correlation analysis method based on transfer entropy is used to calculate the causal contribution of the state parameter changes of each non-startable and non-stop device to the interference intensity of the corresponding electromagnetic monitoring node. Devices whose causal contribution exceeds a preset causal threshold are identified as the second source of interference.

[0043] In this embodiment, the operating status parameter sequence of equipment that cannot be started or stopped at any time in the injection molding workshop is first collected. For example, the collected parameters include real-time dynamic data such as equipment workload, operating power, operating speed, heating output power, and internal circuit on / off status. The data is continuously collected according to a unified collection frequency to form a time-series coherent and dimensionally complete operating status parameter sequence. At the same time, the electromagnetic interference intensity data of electromagnetic monitoring nodes that are spatially close to the equipment that cannot be started or stopped at any time and have a high degree of signal coupling correlation are retrieved.

[0044] Secondly, based on the time sequence, the electromagnetic interference intensity of the corresponding electromagnetic monitoring nodes is discretized and sampled. Specifically, the continuously changing interference values ​​are truncated and normalized according to a predetermined sampling interval. At the same time, a unified timestamp is added to each set of sampled interference data to determine the specific acquisition time and production process node corresponding to each set of interference data. Invalid noise and abnormal jump values ​​generated during the data acquisition process are removed, and finally, a time series of electromagnetic interference intensity with normalized time sequence and clear time sequence correspondence is generated.

[0045] Subsequently, a time-state space is constructed based on the sequence of operating state parameters and the time series of electromagnetic interference intensity. This time-state space uses the time dimension as a unified benchmark, aligns the changing trends of equipment operating parameters and the fluctuation trends of electromagnetic interference intensity in a time sequence, integrates the changing characteristics, fluctuation amplitudes, and changing rhythms of the two types of data, and builds a multi-dimensional time-state space that can reflect the linkage between equipment operating conditions and electromagnetic interference changes.

[0046] After completing the state space construction, an asymmetric causal correlation analysis method based on transfer entropy is used to conduct deep computation. Unlike ordinary correlation analysis, this method can accurately distinguish the unidirectional influence relationship between data, effectively eliminate the interference of data synchronization fluctuations caused by irrelevant environmental factors, and sequentially substitute the time-series data of each device that cannot be started or stopped at any time to quantitatively calculate the causal contribution of the state parameter changes of each device to the interference intensity of the corresponding electromagnetic monitoring node.

[0047] Finally, based on the measured data of the electromagnetic environment in the workshop and industry experience, a causal threshold was set. The causal contribution of each type of equipment was calculated and compared with the preset causal threshold. Equipment with a causal contribution greater than the causal threshold was officially identified as the second source of interference. This completed the classification of all interference equipment that cannot be started or stopped at any time and improved the overall identification system of electromagnetic interference sources in the workshop.

[0048] From the above, it can be concluded that this embodiment collects the operating parameter sequence of non-startable and non-stop equipment and matches it with corresponding electromagnetic interference data to achieve accurate data pairing; it performs discrete sampling and timestamp marking on the interference intensity to ensure accurate time sequence alignment by organizing the time-series data; it constructs a time-state space by combining the two types of sequences to build a complete analytical and computational foundation; it conducts asymmetric causal correlation analysis based on the transfer entropy, which can improve the accuracy of quantifying the impact of equipment operation on electromagnetic interference; and it screens and determines the second interference source by setting a causal threshold, which can uncover hidden interference sources without affecting normal production order, improve the workshop electromagnetic interference database, and effectively enhance the comprehensiveness and reliability of interference identification under complex working conditions.

[0049] In one embodiment of this application, the initial temperature data is filtered based on the interference source and its corresponding interference frequency band, and the filtered data is compensated and corrected based on the thermal inertia of the preform injection mold to obtain the target temperature data, including: The initial temperature data is decomposed into wavelet packets using a preset second time window to obtain wavelet packet coefficients corresponding to multiple frequency bands. Based on the interference frequency band corresponding to the interference source, determine the target frequency band affected by the interference; Based on the preset coupling transfer function of each interference source to the temperature signal transmission line, and the operating status of each interference source, the coupling interference intensity in the target frequency band is estimated. Construct a masking threshold based on the coupling interference strength; Wavelet packet coefficients with moduli less than the masking threshold are set to zero, and wavelet packet coefficients with moduli greater than the masking threshold are subjected to soft threshold shrinkage to obtain the target wavelet packet coefficients. The target wavelet packet coefficients are reconstructed using inverse wavelet packets, and the reconstructed value at the end of the second time window is taken as the filtered temperature data. Based on the thermal inertia constant of the preform injection mold, hysteresis compensation is performed on the filtered temperature data to obtain the target temperature data.

[0050] In this embodiment, the initial temperature data is first decomposed using a preset second time window. During actual execution, a continuous data acquisition interval is defined according to the predetermined second time window. The complete initial temperature time-series signal within the window is then subjected to multi-level refined frequency domain decomposition, ultimately dividing the signal into multiple different frequency intervals. Simultaneously, the wavelet packet coefficients corresponding to each frequency band are calculated. The wavelet packet coefficients are quantitative values ​​representing the signal energy magnitude and fluctuation characteristics within different frequency bands after wavelet packet decomposition. Specifically, based on the duration of each process stage in preform injection molding, the rate of change in mold temperature response, and the period of electromagnetic interference fluctuations in the workshop, and simultaneously matching the temperature data sampling frequency, the second time window is selected to balance the accuracy of wavelet packet decomposition frequency band division with data real-time performance. This selection allows for the complete capture of local temperature fluctuations without causing data lag or accumulation.

