An enameled wire dielectric loss map AI intelligent analysis system and method

Abstract

The embodiment of the application provides a kind of enameled wire dielectric loss atlas AI intelligent analysis system and method, it is applied to insulating material and manufacturing process technical field, this method obtains dielectric loss data by monitoring polarization relaxation signal in the curing process of enameled wire, and oven temperature gradient, take-up tension are constructed coupling atlas by space-time correlation;By differential processing, extract the solidification front moving speed, solvent evaporation interface change and microcapsule distribution density;Based on the feature set, the correspondence between process parameters and curing state is established, and the key process parameter combination that causes dielectric loss anomaly is identified;Generate the repair strategy with adjusting oven outlet cooling fan as the core, trigger microcapsule self-repair by establishing local non-uniform temperature field;Repair effect is evaluated by using terahertz wave scanning, process mapping relationship is established, process parameter optimization direction is deduced reversely, enameled wire curing process monitoring, repair and process optimization are realized, product quality and production efficiency are improved.

Description

Technical Field

[0001] This application relates to the field of insulating materials and manufacturing processes, and in particular to an AI-powered intelligent analysis system and method for dielectric loss graphs of enameled wires. Background Technology

[0002] In the field of enameled wire manufacturing, the curing quality of the insulating varnish is a key factor that determines the product performance. Currently, the industry generally adopts a method of offline sampling inspection combined with manual experience adjustment to control the curing process.

[0003] Traditional testing methods cannot capture changes during the curing process in real time. They can only assess the degree of curing through destructive testing after production is completed, leading to a lag in defect detection. Although existing dielectric loss measurement technologies can reflect certain state characteristics of insulating materials, the measurement data is independent of production process parameters, such as oven temperature distribution and winding tension, forming data silos. This disconnect makes process adjustments lack precise data support and often relies on the experience and judgment of operators, resulting in blind and inaccurate parameter adjustments.

[0004] Traditional methods lack effective online repair mechanisms for coating defects that occur during the production process, usually requiring downtime for repairs, which seriously affects production efficiency and product qualification rate. With the continuous improvement of the performance requirements for enameled wire, this extensive production process control method can no longer meet the needs of high-quality production.

[0005] Therefore, there is an urgent industrial need to develop an intelligent analysis method that can detect, diagnose, and repair defects during the curing process of enameled wire. Summary of the Invention

[0006] This application provides an AI-powered intelligent analysis system and method for dielectric loss graphs of enameled wires, enabling precise control of the enameled wire curing process, self-repair of defects, and self-optimization of process parameters, thereby improving product quality, production efficiency, and product qualification rate. To achieve the above objectives, this application adopts the following technical solution: An AI-powered intelligent analysis method for dielectric loss patterns of enameled wires, the method comprising: When the enameled wire enters the oven curing stage, the polarization relaxation signal of the insulating varnish during the heating process is monitored by a dielectric sensor to obtain the original data of dielectric loss reflecting the degree of polymer crosslinking. The original data of dielectric loss is spatiotemporally correlated with the oven temperature gradient and the winding tension fluctuation to construct a coupling graph of dielectric loss and process parameters. Differential processing is performed on the coupling spectrum to extract characteristic parameters such as the moving rate of the curing front of the insulating varnish, the change of the solvent evaporation interface position, and the fluctuation of the microcapsule distribution density, so as to obtain the coating curing state feature set; Based on the coating curing state feature set, the correspondence between the oven hot air circulation speed, the vibration frequency of the coating mold and the curing front morphology is established, and the key process parameter combination that leads to abnormal media loss is identified. Based on the combination of key process parameters, a repair control strategy is generated with the adjustment of the oven outlet cooling fan as the core. By implementing the repair control strategy, and targeting the identified defect spatial coordinates, the leveling and curing rate of the insulating varnish are changed to release repair monomers from the capsule at the defect spatial coordinates, thereby achieving self-repair of the coating structure. Based on the defect spatial coordinates, the positions of the enameled wire segments that have completed self-repair are marked. When the marked enameled wire section is moved to the terahertz detection station, the marked enameled wire section is scanned with terahertz waves to generate evaluation indicators including repair uniformity and bonding integrity. The evaluation indicators are correlated with the original defect features in the feature set to establish a mapping relationship between defect type, repair process and repair effect; Based on the mapping relationship, the optimization direction of the oven temperature zone setting value and the painting pressure parameter is derived in reverse.

[0007] In some possible implementations, the step of spatiotemporally correlating the raw dielectric loss data with the oven temperature gradient and winding tension fluctuations to construct a coupled spectrum of dielectric loss and process parameters includes: Data from multiple temperature measurement points inside the oven are collected to form a temperature gradient distribution curve; The readings of the take-up tension sensor are recorded synchronously to form tension fluctuation time-series data; A unified timestamp is applied to the raw data of dielectric loss, temperature gradient curves, and tension fluctuation data to generate a multi-source data set; Establish a two-dimensional coordinate field with time as the horizontal axis and the position of the enameled wire as the vertical axis; The multi-source data set is mapped onto the two-dimensional coordinate field to generate a data alignment matrix; The data alignment matrix is ​​fused to generate a coupling graph of dielectric loss and process parameters.

[0008] In some possible implementations, the differential processing of the coupling spectrum extracts characteristic parameters such as the movement rate of the insulating varnish curing front, the change in the solvent evaporation interface position, and the fluctuation in the microcapsule distribution density, to obtain a coating curing state feature set, including: Perform a first-order differential operation on the coupling spectrum along the time axis to generate a time-axis differential spectrum; Identify the extreme points of the rate of change of dielectric loss from the time axis differential graph to form a set of extreme points; Connect the points in the set of extreme points to form a solidified front movement path; Calculate the slope of the curing front movement path to obtain the curing front movement rate of the insulating varnish; Perform first-order differential operations on the coupling spectrum along the spatial axis to generate a spatial axis differential spectrum; Locate the interface where the differential value changes abruptly in the spatial axis differential map to determine the initial position of the solvent evaporation interface; Track changes in the solvent evaporation interface position to form a sequence of solvent evaporation interface position changes; The high-frequency fluctuation component caused by the microcapsules was separated from the coupled spectrum signal; The amplitude distribution of the high-frequency fluctuation components is statistically analyzed to obtain the microcapsule distribution density fluctuation parameter. The curing front movement rate, solvent evaporation interface position change sequence, and microcapsule distribution density fluctuation parameters are combined to form a coating curing state characteristic set.