[0051] Based on the interference source and its corresponding interference frequency band, a matching comparison is performed within all the frequency bands that have been split, and the frequency range that overlaps with the electromagnetic interference frequency band in the workshop and is affected by electromagnetic clutter is locked in order to determine the target frequency band affected by the interference.

[0052] Based on this, the coupling transfer function of each interference source to the temperature signal transmission line is retrieved. According to the real-time operational status of each interference source acquired on-site, multiple influencing factors such as spatial transmission distance, line shielding effectiveness, and equipment operating load are comprehensively summarized, and the actual coupling interference intensity within the target frequency band is estimated. The coupling transfer function is a mathematical model characterizing the strength and conduction law of electromagnetic interference from interfering equipment to the temperature sensing transmission line. The coupling interference intensity is the actual noise magnitude superimposed on the temperature data after magnetic interference couples into the temperature measurement signal along the line. The coupling transfer function of each interference source to the temperature signal transmission line is a mathematical mapping model characterizing the strength and variation law of electromagnetic interference generated by various equipment in the injection molding workshop propagating into the temperature measurement signal line. Basic environmental parameters such as the spatial layout distance between each interfering device and the temperature signal transmission line, the line shielding protection level, the wiring direction, and the layout of on-site metal components were measured. Then, through working condition tests, the electromagnetic signal parameters radiated by the interference source and the interference signal data coupled to the temperature measurement line were collected under different load operating conditions. The electromagnetic coupling transmission mechanism was integrated to establish a basic function model. Then, the parameters of the basic function model were fitted and corrected using multiple sets of measured working condition sample data. At the same time, variables such as the change of workshop equipment operating load and the electromagnetic superposition effect of surrounding equipment were incorporated to complete iterative optimization. Finally, a coupling transfer function that is suitable for the actual working conditions on site and can accurately characterize the interference conduction characteristics was obtained.

[0053] Secondly, the core of constructing a masking threshold based on the coupling interference strength is to establish a dynamically adaptable quantitative judgment standard based on the actual value of the coupling interference strength estimated within the target frequency band, and according to the amplitude characteristics and signal-to-noise ratio requirements of the temperature signal itself. Specifically, this masking threshold is adjusted in real time according to changes in the operating state of the interference source and fluctuations in the coupling interference strength, used to distinguish between useless interference noise and the effective true temperature signal in the signal. Specifically, using the peak value, mean value, and fluctuation variance of the coupling interference strength as core parameters, a masking threshold highly adapted to the current interference environment is constructed through mathematical modeling, enabling this masking threshold to serve as an accurate basis for subsequent wavelet packet coefficient selection.

[0054] Subsequently, according to established processing rules, all wavelet packet coefficients are differentiated. Wavelet packet coefficients with magnitudes less than the masking threshold are zeroed out to remove invalid signal components representing weak electromagnetic interference. Simultaneously, wavelet packet coefficients with magnitudes greater than the masking threshold undergo soft-thresholding contraction, for example, to smooth and compress the abnormal signal amplitude generated by interference superposition, avoiding waveform distortion and feature loss caused by hard-thresholding. The target wavelet packet coefficients are obtained after these two processing methods. The established processing rules include two data processing methods based on the relationship between the masking threshold and the magnitude of the wavelet packet coefficients: first, wavelet packet coefficients with magnitudes less than the masking threshold are directly zeroed out to completely filter out invalid noise components formed by weak electromagnetic interference; second, wavelet packet coefficients with magnitudes greater than the masking threshold undergo soft-thresholding contraction using a set contraction coefficient and contraction amplitude to smooth and compress abnormal signal amplitudes. The execution order and parameter adjustment standards for the two processing methods are also determined. The shrinkage coefficient is determined based on the normal fluctuation range of the preform injection mold temperature, the signal-to-noise ratio requirements of the temperature measurement signal, and the fluctuation range of the electromagnetic coupling interference intensity on site. Under the premise of ensuring that the actual temperature change trend does not deviate, the basic value is determined by referring to the degree of signal distortion caused by interference superposition under different working conditions. Then, the coefficient is finely adjusted by combining the temperature change characteristics of different areas of the cavity, core, and runner. At the same time, it is repeatedly calibrated by combining the soft threshold noise reduction simulation results.

[0055] The inverse wavelet packet reconstruction operation is performed using the processed target wavelet packet coefficients. For example, the various frequency band signals after splitting are re-integrated and restored into a continuous and complete temperature time series signal, thus completing the signal noise reduction and restoration work. Then, the reconstructed value corresponding to the end time of the second time window is uniformly extracted, and the reconstructed value is determined as the filtered temperature data.

[0056] Finally, based on the thermal inertia constant of the preform injection mold, hysteresis compensation is performed on the filtered temperature data to obtain the target temperature data. The thermal inertia constant is a fixed parameter characterizing the inherent thermal properties of the preform injection mold, which exhibits slow heating and delayed cooling, and reflects the degree of temperature response delay of the mold.