[0009] In some possible implementations, the relationship between the oven hot air circulation speed, the vibration frequency of the coating mold, and the morphology of the curing front is established based on the coating curing state feature set, and the key process parameter combinations that lead to abnormal media loss are identified, including: Call the coating curing state feature set; Based on the feature set data, a correspondence is established between the hot air circulation speed of the oven and the moving rate of the curing front, thus obtaining the first correspondence. Based on the feature set data, a correspondence between the vibration frequency of the coating mold and the change sequence of the solvent evaporation interface position is established to obtain a second correspondence. The measured values ​​in the feature set data are compared with the normal ranges in the first and second correspondences to obtain the comparison results. Based on the comparison results, abnormal features that deviate from the normal range are identified, and an abnormal feature set is obtained; Based on the abnormal feature set, the hot air circulation speed and vibration frequency that cause the abnormality are reversed from the first correspondence and the second correspondence to obtain the combination of key process parameters. The combination of key process parameters is marked as the combination of key process parameters that leads to abnormal dielectric loss.

[0010] In some possible implementations, the generation of a repair control strategy based on the combination of key process parameters, with the core being the adjustment of the oven outlet cooling fan, includes: Receive the combination of key process parameters and the spatial coordinates of the defect to obtain the target defect information; Based on the defect type in the target defect information, a preset control strategy library is queried to obtain a matching control strategy; Based on the matching control strategy, a repair control strategy with the core command of adjusting the speed of the cooling fan at the oven outlet section is generated, and an initial repair strategy is obtained. The initial repair strategy is parameterized by setting specific target values ​​for the fan speed and the duration of action to obtain a parameterized repair strategy. The parameterized repair strategy is bound to the defect space coordinates in the target defect information to obtain the bound repair control strategy.

[0011] In some possible implementations, the execution of the repair control strategy, based on the identified defect spatial coordinates, induces the microcapsules to release repair monomers at the defect spatial coordinates by changing the leveling and curing rate of the insulating varnish, thereby achieving self-repair of the coating structure, including: According to the repair control strategy, the cooling fan at the oven outlet is adjusted to the target speed, and a target cooling environment is established at the identified defect spatial coordinates. Based on the target cooling environment, a local non-uniform temperature field is formed on the surface of the enameled wire at the defect spatial coordinates. The target thermal stress induced at the defect by the local non-uniform temperature field causes the microcapsule wall material at the defect to rupture. The repair monomers inside the ruptured microcapsules are released to the coating defects, where they undergo a polymerization reaction to complete the self-repair of the coating structure.

[0012] In some possible implementations, when the marked enameled wire segment reaches the terahertz detection station, the marked enameled wire segment is scanned using terahertz waves to generate evaluation indicators including repair uniformity and bonding integrity, including: When the enameled wire section carrying the location marker arrives at the terahertz detection station, the terahertz scanning command is triggered. According to the terahertz scanning command, the marked segment is scanned using a terahertz wave to obtain the original terahertz signal; The original terahertz signal that penetrates the coating is received and processed to generate a terahertz time-domain spectrum; Based on the terahertz time-domain spectrum, the dielectric constant of multiple micropoints in the repair region is measured to obtain a dielectric constant dataset. The standard deviation of the dielectric constant dataset is calculated to obtain the repair uniformity index; Based on the terahertz time-domain spectrum, the reflectivity of the terahertz wave is measured at the repair interface to obtain interface reflectivity data. The degree of agreement between the interface reflectivity data and the theoretical complete interface reflectivity is calculated to obtain the integrity index of the interface. By integrating the repair uniformity index and the bonding integrity index, an evaluation index for repair effect is output.

[0013] In some possible implementations, the optimization direction for deriving the oven temperature zone setpoint and coating pressure parameters based on the mapping relationship includes: The evaluation data is obtained by calling the aforementioned repair effect evaluation indicators; The original defect features in the coating curing state feature set are called to obtain defect feature data; The evaluation data, the defect feature data, and the repair process parameters used in the repair are correlated to generate a triplet dataset; Analyze the triplet dataset, establish the mapping relationship between defect type, repair process and repair effect, and obtain the repair effect mapping relationship table; Based on the repair effect mapping table, historical records in which the evaluation data reached the preset quality standard were selected to obtain a set of compliant repair records. Extract the corresponding oven temperature zone setting and painting pressure parameters from the set of qualified repair records to obtain a set of reference process parameters; Analyze the statistical distribution of the reference process parameters for the oven temperature zone settings to determine the optimal parameter range for the oven temperature zone settings; Analyze the statistical distribution of the coating pressure parameters in the reference process parameter set to determine the optimal parameter range for the coating pressure parameters.

[0014] In some possible implementations, the method further includes: Based on the optimized parameter range of the oven temperature zone setting value and the optimized parameter range of the coating pressure parameter, optimized process parameters are generated. The optimized process parameters are written into the production line control parameters to complete the parameter update; the updated process parameters are executed to continue producing new enameled wires under the new process conditions. Collect the dielectric loss data of the new enameled wire to form a new round of dielectric loss data; The new round of dielectric loss data is correlated with the optimized process parameters to generate verification results; Based on the verification results, the optimized process parameters are fine-tuned to form an adaptive control closed loop.

[0015] An AI-powered intelligent analysis system for dielectric loss patterns of enameled wires, the system comprising: The dielectric loss monitoring module is configured to monitor the polarization relaxation signal of the insulating varnish during the heating process by a dielectric sensor when the enameled wire enters the oven curing stage, and obtain the raw data of dielectric loss reflecting the degree of polymer crosslinking. The coupling graph construction module is used to spatiotemporally correlate the raw data of dielectric loss with the oven temperature gradient and the winding tension fluctuation to construct a coupling graph of dielectric loss and process parameters. The feature extraction module is used to perform differential processing on the coupling spectrum to extract feature parameters such as the moving rate of the curing front of the insulating varnish, the change of the solvent evaporation interface position, and the fluctuation of the microcapsule distribution density, so as to obtain the feature set of the coating curing state. The key parameter identification module is used to establish the correspondence between the oven hot air circulation speed, the vibration frequency of the coating mold and the curing front morphology based on the coating curing state feature set, and to identify the key process parameter combinations that cause abnormal media loss. The repair strategy generation module is used to generate a repair control strategy based on the combination of key process parameters, with the core being the adjustment of the oven outlet cooling fan. The self-healing execution module is used to execute the repair control strategy. Based on the identified defect spatial coordinates, it changes the leveling and curing rate of the insulating varnish to induce the microcapsules to release repair monomers at the defect spatial coordinates, thereby achieving self-healing of the coating structure. The location marking module is used to mark the location of the self-healing enameled wire segment based on the defect spatial coordinates. The repair evaluation module is used to scan the marked enameled wire section with terahertz waves when the marked enameled wire section is run to the terahertz detection station, and generate evaluation indicators including repair uniformity and bonding integrity. The mapping relationship establishment module is used to perform correlation analysis between the evaluation index and the original defect features in the feature set, and establish a mapping relationship between defect type, repair process and repair effect; The process optimization module is used to deduce the optimization direction of the oven temperature zone setting value and the coating pressure parameter based on the mapping relationship.

[0016] As can be seen from the above technical solution, this application has the following beneficial effects: 1. This method constructs an intelligent closed-loop system integrating monitoring, diagnosis, repair, and optimization, improving the quality control level of enameled wire production. It collects dielectric loss data through dielectric sensors and integrates it with oven temperature and take-up tension in a spatiotemporal manner to generate a coupled spectrum, breaking down traditional data silos and achieving global visual monitoring of the impact of multiple parameters on the curing process. The system can accurately locate the root cause of defects and establish a local non-uniform temperature field at specific defects by adjusting the cooling fan at the oven outlet. It utilizes thermal stress to trigger a microcapsule self-healing mechanism, achieving online repair without stopping the machine. Terahertz detection is used to quantitatively evaluate the repair effect and reverse-optimize process parameters, forming a complete intelligent control loop that improves production efficiency and product consistency.

[0017] 2. This method, through spatiotemporal correlation and differential processing, accurately extracts feature parameters such as the curing front movement rate and the solvent evaporation interface position from the coupled spectrum, realizing multi-dimensional quantitative characterization of complex curing processes. It solves the problem of incomplete feature extraction in traditional methods, establishes the correspondence between process parameters and curing features, and can accurately identify key process parameter combinations that cause anomalies. This allows process adjustment to shift from relying on experience to making precise decisions based on data. It organically combines process control with material self-healing technology, precisely triggering point repair by controlling the local temperature field, and using terahertz non-destructive testing to provide quantitative feedback, forming a complete technical closed loop of continuous self-optimization, thus improving the accuracy and adaptability of process control. Attached Figure Description

[0018] The invention will now be further described with reference to the accompanying drawings.

[0019] Figure 1 A first flowchart provided for an embodiment of this application; Figure 2 A second flowchart provided for embodiments of this application; Figure 3 A third flowchart provided for embodiments of this application; Figure 4 The fourth flowchart provided for the embodiments of this application; Figure 5 The fifth flowchart provided for the embodiments of this application; Figure 6 The sixth flowchart provided for the embodiments of this application. Detailed Implementation

[0020] The terms "first," "second," and "third," etc., used in this application specification, claims, and drawings are for distinguishing different objects, not for specifying a particular order.

[0021] In the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0022] Research has revealed that traditional methods typically rely on offline testing and manual adjustment of process parameters based on experience, making it difficult to monitor the curing state in real time and repair defects promptly. While existing technologies can reflect the state of insulating materials through dielectric loss measurement, they lack a dynamic correlation with process parameters, leading to inaccurate defect identification and poor repair results.

[0023] To address the aforementioned issues, this application provides an AI-powered intelligent analysis system and method for enameled wire dielectric loss graphs: Example 1 To solve the above problems, such as Figure 1 As shown, the curing quality of the insulating varnish directly affects the electrical and mechanical properties of enameled wire during its production. This application utilizes AI-powered intelligent analysis of dielectric loss patterns to achieve real-time monitoring of the curing process, defect self-repair, and process optimization, thereby improving the quality and production efficiency of enameled wire.

[0024] This embodiment provides an AI-powered intelligent analysis method for dielectric loss patterns of enameled wires, with the following specific steps: During the oven curing stage of the enameled wire, the polarization relaxation signal of the insulating varnish during the heating process is monitored by a dielectric sensor to obtain raw data on dielectric loss reflecting the degree of polymer crosslinking. On enameled wire production lines, after the enameled wire enters the drying oven, dielectric sensors, such as capacitive sensors or impedance analyzers, are used to monitor the dielectric properties of the insulating varnish in real time. Polarization relaxation signal refers to the change in electrical signal generated during the cross-linking process of the insulating varnish due to the orientation and relaxation of polar groups in the polymer molecular chains under the influence of an electric field. The raw data for dielectric loss includes the dielectric constant and loss tangent, which change with increasing temperature and directly reflect the degree of polymer cross-linking. For example, when the degree of cross-linking increases, molecular chain movement is restricted, the polarization relaxation time shortens, leading to a decrease in the loss tangent.

[0025] Beneficial effects: By monitoring polarization relaxation signals in real time, this application can capture the dynamic changes of insulating varnish during the curing process, solving the problem that traditional methods cannot reflect the crosslinking state in real time. The principle is that polymer crosslinking leads to a decrease in the mobility of molecular chain segments, thereby changing the polarization relaxation behavior; by analyzing dielectric loss data, the degree of crosslinking can be indirectly quantified, providing basic data for subsequent spectrum construction. This real-time monitoring allows for the immediate detection of curing anomalies during production, rather than waiting until offline detection to discover the problem.

[0026] By spatiotemporally correlating the raw data of dielectric loss with the oven temperature gradient and the fluctuation of winding tension, a coupling graph of dielectric loss and process parameters is constructed: The oven temperature gradient is collected through multiple temperature measurement points, such as thermocouples, to form a temperature distribution curve. Wire tension fluctuations are recorded by tension sensors, such as strain gauges, to form tension time-series data. Spatiotemporal correlation refers to combining the time-series data with spatial location, specifically the movement path of the enameled wire within the oven. All data is synchronized using a unified timestamp and mapped onto a two-dimensional coordinate field with time as the horizontal axis and the enameled wire's position as the vertical axis, generating a data alignment matrix. Then, through data fusion techniques, such as weighted averaging or principal component analysis, a coupled spectrum is generated. This spectrum visually displays the relationship between medium loss, temperature, and tension.