[0057] From the above, it can be concluded that this embodiment uses a second time window to capture data and complete wavelet packet decomposition, which can orderly split the signal frequency band and facilitate layered data processing; it locks the disturbed target frequency band based on the known interference frequency band, achieving precise targeted noise reduction; it estimates the interference intensity based on the coupling transfer function and equipment operating conditions, making the noise judgment more consistent with the actual transmission law; it processes the data by setting the coefficients to zero and using soft threshold shrinkage, which can remove interference noise while retaining the true temperature change characteristics; it obtains clean filtered data through inverse wavelet packet reconstruction, ensuring the integrity of the temperature sensing electrical signal; and it completes hysteresis compensation based on the thermal inertia constant, effectively reducing the mold temperature measurement delay deviation and further improving the accuracy of temperature monitoring.

[0058] In one embodiment of this application, based on the thermal inertia constant of the preform injection mold, hysteresis compensation is performed on the filtered temperature data to obtain target temperature data, including: Obtain the three-dimensional geometric model, material thermal properties parameters, and wall thickness distribution data of the preform injection mold, and establish a three-dimensional unsteady-state heat conduction finite element model; Based on the wall thickness distribution data, the cavity surface is divided into multiple parameter zones; Based on the correspondence between the wall thickness value and the thermal inertia constant of each parameter region, the thermal inertia constant correction factor for each parameter region is determined. The thermal inertia constant correction factor was calibrated by global inversion using the adjoint variable method, resulting in a target three-dimensional unsteady heat conduction finite element model. The target three-dimensional unsteady heat conduction finite element model is embedded in a Kalman filter, and the filtered temperature data is used as input to estimate the true temperature distribution on the cavity surface, thus obtaining the target temperature data.

[0059] In this embodiment, the core basic data of the preform injection mold is first obtained. This core basic data includes a three-dimensional geometric model, material thermal properties, and wall thickness distribution data. The three-dimensional geometric model is obtained through 3D scanning or reconstruction from design drawings, and includes spatial structural data of the cavity, core, runner, and cooling channels. The material thermal properties are determined based on material testing reports or industry standard databases, including the thermal conductivity, specific heat capacity, and density of the mold material. The wall thickness distribution data is obtained through wall thickness detection equipment, including thickness information at key locations such as cavity corners, runner intersections, and sidewalls. Based on this data, a three-dimensional unsteady-state heat conduction finite element model is constructed using the finite element analysis method. This model can simulate the dynamic heat transfer process of the mold during the injection molding cycle, including heat transfer during melt filling and heat dissipation during the cooling stage.

[0060] Next, the cavity surface is divided into zones based on the wall thickness distribution data. Since differences in wall thickness across different areas of the mold affect heat transfer rate and heat storage capacity—with greater wall thickness resulting in stronger thermal inertia and a more delayed temperature response—the cavity surface is divided into multiple parameter zones based on the principles of wall thickness uniformity and structural correlation. This ensures that the wall thickness fluctuation within each parameter zone is controlled within a preset range. This preset range is based on the typical wall thickness difference range between different molding areas of the preform injection mold, the tolerance for heat transfer rate deviation within the same area, the mold zone temperature measurement accuracy requirements, and the criteria for judging thermal inertia uniformity. Furthermore, reasonable fluctuation limits are determined by referring to a large amount of measured wall thickness distribution data from similar molds. The tolerance for heat transfer rate deviation is the maximum reasonable deviation allowed when there are differences in heat transfer speed at different points within the same area of ​​the mold. Exceeding the tolerance for heat transfer rate deviation will cause excessive differences in thermal inertia, making it impossible to uniformly compensate for temperature lag.

[0061] Subsequently, based on the preset correspondence between the wall thickness values ​​and thermal inertia constants of each parameter zone, a thermal inertia constant correction factor for each parameter zone is determined. This correction factor is used as an adjustable proportionality coefficient to correct deviations in the standard thermal inertia constant. The preset correspondence is obtained through mold thermal characteristic experiments and data fitting, showing a positive correlation between wall thickness and thermal inertia constant. After obtaining the thermal inertia constant correction factor, the adjoint variable method is used to perform an overall inversion calibration of the correction factor, resulting in the target three-dimensional unsteady-state heat conduction finite element model. The three-dimensional unsteady-state heat conduction encompasses the mold's length, width, and thickness in three dimensions, while also considering the dynamic temperature changes over time during production.

[0062] Finally, the target 3D unsteady-state heat conduction finite element model is embedded into a Kalman filter to construct a fusion compensation framework of the physical model and the filtering algorithm. Using the filtered temperature data as the filter's observation input, the target 3D unsteady-state heat conduction finite element model is used to predict the temperature change trend of the mold cavity. Simultaneously, the Kalman filter's state estimation and update mechanism corrects the deviation between the predicted and observed values. This process not only uses the physical model to reconstruct the temperature hysteresis caused by thermal inertia but also uses the filtering algorithm to offset the effects of modeling errors and residual interference. The final output temperature data is the target temperature data, which reflects the actual temperature state of the mold cavity.