[0027] Beneficial effects: By using spatiotemporal correlation, this application dynamically links dielectric loss with process parameters, solving the data silo problem in traditional methods. The principle is that the curing process is affected by the combined influence of temperature and tension; uneven temperature gradients may lead to localized insufficient curing, and tension fluctuations may cause changes in film thickness, all of which are reflected in dielectric loss. Coupling plots can visually display these effects, thereby helping to identify abnormal patterns; for example, if the temperature in a certain area is too low, the coupling plot will show that the dielectric loss value in that area is abnormally high, indicating insufficient curing.

[0028] By differentiating the coupling spectrum, characteristic parameters such as the moving rate of the curing front of the insulating varnish, the change in the position of the solvent evaporation interface, and the fluctuation of the microcapsule distribution density are extracted to obtain the feature set of the coating curing state: The coupling spectrum is subjected to first-order differentiation along the time axis to calculate the rate of change of dielectric loss and identify extreme points, such as local maxima or minima, which correspond to the positions of the curing front. Connecting these extreme points forms the path of the curing front movement, and the slope of this path yields the curing front movement rate. Similarly, differentiation is performed along the spatial axis to locate abrupt changes in differential values, determine the initial position of the solvent evaporation interface, and track its change sequence. Furthermore, high-frequency fluctuation components are separated from the coupling spectrum signal using a bandpass filter. These fluctuations are caused by uneven microcapsule distribution or rupture, and the amplitude distribution is statistically analyzed to obtain the microcapsule distribution density fluctuation parameter. The set of all characteristic parameters forms the coating curing state feature set.

[0029] Beneficial Effects: By extracting key features through differential processing, this application can accurately characterize the curing kinetics process, solving the problem of incomplete feature extraction in traditional methods. The principle is as follows: the curing front movement rate reflects the curing reaction rate, changes in the solvent evaporation interface affect the quality of film formation, and fluctuations in microcapsule distribution are related to repair potential; these features collectively provide a multi-dimensional view of the curing state, providing a basis for anomaly diagnosis. For example, an excessively slow curing front movement rate may indicate insufficient temperature, while uneven microcapsule distribution may lead to repair failure.

[0030] Based on the aforementioned coating curing state feature set, a correspondence was established between the oven hot air circulation speed, the vibration frequency of the coating mold, and the morphology of the curing front, identifying key process parameter combinations that lead to abnormal media loss: Using feature set data, a correspondence between hot air circulation speed and curing front movement rate (first correspondence) and a correspondence between coating mold vibration frequency and solvent evaporation interface position change sequence are established through regression analysis or machine learning models (second correspondence). Then, the measured values ​​are compared with the normal range to identify abnormal features, such as curing rate deviation from the standard value, and the hot air circulation speed and vibration frequency that cause the abnormality are traced back to form a combination of key process parameters.

[0031] Beneficial Effects: This application establishes a correspondence between process parameters and curing characteristics, achieving precise location of defect root causes and solving the problem of blind parameter adjustment in traditional methods. The principle is as follows: the hot air circulation speed affects the temperature uniformity within the oven, thereby altering the curing front movement rate; the vibration frequency of the coating mold affects the leveling properties of the paint, thus changing the solvent evaporation interface position. By identifying key parameter combinations, the process can be adjusted specifically, avoiding the resource waste caused by overall adjustments.

[0032] Based on the combination of key process parameters, a repair control strategy is generated, with the adjustment of the oven outlet cooling fan as the core: Based on the combination of key process parameters and the spatial coordinates of defects, a preset control strategy library is queried. For example, the database stores control strategies corresponding to different defect types. A repair control strategy is generated with the core instruction of adjusting the speed of the cooling fan at the oven outlet. The strategy parameterization includes setting the target value of the fan speed and the duration of action, and binding it with the defect coordinates.

[0033] Beneficial effects: This application achieves rapid response repair of local defects by specifically adjusting the cooling fan, solving the problem of low efficiency in traditional overall adjustment. The principle is that the cooling fan speed changes the cooling rate of the enameled wire surface, affecting the leveling and curing rate of the insulating varnish; creating a local cooling environment at the defect site can induce thermal stress, prompting the microcapsules to rupture and release repair monomers.

[0034] By implementing the aforementioned repair control strategy, and targeting the identified defect spatial coordinates, the leveling and curing rate of the insulating varnish are altered to induce the microcapsules to release repair monomers at the defect spatial coordinates, thereby achieving self-repair of the coating structure. Adjust the cooling fan at the oven outlet to the target speed to establish a localized non-uniform temperature field at the defect coordinates. The non-uniform temperature field generates thermal stress, causing the microcapsule wall material at the defect to rupture and release repair monomers, such as epoxy resin monomers. The repair monomers undergo a polymerization reaction at the defect, filling cracks or pores and completing self-repair.

[0035] Beneficial Effects: This application utilizes microencapsulation technology to achieve self-healing of the coating structure, overcoming the drawback of traditional methods that require downtime for repairs. The principle lies in the fact that the microcapsule wall material is designed to rupture under specific thermal stress, releasing monomers that repair defects through polymerization, restoring the coating's insulation properties. This self-healing process requires no manual intervention, improving production efficiency and product quality.

[0036] Based on the aforementioned defect spatial coordinates, the positions of the self-healed enameled wire segments are marked: Use marking systems, such as RFID tags or inkjet printers, to mark the sections corresponding to the defect coordinates on the enameled wire for subsequent inspection.

[0037] Beneficial effects: Location marking ensures that subsequent inspections can target the repair sections, avoiding the waste of resources in full-line inspections and improving inspection efficiency.

[0038] When the marked enameled wire segment reaches the terahertz testing station, the marked enameled wire segment is scanned using terahertz waves to generate evaluation indicators including repair uniformity and bonding integrity: When the marked section reaches the terahertz detection station, a terahertz scan is triggered. The terahertz wave penetrates the coating, acquires the original signal, processes it to generate a terahertz time-domain spectrum, measures the dielectric constant of multiple micropoints within the repair area, and calculates the standard deviation to obtain the repair uniformity index; the terahertz wave reflectivity is measured at the repair interface and compared with the reflectivity of the theoretical intact interface to obtain the bonding integrity index.

[0039] Beneficial effects: By using terahertz nondestructive testing, this application can quantify the repair effect, solving the problem of inaccuracy in traditional visual inspection. The principle is that terahertz waves are sensitive to dielectric materials, the uniformity of the dielectric constant of the repaired area reflects the repair quality, and the interface reflectivity reflects the bonding strength; these indicators objectively evaluate the repair effect and provide data support for process optimization.

[0040] The evaluation indicators are correlated with the original defect features in the feature set to establish a mapping relationship between defect type, repair process, and repair effect: The evaluation metrics, original defect characteristics, and repair process parameters are correlated to generate a triplet dataset. Data mining techniques, such as cluster analysis or decision trees, are used to build a mapping table that displays the repair effects corresponding to different defect types and repair processes.

[0041] Beneficial effects: By establishing mapping relationships through correlation analysis, the causal relationships of defect repair are revealed, solving the problem of traditional methods relying on experience for repair process selection. The principle is that mapping relationships help understand which repair process is most effective for a specific defect type, thereby optimizing the repair strategy.

[0042] Based on the mapping relationship, the optimization direction for the oven temperature zone setpoint and coating pressure parameters is derived in reverse: Filter records that meet the repair effect standards from the mapping table, extract the corresponding oven temperature zone settings and painting pressure parameters, analyze the statistical distribution, such as histograms or confidence intervals, and determine the optimal parameter range.

[0043] Beneficial effects: By optimizing process parameters through data-driven approaches, continuous improvement of the production process is achieved, overcoming the lag inherent in traditional experience-based adjustments. The principle lies in: extracting the optimal range of process parameters from successful repair cases through reverse engineering, guiding parameter settings, and reducing the occurrence of defects.

[0044] Example 2 like Figures 2-6 As shown, the specific process of constructing the coupling graph of dielectric loss and process parameters is as follows: Data from multiple temperature measurement points inside the oven is collected to form a temperature gradient distribution curve: Multiple thermocouples are arranged along the length of the oven to collect temperature data and form a curve to show the temperature changes at different locations.

[0045] Synchronously record the readings of the tension sensor during take-up to form tension fluctuation time series data: use the tension sensor to monitor the tension changes during take-up and record them as time series data.

[0046] To ensure time consistency, a unified timestamp is applied to the raw data of dielectric loss, temperature gradient curves, and tension fluctuation data: all data sources are synchronized using a high-precision clock.

[0047] A two-dimensional coordinate field is established with time as the horizontal axis and the position of the enameled wire as the vertical axis: the position of the enameled wire is obtained through an encoder, reflecting its real-time position in the oven.

[0048] The multi-source data set is mapped onto a two-dimensional coordinate field to generate a data alignment matrix: each matrix element corresponds to the medium loss value, temperature value, and tension value at a specific time and location.

[0049] The data alignment matrix is ​​fused to generate a coupled spectrum of dielectric loss and process parameters: Data fusion algorithms, such as Kalman filtering or neural networks, are used to integrate the data and generate a comprehensive spectrum.

[0050] Beneficial effects: Through multi-source data alignment and fusion, coupled graphs can accurately reflect the spatiotemporal impact of process parameters on dielectric loss, solving the problem of incompleteness in traditional single-point measurements. Specifically, temperature gradients affect curing uniformity, and tension fluctuations may cause paint film deformation. Coupled graphs can identify these interactions, thereby improving the accuracy of defect detection. For example, if tension fluctuations are frequent, the coupled graph will show that the dielectric loss value changes drastically with location, indicating uneven paint film thickness, requiring adjustment of the painting process.

[0051] The specific process of extracting the feature set of the coating curing state: Perform first-order differential operations on the coupling spectrum along the time axis to generate a time-axis differential spectrum: use numerical differential methods, such as the central difference method, to calculate the rate of change of dielectric loss at each time point.

[0052] Identify the extreme points of the rate of change of dielectric loss from the time axis differential plot to form a set of extreme points: extreme points correspond to key events of the curing reaction, such as the gel point or the curing start point.

[0053] Connect the points in the set of extreme points to form a solidified frontier path: use linear interpolation or curve fitting to connect the extreme points.

[0054] Calculate the slope of the curing front movement path to obtain the curing front movement rate of the insulating varnish: the slope represents the movement speed of the curing front.

[0055] Perform first-order differential operations on the coupling spectrum along the spatial axis to generate a spatial axis differential spectrum: perform differential operations along the length of the enameled wire to calculate the rate of change of dielectric loss at each location.

[0056] Locate the interface where the differential value changes abruptly in the spatial axis differential map to determine the initial position of the solvent evaporation interface: the abrupt change point represents the solvent evaporation interface, where the medium loss value changes significantly.

[0057] Track changes in the solvent evaporation interface position to form a sequence of solvent evaporation interface position changes: use tracking algorithms, such as particle filtering, to follow the interface movement.

[0058] Separating high-frequency fluctuation components caused by microcapsules from the coupled spectrum signal: using a bandpass filter to separate the signal in a specific frequency band that corresponds to the microcapsule activity.

[0059] The amplitude distribution of high-frequency fluctuation components is statistically analyzed to obtain the distribution density fluctuation parameter of microcapsules: the variance or entropy of the amplitude is calculated to represent the uniformity of distribution.

[0060] The curing front movement rate, solvent evaporation interface position change sequence, and microcapsule distribution density fluctuation parameters are combined to form a coating curing state feature set: the feature set is stored as a vector for subsequent analysis.

[0061] Beneficial effects: Through differential processing and high-frequency separation, the dynamic characteristics of the curing process can be accurately extracted, solving the problem that traditional methods struggle to capture subtle changes. Specifically, the curing front movement rate helps assess curing efficiency, changes in the solvent evaporation interface reflect the film drying process, and fluctuations in microcapsule distribution indicate repair capability.

[0062] The specific process of identifying key process parameter combinations: Retrieve coating curing state feature set: Read feature set data from database or memory.

[0063] Based on the feature set data, a correspondence between the oven hot air circulation speed and the curing front movement rate is established, resulting in the first correspondence: a regression model is used to fit the relationship between the hot air circulation speed and the curing rate. For example, an increase in hot air speed may lead to an increase in curing rate.