[0063] From the above, it can be concluded that this embodiment constructs a finite element model of heat conduction based on the mold's three-dimensional model, thermal properties, and wall thickness data, which can realistically reproduce the overall heat transfer state of the mold. Furthermore, dividing the cavity into parameter zones based on wall thickness enables refined temperature analysis by zone. Matching thermal inertia correction factors to wall thickness adapts to the differences in heat transfer across different regions. Inverting calibration factors using the adjoint variable method effectively improves the accuracy of model parameters. Embedding the optimized model into a Kalman filter and extrapolating the actual cavity temperature from the filtered temperature data effectively compensates for the temperature measurement lag error caused by the mold's thermal inertia, further improving the accuracy of the overall mold temperature measurement and providing temperature data support for temperature anomaly detection and production process control.

[0064] In one embodiment of this application, the adjoint variable method is used to perform a global inversion calibration of the thermal inertia constant correction factor to obtain a target three-dimensional unsteady-state heat conduction finite element model, including: The thermal inertia constant correction factor was used as the initial guess value, and the forward transient heat conduction calculation was performed based on the three-dimensional unsteady heat conduction finite element model to obtain the simulated temperature field value. Based on the standard temperature data of historical preform injection molds under injection molding, a temperature residual functional is constructed between the simulated temperature field value and the temperature data. Based on the temperature residual functional, the adjoint equation is derived and solved to obtain the adjoint temperature field. Based on the accompanying temperature field and the simulated temperature field values, the sensitivity of the temperature residual functional to the correction factor of the thermal inertia constant in each parameter region is calculated. Based on sensitivity, the thermal inertia constant correction factor for each parameter region is iteratively updated using the gradient descent optimization algorithm until the temperature residual functional converges to the expected accuracy, thus obtaining the converged thermal inertia constant correction factor. The converged thermal inertia constant correction factor is updated into the three-dimensional unsteady heat conduction finite element model to obtain the target three-dimensional unsteady heat conduction finite element model.

[0065] In this embodiment, the pre-determined thermal inertia constant correction factor is first set as the initial guess value for model calculation. Based on the completed three-dimensional unsteady heat conduction finite element model, the global forward transient heat conduction calculation is carried out to completely replicate the dynamic changes of heat conduction, heat accumulation and heat dissipation inside the mold during the entire injection molding process. The simulated temperature field values ​​corresponding to different times within the overall space of the mold are accurately solved, and the forward temperature simulation of the model is completed.

[0066] Subsequently, the standard temperature data corresponding to the injection molds of the preforms stored in the workshop under injection conditions were retrieved. This type of data is all reliable process temperature data that has been measured and verified on-site. Using this as the real temperature reference benchmark, and in accordance with the established mathematical construction rules, the temperature residual functional between the simulated temperature field value and the measured standard temperature data was built. This intuitively quantifies the overall deviation between the model simulation results and the actual operating temperature, and establishes a clear convergence judgment basis for subsequent parameter optimization.

[0067] After constructing the temperature residual functional, the formula derivation was completed by using this functional as the core derivation basis and combining it with the physical control equation of three-dimensional unsteady heat conduction. The adjoint equation adapted to this heat conduction simulation scenario was then obtained. The adjoint equation was then calculated and analyzed by numerical solution, and the corresponding adjoint temperature field was finally obtained. The adjoint temperature field can be used to efficiently build a bridge between temperature deviation and internal parameters of the model, simplifying the subsequent sensitivity solution process.

[0068] Next, based on the already solved temperature field and the simulated temperature field values ​​obtained in the previous calculations, joint calculations are carried out according to the established integration rules. The sensitivity values ​​of the temperature residual functional corresponding to the thermal inertia constant correction factors for each parameter region are accurately calculated. The magnitude of the change in the overall temperature simulation deviation when the correction factor of each parameter region changes is clearly quantified, and the priority and direction of optimization adjustment of each parameter are clarified.

[0069] After obtaining the sensitivity results for all parameters, the gradient descent optimization algorithm is officially activated. A reasonable iteration step size is determined in combination with the sensitivity values. The thermal inertia constant correction factor for each parameter region is iteratively updated and adjusted one by one. After each round of parameter update, the forward heat conduction calculation is carried out again and the temperature residual functional value is verified. The optimization process is continuously iterated and cyclically optimized until the temperature residual functional is stably converged to within the preset accuracy range. At this point, the converged thermal inertia constant correction factor can be determined.

[0070] Finally, the thermal inertia constant correction factor, after multiple rounds of iterative optimization and convergence to meet the expected accuracy requirements, is batch-updated and synchronously entered into the original three-dimensional unsteady-state heat conduction finite element model. This completes the overall replacement and optimization upgrade of the model's core thermal characteristic parameters. The resulting new simulation model, after parameter updates, is a more accurate and realistic three-dimensional unsteady-state heat conduction finite element model, providing accurate and reliable simulation model support for subsequent mold temperature lag compensation. The expected accuracy is a pre-set convergence criterion for the temperature residual functional, referring to the maximum allowable overall deviation limit between the simulated mold temperature field value and the actual measured standard temperature data. When the temperature residual functional value shrinks and stabilizes below this expected accuracy during iterative optimization, it indicates that the thermal inertia constant correction factor has been properly adjusted, and the temperature error between the simulated temperature field value and the operating temperature meets the requirements for injection molding temperature measurement. The iteration can then be stopped to complete parameter calibration. The setting is based on the temperature control tolerance of preform injection molding, the allowable error range of temperature measurement in different areas of the mold, the required accuracy level for temperature lag compensation, and also considers simulation calculation efficiency, defining a reasonable deviation threshold.