[0064] Based on the feature set data, a correspondence between the vibration frequency of the coating mold and the change sequence of the solvent evaporation interface position is established, resulting in a second correspondence: time series analysis is used to establish the relationship between vibration frequency and interface position change. For example, excessively high vibration frequency may lead to interface position instability.

[0065] The measured values ​​in the feature set data are compared with the normal ranges in the first and second correspondences to obtain the comparison results: the normal ranges are set based on historical data or standard values, and the comparison results show which features deviate from the normal range.

[0066] Based on the comparison results, abnormal features that deviate from the normal range are identified, resulting in an abnormal feature set: for example, the curing rate is lower than the threshold or the interface position changes too much.

[0067] Based on the abnormal feature set, the hot air circulation speed and vibration frequency that caused the abnormality were reverse-searched from the first and second correspondences to obtain the key process parameter combination: for example, if the curing rate is too slow, the reverse search found that the hot air circulation speed is too low.

[0068] The key process parameter combinations are marked as those that cause abnormal media loss: these markings are used for subsequent repair strategies.

[0069] Beneficial effects: By establishing correspondences and reverse lookup, the process parameters causing defects can be accurately located, solving the problem of trial and error in parameter adjustment in traditional methods. Specifically, the hot air circulation speed affects the temperature distribution, thus affecting the curing rate; the vibration frequency of the coating mold affects the uniformity of the paint film, thus affecting solvent evaporation; by identifying key parameter combinations, these parameters can be quickly adjusted to avoid production interruptions.

[0070] The specific process of generating a repair control strategy: Receive key process parameter combinations and defect spatial coordinates to obtain target defect information: obtain key parameters and defect locations from upstream steps.

[0071] Based on the defect type in the target defect information, the preset control strategy library is queried to obtain the matching control strategy: The control strategy library stores different defect types, such as insufficient curing and uneven paint film, and the corresponding control strategies. For example, for insufficient curing, the strategy may include increasing the cooling fan speed.

[0072] Based on the matching control strategy, a repair control strategy is generated with the core instruction of adjusting the speed of the cooling fan at the oven outlet section, resulting in an initial repair strategy: the initial strategy specifies the fan to be adjusted and the approximate speed range.

[0073] The initial repair strategy is parameterized by setting specific target values ​​for the fan speed and the duration of action, resulting in a parameterized repair strategy.

[0074] The parameterized repair strategy is bound to the defect spatial coordinates in the target defect information to obtain the bound repair control strategy: ensuring that the strategy is only applied to the segment corresponding to the defect coordinates.

[0075] Beneficial effects: By querying the strategy library and parameterized binding, precise repair strategies for specific defects can be generated, solving the problem of traditional one-size-fits-all adjustments. Specifically, adjusting the cooling fan can locally change the curing environment, prompting the microcapsules to release repair monomers at the defect site, thereby achieving efficient self-repair.

[0076] The specific process of performing self-repair: According to the repair control strategy, the cooling fan at the oven outlet is adjusted to the target speed, and a target cooling environment is established at the identified defect spatial coordinates: the control system adjusts the fan speed to create a cooling zone at the defect.

[0077] Based on the target cooling environment, a local non-uniform temperature field is formed on the surface of the enameled wire at the defect spatial coordinates: cooling causes uneven temperature distribution on the surface of the enameled wire, generating thermal stress.

[0078] The target thermal stress induced at the defect by the local non-uniform temperature field causes the microcapsule wall material at the defect to rupture: the thermal stress exceeds the strength threshold of the microcapsule wall material, leading to rupture.

[0079] The repair monomers inside the ruptured microcapsules are released to the coating defects, where they polymerize to complete the self-repair of the coating structure: the repair monomers, such as dicyclopentadiene, polymerize upon contact with the catalyst to fill the defects.

[0080] Beneficial effects: By inducing thermal stress through localized cooling, microcapsule rupture and repair monomer release can be triggered, solving the problem that traditional self-healing technologies require external stimulation.

[0081] The specific process of generating evaluation indicators: When the enameled wire section carrying the location marker arrives at the terahertz detection station, the terahertz scanning command is triggered: the terahertz scan is automatically triggered by the sensor detecting the marked section.

[0082] According to the terahertz scanning command, the marked section is scanned with terahertz waves to obtain the original terahertz signal: the terahertz transmitter emits a beam, and the receiver receives the signal that penetrates the coating.

[0083] The raw terahertz signal penetrating the coating is received and processed to generate a terahertz time-domain spectrum: signal processing algorithms, such as Fourier transform, are used to convert the time-domain signal into a frequency-domain spectrum.

[0084] Based on the terahertz time-domain spectrum, the dielectric constant of multiple micropoints within the repair region is measured to obtain a dielectric constant dataset: the dielectric constant of each micropoint is calculated to form the dataset.

[0085] Calculate the standard deviation of the dielectric constant dataset to obtain the repair uniformity index: a smaller standard deviation indicates uniform repair.

[0086] Based on the terahertz time-domain spectrum, the reflectivity of terahertz waves is measured at the repair interface to obtain interface reflectivity data: reflectivity reflects the bonding condition of the interface.

[0087] The degree of agreement between the calculated interface reflectivity data and the theoretical complete interface reflectivity is used to obtain the bonding integrity index: a high degree of agreement indicates good bonding.

[0088] By integrating the uniformity index and the integrity index of the bond, an evaluation index for the repair effect is output: the evaluation index is used for subsequent analysis.

[0089] Beneficial effects: Terahertz scanning and index calculation enable objective evaluation of repair quality, solving the problem of inaccurate traditional subjective assessments. Specifically, repair uniformity reflects the distribution of repair materials, while integrity reflects the bonding strength between the repair layer and the substrate; these indicators provide a quantitative basis for process optimization.

[0090] The specific process of reverse derivation of the optimization direction: Call the repair effect evaluation indicators to obtain evaluation data: obtain the evaluation indicator values ​​from the detection system.

[0091] The original defect features in the coating curing state feature set are called to obtain defect feature data: the defect type and feature value are obtained from the feature set.

[0092] The evaluation data, defect feature data, and repair process parameters used in the repair are correlated to generate a triplet dataset: each record includes defect features, repair process, and repair effect.

[0093] Analyze the triplet dataset to establish a mapping relationship between defect type, repair process and repair effect, and obtain a repair effect mapping relationship table: use data mining techniques, such as association rule learning, to discover patterns.

[0094] Based on the repair effect mapping table, historical records in which the evaluation data reached the preset quality standards were selected to obtain a set of compliant repair records.

[0095] From the set of compliant repair records, extract the corresponding oven temperature zone settings and painting pressure parameters to obtain a set of reference process parameters: collect these parameter values.

[0096] Analyze the statistical distribution of the oven temperature zone setpoints of the reference process parameters to determine the optimal parameter range of the oven temperature zone setpoints: calculate the mean ± standard deviation or use a box plot to determine the range.

[0097] Analyze the statistical distribution of the coating pressure parameters in the reference process parameter set to determine the optimal parameter range for the coating pressure parameters; the same method is used to determine the pressure optimization range.

[0098] Beneficial effects: Through correlation analysis and statistical distribution, optimized process parameters can be derived from successful cases, solving the problem of traditional optimization relying on trial and error. Specifically, the oven temperature zone setting affects the curing process, and the painting pressure affects the paint film thickness; by determining the optimization range, parameters can be set to reduce defects and improve product quality consistency.

[0099] The foregoing has shown and described the basic principles, main features, and advantages of this application. Those skilled in the art should understand that this application is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of this application. Various changes and modifications can be made to this application without departing from the spirit and scope thereof, and all such changes and modifications fall within the scope of this application as claimed. The scope of protection of this application is defined by the appended claims and their equivalents.

Claims

1. An AI-powered intelligent analysis method for dielectric loss patterns of enameled wires, characterized in that, The method includes: When the enameled wire enters the oven curing stage, the polarization relaxation signal of the insulating varnish during the heating process is monitored by a dielectric sensor to obtain the original data of dielectric loss reflecting the degree of polymer crosslinking. The original data of dielectric loss is spatiotemporally correlated with the oven temperature gradient and the winding tension fluctuation to construct a coupling graph of dielectric loss and process parameters. Differential processing is performed on the coupling spectrum to extract characteristic parameters such as the moving rate of the curing front of the insulating varnish, the change of the solvent evaporation interface position, and the fluctuation of the microcapsule distribution density, so as to obtain the coating curing state feature set; Based on the coating curing state feature set, the correspondence between the oven hot air circulation speed, the vibration frequency of the coating mold and the curing front morphology is established, and the key process parameter combination that leads to abnormal media loss is identified. Based on the combination of key process parameters, a repair control strategy is generated with the adjustment of the oven outlet cooling fan as the core. By implementing the repair control strategy, and targeting the identified defect spatial coordinates, the leveling and curing rate of the insulating varnish are changed to release repair monomers from the capsule at the defect spatial coordinates, thereby achieving self-repair of the coating structure. Based on the defect spatial coordinates, the positions of the enameled wire segments that have completed self-repair are marked. When the marked enameled wire section is moved to the terahertz detection station, the marked enameled wire section is scanned with terahertz waves to generate evaluation indicators including repair uniformity and bonding integrity. The evaluation indicators are correlated with the original defect features in the feature set to establish a mapping relationship between defect type, repair process and repair effect; Based on the mapping relationship, the optimization direction of the oven temperature zone setting value and the painting pressure parameter is derived in reverse.

2. The method according to claim 1, characterized in that, The step of spatiotemporally correlating the raw data of dielectric loss with the oven temperature gradient and the winding tension fluctuation to construct a coupled spectrum of dielectric loss and process parameters includes: Data from multiple temperature measurement points inside the oven are collected to form a temperature gradient distribution curve; The readings of the take-up tension sensor are recorded synchronously to form tension fluctuation time-series data; A unified timestamp is applied to the raw data of dielectric loss, temperature gradient curves, and tension fluctuation data to generate a multi-source data set; Establish a two-dimensional coordinate field with time as the horizontal axis and the position of the enameled wire as the vertical axis; The multi-source data set is mapped onto the two-dimensional coordinate field to generate a data alignment matrix; The data alignment matrix is ​​fused to generate a coupling graph of dielectric loss and process parameters.

3. The method according to claim 1, characterized in that, The coupling spectrum is differentiated to extract characteristic parameters such as the movement rate of the curing front of the insulating varnish, the change in the position of the solvent evaporation interface, and the fluctuation of the microcapsule distribution density, thus obtaining a feature set of the coating curing state, including: Perform a first-order differential operation on the coupling spectrum along the time axis to generate a time-axis differential spectrum; Identify the extreme points of the rate of change of dielectric loss from the time axis differential graph to form a set of extreme points; Connect the points in the set of extreme points to form a solidified front movement path; Calculate the slope of the curing front movement path to obtain the curing front movement rate of the insulating varnish; Perform first-order differential operations on the coupling spectrum along the spatial axis to generate a spatial axis differential spectrum; Locate the interface where the differential value changes abruptly in the spatial axis differential map to determine the initial position of the solvent evaporation interface; Track changes in the solvent evaporation interface position to form a sequence of solvent evaporation interface position changes; The high-frequency fluctuation component caused by the microcapsules was separated from the coupled spectrum signal; The amplitude distribution of the high-frequency fluctuation components is statistically analyzed to obtain the microcapsule distribution density fluctuation parameter. The curing front movement rate, solvent evaporation interface position change sequence, and microcapsule distribution density fluctuation parameters are combined to form a coating curing state characteristic set.

4. The method according to claim 2, characterized in that, Based on the coating curing state feature set, a correspondence is established between the oven hot air circulation speed, the vibration frequency of the coating mold, and the curing front morphology. This identifies key process parameter combinations that lead to abnormal media loss, including: Call the coating curing state feature set; Based on the feature set data, a correspondence is established between the hot air circulation speed of the oven and the moving rate of the curing front, thus obtaining the first correspondence. Based on the feature set data, a correspondence between the vibration frequency of the coating mold and the change sequence of the solvent evaporation interface position is established to obtain a second correspondence. The measured values ​​in the feature set data are compared with the normal ranges in the first and second correspondences to obtain the comparison results. Based on the comparison results, abnormal features that deviate from the normal range are identified, and an abnormal feature set is obtained; Based on the abnormal feature set, the hot air circulation speed and vibration frequency that cause the abnormality are reversed from the first correspondence and the second correspondence to obtain the combination of key process parameters. The combination of key process parameters is marked as the combination of key process parameters that leads to abnormal dielectric loss.