[0071] From the above, it can be concluded that this embodiment uses the correction factor as the initial value to carry out forward heat conduction calculations, thereby improving the efficiency of obtaining the initial temperature simulation results; constructing a residual functional based on standard temperature data improves the accuracy of quantifying the deviation between the simulated temperature and the actual temperature; obtaining the accompanying temperature field by solving the adjoint equation effectively establishes the basis for parameter optimization analysis; calculating the parameter sensitivity based on the two types of temperature fields determines the degree of influence of factor adjustment; iteratively optimizing the correction factor using the gradient descent algorithm steadily reduces the numerical deviation; and completing the parameter update and model finalization after the residual convergence effectively improves the degree to which the heat conduction model fits the actual working conditions.

[0072] In one embodiment of this application, the sensitivity of the temperature residual functional to the correction factor of the thermal inertia constant in each parameter region is calculated based on the accompanying temperature field and the simulated temperature field value, including: The product field is obtained by multiplying the simulated temperature field value with the accompanying temperature field value at the same spatial location and the same time node point by point. The product field is integraled in volume over the spatial range of each parameter region, and then integraled over a preset third time window on the time axis to obtain the initial sensitivity corresponding to each parameter region. Based on the wall thickness gradient direction of each parameter interval, the initial sensitivity of adjacent parameter intervals is spatially smoothed to obtain the sensitivity.

[0073] In this embodiment, the simulated temperature field value is first matched with the accompanying temperature field, and corresponding values ​​at the same spatial location and time node are selected. Then, a point-by-point multiplication operation is performed across the entire domain to obtain the product field. This product field can reflect the coupling relationship between the simulated temperature distribution and the deviation correlation field in the spatiotemporal dimension, and completely retain the linkage information of the two types of field quantities at different locations and times.

[0074] After constructing the product field, spatial and temporal integration operations are performed sequentially to obtain the basic parameters. Specifically, the spatial boundary ranges corresponding to each pre-defined parameter region are first delineated, and volume integration is performed on the product field within the corresponding spatial range. Simultaneously, temporal integration is performed within the complete time axis interval corresponding to the preset third time window, uniformly aggregating the dynamic effects under the time dimension. After the double integration operation, the initial sensitivity corresponding to each parameter region can be solved one by one. The initial sensitivity is used to preliminarily determine the basic influence of the correction factor of different regions on the overall temperature residual. Among them, the preset third time window is based on the complete molding process cycle of preform injection molding, selecting the duration covering the heat transfer stages such as melt filling, holding pressure, and cooling. By matching the dynamic response duration of mold heat transfer and temperature field changes, the entire process of temperature residual evolution is taken into account. At the same time, it meets the data integration requirements of thermal inertia parameter correction, and the continuous duration that can completely aggregate the dynamic effects of heat conduction under the time dimension is selected for setting.

[0075] Because the wall thickness of the preform injection mold cavity exhibits a continuous and gradual change, the heat conduction state between adjacent parameter areas influences each other. Using the initial sensitivity can lead to abrupt numerical changes and large deviations in regional connections. Therefore, optimization based on the actual structural characteristics is necessary. For example, using the actual wall thickness gradient direction within each parameter interval as the adjustment basis, and following the trend of wall thickness change from thick to thin or from thin to thick, spatial smoothing is performed on the initial sensitivities obtained from adjacent parameter areas. This weakens abrupt fluctuations in sensitivity values ​​at interval boundaries, balances the correlation of parameter influence between adjacent regions, and after smoothing correction, a sensitivity that satisfies the heat transfer distribution law and has a continuous and stable numerical distribution is finally obtained.

[0076] From the above, it can be concluded that this embodiment constructs a product field by multiplying the numerical values ​​of the two locations simultaneously, which can improve the accuracy of the correlation between the two data fields. Furthermore, performing spatial volume integration and time integration within a specified time window on the product field can improve the efficiency of calculating the initial sensitivity values ​​of each region. By performing spatial smoothing processing on the initial sensitivity of adjacent regions based on the wall thickness gradient direction, the numerical abrupt deviation between regions can be reduced, making the sensitivity results more consistent with the actual heat transfer law of the mold structure. This effectively improves the rationality of subsequent factor iteration calibration and the overall model heat transfer fitting accuracy, ensuring the stable and reliable temperature hysteresis compensation effect.

[0077] In one embodiment of this application, before generating a temperature anomaly warning signal, the method further includes: Obtain production plan information from the centralized monitoring platform of the injection molding workshop. The production plan information includes the planned equipment start-up and shutdown times and process changeover times. The time when the temperature deviation value exceeds the first interval is matched with the planned equipment start-up and shutdown time and process switch time. If the time difference between the moment when the temperature deviation exceeds the first interval and any planned time point is less than the preset fourth time window, the temperature deviation is determined to be a planned fluctuation, and the generation of a temperature anomaly warning signal is suppressed. If the time difference between the moment when the temperature deviation exceeds the first interval and all planned time points is greater than or equal to the fourth time window, a temperature anomaly warning signal will be generated normally.

[0078] In this embodiment, the production plan information issued by the centralized monitoring platform of the injection molding workshop is first obtained. This production plan information integrates the overall production scheduling and process adjustment arrangements of the workshop, including the planned equipment start-up and shutdown times and process switching times. At the same time, the process content, adjustment type and preset temperature fluctuation range corresponding to various planned time points are stored simultaneously.