5. The method according to claim 1, characterized in that, The repair control strategy, which is based on the combination of key process parameters and centers on adjusting the cooling fan at the oven outlet, includes: Receive the combination of key process parameters and the spatial coordinates of the defect to obtain the target defect information; Based on the defect type in the target defect information, a preset control strategy library is queried to obtain a matching control strategy; Based on the matching control strategy, a repair control strategy with the core command of adjusting the speed of the cooling fan at the oven outlet section is generated, and an initial repair strategy is obtained. The initial repair strategy is parameterized by setting specific target values ​​for the fan speed and the duration of action to obtain a parameterized repair strategy. The parameterized repair strategy is bound to the defect space coordinates in the target defect information to obtain the bound repair control strategy.

6. The method according to claim 5, characterized in that, The execution of the repair control strategy, based on the identified defect spatial coordinates, involves altering the leveling and curing rate of the insulating varnish to induce the microcapsules to release repair monomers at the defect spatial coordinates, thereby achieving self-repair of the coating structure. This includes: According to the repair control strategy, the cooling fan at the oven outlet is adjusted to the target speed, and a target cooling environment is established at the identified defect spatial coordinates. Based on the target cooling environment, a local non-uniform temperature field is formed on the surface of the enameled wire at the defect spatial coordinates. The target thermal stress induced at the defect by the local non-uniform temperature field causes the microcapsule wall material at the defect to rupture. The repair monomers inside the ruptured microcapsules are released to the coating defects, where they undergo a polymerization reaction to complete the self-repair of the coating structure.

7. The method according to claim 1, characterized in that, When the marked enameled wire segment reaches the terahertz detection station, the marked enameled wire segment is scanned using terahertz waves to generate evaluation indicators including repair uniformity and bonding integrity, including: When the enameled wire section carrying the location marker arrives at the terahertz detection station, the terahertz scanning command is triggered. According to the terahertz scanning command, the marked segment is scanned using a terahertz wave to obtain the original terahertz signal; The original terahertz signal that penetrates the coating is received and processed to generate a terahertz time-domain spectrum; Based on the terahertz time-domain spectrum, the dielectric constant of multiple micropoints in the repair region is measured to obtain a dielectric constant dataset. The standard deviation of the dielectric constant dataset is calculated to obtain the repair uniformity index; Based on the terahertz time-domain spectrum, the reflectivity of the terahertz wave is measured at the repair interface to obtain interface reflectivity data. The degree of agreement between the interface reflectivity data and the theoretical complete interface reflectivity is calculated to obtain the integrity index of the interface. By integrating the repair uniformity index and the bonding integrity index, an evaluation index for repair effect is output.

8. The method according to claim 7, characterized in that, The optimization direction for the oven temperature zone setpoint and coating pressure parameters, derived in reverse based on the mapping relationship, includes: The evaluation data is obtained by calling the aforementioned repair effect evaluation indicators; The original defect features in the coating curing state feature set are called to obtain defect feature data; The evaluation data, the defect feature data, and the repair process parameters used in the repair are correlated to generate a triplet dataset; Analyze the triplet dataset, establish the mapping relationship between defect type, repair process and repair effect, and obtain the repair effect mapping relationship table; Based on the repair effect mapping table, historical records in which the evaluation data reached the preset quality standard were selected to obtain a set of compliant repair records. Extract the corresponding oven temperature zone setting and painting pressure parameters from the set of qualified repair records to obtain a set of reference process parameters; Analyze the statistical distribution of the reference process parameters for the oven temperature zone settings to determine the optimal parameter range for the oven temperature zone settings; Analyze the statistical distribution of the coating pressure parameters in the reference process parameter set to determine the optimal parameter range for the coating pressure parameters.

9. The method according to claim 1, characterized in that, The method further includes: Based on the optimized parameter range of the oven temperature zone setting value and the optimized parameter range of the coating pressure parameter, optimized process parameters are generated. The optimized process parameters are written into the production line control parameters to complete the parameter update; the updated process parameters are executed to continue producing new enameled wires under the new process conditions. Collect the dielectric loss data of the new enameled wire to form a new round of dielectric loss data; The new round of dielectric loss data is correlated with the optimized process parameters to generate verification results; Based on the verification results, the optimized process parameters are fine-tuned to form an adaptive control closed loop.

10. An AI-powered intelligent analysis system for enameled wire dielectric loss spectrum, characterized in that, The system includes: The dielectric loss monitoring module is configured to monitor the polarization relaxation signal of the insulating varnish during the heating process by a dielectric sensor when the enameled wire enters the oven curing stage, and obtain the raw data of dielectric loss reflecting the degree of polymer crosslinking. The coupling graph construction module is used to spatiotemporally correlate the raw data of dielectric loss with the oven temperature gradient and the winding tension fluctuation to construct a coupling graph of dielectric loss and process parameters. The feature extraction module is used to perform differential processing on the coupling spectrum to extract feature parameters such as the moving rate of the curing front of the insulating varnish, the change of the solvent evaporation interface position, and the fluctuation of the microcapsule distribution density, so as to obtain the feature set of the coating curing state. The key parameter identification module is used to establish the correspondence between the oven hot air circulation speed, the vibration frequency of the coating mold and the curing front morphology based on the coating curing state feature set, and to identify the key process parameter combinations that cause abnormal media loss. The repair strategy generation module is used to generate a repair control strategy based on the combination of key process parameters, with the core being the adjustment of the oven outlet cooling fan. The self-healing execution module is used to execute the repair control strategy. Based on the identified defect spatial coordinates, it changes the leveling and curing rate of the insulating varnish to induce the microcapsules to release repair monomers at the defect spatial coordinates, thereby achieving self-healing of the coating structure. The location marking module is used to mark the location of the self-healing enameled wire segment based on the defect spatial coordinates. The repair evaluation module is used to scan the marked enameled wire section with terahertz waves when the marked enameled wire section is run to the terahertz detection station, and generate evaluation indicators including repair uniformity and bonding integrity. The mapping relationship establishment module is used to perform correlation analysis between the evaluation index and the original defect features in the feature set, and establish a mapping relationship between defect type, repair process and repair effect; The process optimization module is used to deduce the optimization direction of the oven temperature zone setting value and the coating pressure parameter based on the mapping relationship.

Images