[0079] Once a temperature deviation exceeding the first range is identified during monitoring, the specific time when the temperature deviation occurred is retrieved. This specific time of anomaly is then precisely matched with the planned equipment start-up and shutdown times and process switching times to calculate the time interval between them. This time-series-level benchmarking and screening work is completed to determine whether the temperature deviation is related to the preset production scheduling actions.

[0080] In this embodiment, after time matching and comparison, if the time difference between the moment when the temperature deviation exceeds the first interval and any planned time point is less than the preset fourth time window, it indicates that the temperature deviation occurs precisely during the routine production adjustment phase of equipment start-up and shutdown and process parameter switching. This is a normal dynamic temperature change caused by production process changes, and the temperature deviation is immediately determined to be a planned fluctuation. Lockout control logic is then executed simultaneously to actively suppress the generation of temperature anomaly warning signals. The fourth time window integrates the duration of natural mold temperature fluctuations during injection molding machine start-up and shutdown and process switching, matches the mold temperature adjustment response delay with the duration of temperature fluctuations caused by production process adjustments, and then statistically analyzes the average duration and longest duration of normal temperature deviations before and after each process switch and equipment start-up / shutdown. This final time range is used to distinguish between planned process disturbances and actual temperature anomalies.

[0081] If, after comprehensive comparison and confirmation, the time difference between the moment when the temperature deviation exceeds the first interval and all planned time points is greater than or equal to the fourth time window, it indicates that the temperature deviation is not caused by routine process adjustments within the production plan, but rather by an unexpected temperature anomaly that deviates from the preset production rhythm. There is no reasonable cause for the process change. At this time, a temperature anomaly warning signal is generated normally according to the established monitoring logic, and an anomaly prompt is pushed out in a timely manner to ensure that the workshop can detect and deal with the real mold temperature failure problem as soon as possible.

[0082] As can be seen from the above, this embodiment, by retrieving production plan information from the workshop monitoring platform, grasps the predetermined time nodes for equipment start-up, shutdown, and process switching, and compares the time of temperature exceeding the standard with the planned time, which can improve the efficiency of distinguishing the causes of temperature fluctuations. By using a fourth time window to complete time difference determination, it effectively identifies normal planned temperature fluctuations within the production process; it proactively suppresses the triggering of warnings for compliant fluctuations, issuing warnings only when temperature anomalies occur during unplanned periods. This pre-judgment logic can effectively distinguish between normal process fluctuations and actual temperature faults, reducing invalid alarms and false warnings from the source, avoiding interference from redundant information for staff, ensuring timely detection and handling of abnormal temperature problems, and improving the practicality and accuracy of the workshop temperature monitoring and early warning system.

[0083] Corresponding to the temperature monitoring method for preform injection molds described in the above embodiments, Figure 2 This is a structural block diagram of a temperature monitoring system for a preform injection mold, provided as an embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown. References Figure 2The temperature monitoring system 20 applied to preform injection molds includes: an electromagnetic interference detection and identification module 21, a temperature data isolation and acquisition module 22, a data filtering compensation and correction module 23, a temperature anomaly judgment and early warning module 24, and a monitoring report generation and uploading module 25.

[0084] Among them, the electromagnetic interference detection and identification module 21 is used to measure the electromagnetic interference intensity and frequency range of each piece of equipment in the injection molding workshop under different operating conditions using electromagnetic environment detection equipment, and to identify the interference source and its corresponding interference frequency band. Temperature data isolation acquisition module 22 is used to acquire temperature data and transmit the temperature data to the data acquisition circuit after photoelectric isolation and magnetic isolation to obtain initial temperature data. The data filtering compensation and correction module 23 is used to filter the initial temperature data based on the interference source and its corresponding interference frequency band, and to compensate and correct the filtered data based on the thermal inertia of the preform injection mold to obtain the target temperature data. The temperature anomaly detection and early warning module 24 is used to compare the target temperature data with the preset standard temperature curve and calculate the temperature deviation value and the deviation change rate. If the temperature deviation value exceeds the first interval and the duration is greater than the preset first time window, a temperature anomaly early warning signal is generated, and the temperature anomaly trend type is determined according to the deviation change rate to obtain the target anomaly trend type. The monitoring report generation and upload module 25 is used to generate a temperature monitoring report based on the temperature anomaly warning signal and the target anomaly trend type, and upload it to the centralized monitoring platform of the injection molding workshop.

[0085] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the modules in the aforementioned device embodiments, for example... Figure 2 The electromagnetic interference detection and identification module 21, temperature data isolation and acquisition module 22, data filtering compensation and correction module 23, temperature anomaly judgment and early warning module 24, and monitoring report generation and uploading module 25 shown have the following functions.

[0086] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0087] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0088] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store device type information.

[0089] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation methods described in any embodiment of the temperature monitoring method for preform injection molds provided in the embodiments of this application, or they can execute the implementation methods of the electronic devices described in the embodiments of this application, which will not be repeated here.

[0090] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0091] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0092] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0093] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0094] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device 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. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces or units, or it may be an electrical, mechanical, or other form of connection.

[0095] 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 the embodiments of this application, depending on actual needs.

[0096] 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.

[0097] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for temperature monitoring applied to preform injection molds, characterized in that, include: Electromagnetic environment detection equipment was used to measure the electromagnetic interference intensity and frequency range of each piece of equipment in the injection molding workshop under different operating conditions, and the interference source and its corresponding interference frequency band were identified. Temperature data is acquired and transmitted to the data acquisition circuit after photoelectric and magnetic isolation processing to obtain initial temperature data. Based on the interference source and its corresponding interference frequency band, the initial temperature data is filtered, and the filtered data is compensated and corrected based on the thermal inertia of the preform injection mold to obtain the target temperature data. The target temperature data is compared with a preset standard temperature curve, and the temperature deviation value and the deviation change rate are calculated. If the temperature deviation value exceeds the first interval and the duration is greater than the preset first time window, a temperature anomaly warning signal is generated, and the temperature anomaly trend type is determined based on the deviation change rate to obtain the target anomaly trend type. Based on the temperature anomaly warning signal and the target anomaly trend type, a temperature monitoring report is generated and uploaded to the centralized monitoring platform of the injection molding workshop.

2. The temperature monitoring method for preform injection molds according to claim 1, characterized in that, The equipment in the injection molding workshop includes injection molding machines, robotic arms, mold temperature controllers, and dryers; The method of using electromagnetic environment detection equipment to measure the electromagnetic interference intensity and frequency range of equipment in the injection molding workshop under different operating conditions, and to identify the interference source and its corresponding interference frequency band, includes: According to the preset measurement cycle, the equipment in the injection molding workshop is controlled to run individually and in combination in sequence to obtain multiple operating states; In each operating state, multiple electromagnetic environment detection nodes are arranged in the injection molding workshop using a grid-based layout, and electromagnetic interference signals of each node are collected. Time-frequency analysis was performed on the electromagnetic interference signals of each node, and the electromagnetic interference intensity and frequency range of different devices under each operating state were extracted. Based on the electromagnetic interference intensity and frequency range of different nodes under different operating conditions, a heat map of electromagnetic interference intensity distribution for each device under its operating condition is drawn. Based on the electromagnetic interference intensity distribution heat map, high interference areas are located, and interference sources are identified according to the step response discrimination method of startable and stopable equipment and the transfer entropy causal analysis of equipment that cannot be started or stopped at any time. The spectral features of the identified interference sources are extracted, and the center frequency and frequency bandwidth of each interference source are obtained through frequency domain peak search and bandwidth analysis. The results are then summarized to form the interference frequency band.

3. The temperature monitoring method for preform injection molds according to claim 2, characterized in that, The process of locating high-interference areas based on the electromagnetic interference intensity distribution heatmap, and identifying interference sources based on the step response discrimination method for startable / stoppable devices and the transfer entropy causal analysis of devices that cannot be started / stopped at any time, includes: Based on the electromagnetic interference intensity distribution heat map, the region where the electromagnetic interference intensity is greater than the preset high interference threshold is located as the interference region. For devices that can be independently started and stopped within the interference area, start-stop control is performed. Devices whose electromagnetic interference intensity in the interference area increases stepwise after startup and decreases stepwise after shutdown are identified as the first interference source. For devices that cannot be started and stopped at any time, a state space is constructed based on the time series of the operating state parameters of the devices that cannot be started and stopped at any time and the corresponding electromagnetic interference intensity, and the causal contribution is calculated using transfer entropy to obtain the second interference source. The interference source includes the first interference source and the second interference source. The spectral features of the interference sources are extracted to determine the center frequency and frequency bandwidth of each interference source, thus obtaining the interference frequency band.

4. The temperature monitoring method for preform injection molds according to claim 3, characterized in that, For equipment that cannot be started or stopped at any time, a state space is constructed based on the time series of the operating state parameters of the equipment and the corresponding electromagnetic interference intensity, and the causal contribution is calculated using transfer entropy to obtain the second interference source, including: Collect the sequence of operating status parameters of equipment that cannot be started or stopped at any time in the injection molding workshop, and retrieve the electromagnetic interference intensity of the corresponding electromagnetic monitoring node; The electromagnetic interference intensity of the corresponding electromagnetic monitoring node is discretized and timestamped according to the time sequence to obtain the electromagnetic interference intensity time series. Based on the operating state parameter sequence and the electromagnetic interference intensity time sequence, a time state space is constructed; In the state space, an asymmetric causal correlation analysis method based on transfer entropy is used to calculate the causal contribution of the state parameter changes of each non-startable and non-stop device to the interference intensity of the corresponding electromagnetic monitoring node. Devices whose causal contribution is greater than a preset causal threshold are identified as the second interference source.

5. The temperature monitoring method for preform injection molds according to claim 1, characterized in that, The process of filtering the initial temperature data based on the interference source and its corresponding interference frequency band, and compensating and correcting the filtered data based on the thermal inertia of the preform injection mold to obtain the target temperature data, includes: The initial temperature data is subjected to wavelet packet decomposition using a preset second time window to obtain wavelet packet coefficients corresponding to multiple frequency bands; Based on the interference frequency band corresponding to the interference source, determine the target frequency band affected by the interference; Based on the preset coupling transfer function of each interference source to the temperature signal transmission line, and the operating status of each interference source, the coupling interference intensity in the target frequency band is estimated. Construct a masking threshold based on the coupling interference strength; Wavelet packet coefficients whose modulus is less than the masking threshold are set to zero, and wavelet packet coefficients whose modulus is greater than the masking threshold are subjected to soft threshold shrinkage to obtain the target wavelet packet coefficients. The target wavelet packet coefficients are reconstructed using inverse wavelet packets, and the reconstructed value at the end of the second time window is taken as the filtered temperature data. Based on the thermal inertia constant of the preform injection mold, hysteresis compensation is performed on the filtered temperature data to obtain the target temperature data.

6. The temperature monitoring method for preform injection molds according to claim 5, characterized in that, The target temperature data is obtained by performing hysteresis compensation on the filtered temperature data based on the thermal inertia constant of the preform injection mold, including: Obtain the three-dimensional geometric model, material thermal properties parameters, and wall thickness distribution data of the preform injection mold, and establish a three-dimensional unsteady-state heat conduction finite element model; Based on the wall thickness distribution data, the cavity surface is divided into multiple parameter regions; Based on the correspondence between the wall thickness value and the thermal inertia constant of each parameter region, the thermal inertia constant correction factor for each parameter region is determined. The thermal inertia constant correction factor was calibrated by global inversion using the adjoint variable method to obtain the target three-dimensional unsteady heat conduction finite element model; The target three-dimensional unsteady heat conduction finite element model is embedded in a Kalman filter, and the filtered temperature data is used as input to estimate the true temperature distribution on the cavity surface, thereby obtaining the target temperature data.

7. The temperature monitoring method for preform injection molds according to claim 6, characterized in that, The method of using adjoint variables to perform a global inversion calibration of the thermal inertia constant correction factor yields the target three-dimensional unsteady-state heat conduction finite element model, including: Using the thermal inertia constant correction factor as the initial guess value, and performing forward transient heat conduction calculations based on the three-dimensional unsteady heat conduction finite element model, the simulated temperature field value is obtained. Based on the standard temperature data of historical preform injection molds under injection molding, a temperature residual functional is constructed between the simulated temperature field value and the temperature data. Based on the temperature residual functional, the adjoint equation is derived and solved to obtain the adjoint temperature field. Based on the accompanying temperature field and the simulated temperature field values, the sensitivity of the temperature residual functional to the correction factor of the thermal inertia constant in each parameter region is calculated. Based on the sensitivity, the thermal inertia constant correction factor for each parameter region is iteratively updated using the gradient descent optimization algorithm until the temperature residual functional converges to the expected accuracy, thus obtaining the converged thermal inertia constant correction factor. The converged thermal inertia constant correction factor is updated in the three-dimensional unsteady heat conduction finite element model to obtain the target three-dimensional unsteady heat conduction finite element model.

8. The temperature monitoring method for preform injection molds according to claim 7, characterized in that, The calculation of the sensitivity of the temperature residual functional to the correction factor of the thermal inertia constant in each parameter region based on the accompanying temperature field and the simulated temperature field includes: The simulated temperature field value and the accompanying temperature field value at the same spatial location and the same time node are multiplied point by point to obtain the product field; The product field is integraled in volume over the spatial range of each parameter region, and then integraled over a preset third time window on the time axis to obtain the initial sensitivity corresponding to each parameter region. Based on the wall thickness gradient direction of each parameter interval, the initial sensitivity of adjacent parameter intervals is spatially smoothed to obtain the sensitivity.

9. The temperature monitoring method for preform injection molds according to claim 1, characterized in that, Before generating the abnormal temperature warning signal, the following steps are also included: Obtain production plan information from the centralized monitoring platform of the injection molding workshop, including planned equipment start-up and shutdown times and process switching times; The time when the temperature deviation value exceeds the first interval is matched with the planned equipment start-up and shutdown time points and process switching time points. If the time difference between the moment when the temperature deviation value exceeds the first interval and any planned time point is less than the preset fourth time window, then the temperature deviation is determined to be a planned fluctuation, and the generation of the temperature abnormality warning signal is suppressed. If the time difference between the moment when the temperature deviation value exceeds the first interval and all planned time points is greater than or equal to the fourth time window, then the temperature anomaly warning signal is generated normally.

10. A temperature monitoring system for use in preform injection molds, characterized in that, include: The electromagnetic interference detection and identification module is used to measure the electromagnetic interference intensity and frequency range of each piece of equipment in the injection molding workshop under different operating conditions using electromagnetic environment detection equipment, and to identify the interference source and its corresponding interference frequency band. A temperature data isolation acquisition module is used to acquire temperature data and transmit the temperature data to the data acquisition circuit after photoelectric isolation and magnetic isolation to obtain initial temperature data. The data filtering compensation and correction module is used to filter the initial temperature data based on the interference source and its corresponding interference frequency band, and to compensate and correct the filtered data based on the thermal inertia of the preform injection mold to obtain the target temperature data. The temperature anomaly detection and early warning module is used to compare the target temperature data with a preset standard temperature curve and calculate the temperature deviation value and the deviation change rate; if the temperature deviation value exceeds the first interval and the duration is greater than the preset first time window, a temperature anomaly early warning signal is generated, and the temperature anomaly trend type is determined according to the deviation change rate to obtain the target anomaly trend type. The monitoring report generation and upload module is used to generate a temperature monitoring report based on the temperature anomaly warning signal and the target anomaly trend type, and upload it to the centralized monitoring platform of the injection molding workshop.

Images