A method and system for on-line monitoring of the production cycle of a high voltage cable joint

Abstract

The application discloses a production cycle online monitoring method and system for a high-voltage cable joint. The method comprises the following steps: configuring an identification code for the joint and collecting extrusion process parameters and fastening assembly parameters; converting the parameters into an initial intrinsic state parameter set containing residual stress of an insulation layer, interface roughness and contact pressure through a mapping model; constructing an electric-thermal-mechanical coupling degradation model with the parameter set as an initial boundary; collecting temperature, partial discharge and load current data in a running stage; inversely solving dielectric performance decay and contact resistance increment at the current time from the measured data; comparing the current state parameters with the initial state parameters to obtain an offset and a change rate, and determining an insulation performance degradation grade and simulating and predicting a remaining life according to the offset and the change rate. The application realizes individualized state evaluation and life prediction by establishing a physical correlation between manufacturing information and running data.

Description

Technical Field

[0001] This application relates to the field of high-voltage cable technology, specifically to a method and system for online monitoring of the production cycle of high-voltage cable joints. Background Technology

[0002] High-voltage cable joints are critical connecting components in transmission lines, and their operational reliability directly impacts the safety and stability of the entire power grid. Currently, condition monitoring technologies for high-voltage cable joints primarily focus on the post-commissioning operational phase. This involves deploying devices such as temperature sensors and partial discharge sensors to collect real-time data and then using preset thresholds or data-driven models to provide early warnings of faults. These methods can, to some extent, detect abnormal operating conditions and provide a reference for fault handling.

[0003] However, existing monitoring technologies still have shortcomings in practical applications. On the one hand, process data during manufacturing and mechanical assembly data during on-site installation are usually only stored as static archives, lacking effective correlation with dynamic monitoring data during operation. When anomalies occur during operation, it is difficult to determine whether the anomaly is related to specific factors in the manufacturing or installation process, and it is also impossible to quantify and assess the impact of initial performance differences of the joints before commissioning on subsequent service life. On the other hand, existing service life prediction methods mostly use statistical models, and their input parameters and output results lack a clear physical correlation, making it difficult to explain the inherent laws of performance degradation. Consequently, the individual specificity and confidence level of the prediction results need to be improved.

[0004] Therefore, how to establish a data association mechanism that spans the entire life cycle of high-voltage cable joints, and on this basis, realize individualized condition assessment and life prediction with physical interpretability, has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This application provides a method and system for online monitoring of the production cycle of high-voltage cable joints, which at least addresses the problems existing in the prior art.

[0006] A first aspect of this application provides a method for online monitoring of the production cycle of high-voltage cable joints, comprising the following steps: S1: Configure an identification code for the high-voltage cable joint, and collect the process parameters of the joint in the insulation extrusion process and the assembly parameters in the bolt fastening process based on the identification code; S2: Input the process parameters and assembly parameters into a pre-established mapping model, and output the initial intrinsic state parameter set of the joint from the mapping model. The initial intrinsic state parameter set includes the residual stress distribution of the insulating layer and the interface roughness of the semiconductive layer calculated from the process parameters, as well as the contact interface pressure distribution calculated from the assembly parameters. S3: Establish an electro-thermal-mechanical coupled degradation model for the high-voltage cable joint. The coupled degradation model uses the initial intrinsic state parameter set as the initial boundary condition and integrates the electrothermal aging sub-model of the insulating material and the oxidation resistance-increasing sub-model of the contact interface. S4: Collect real-time monitoring data of the high-voltage cable joint under operating conditions. The real-time monitoring data includes at least the joint body temperature, partial discharge pulse, and load current. S5: Substitute the real-time monitoring data into the coupling degradation model for inverse solution, and calculate the real-time intrinsic state parameter set at the current moment through an iterative algorithm. The real-time intrinsic state parameter set includes the current dielectric performance attenuation of the insulation layer and the current contact resistance increase. S6: Compare each parameter in the real-time intrinsic state parameter set with the corresponding parameters in the initial intrinsic state parameter set one by one, and calculate the current offset of each parameter; fit the change rate of each parameter according to the calculated value of the current offset at multiple consecutive sampling times; determine the insulation performance degradation level of the high-voltage cable joint according to the current offset and the change rate, and simulate and calculate its remaining service life based on the coupled degradation model.

[0007] By configuring identification codes for high-voltage cable joints and associating these codes with process and assembly parameters collected during production and installation, a data traceability foundation spanning the entire lifecycle of the joint was established. This ensures that all subsequent monitoring and analysis correspond to the individual historical information of a specific joint. By inputting process and assembly parameters into a mapping model and outputting an initial intrinsic state parameter set containing residual stress distribution in the insulation layer, interface roughness of the semiconductive layer, and pressure distribution at the contact interface, discrete manufacturing process information was transformed into physically meaningful initial conditions for the model. This allowed the subsequently established coupled degradation model to carry the unique characteristics of the individual joint from the outset. By establishing an electro-thermal-mechanical coupled degradation model and using the initial intrinsic state parameter set as initial boundary conditions, a simulation framework capable of reflecting multi-physics interactions and matching the initial state of the individual joint was constructed, providing a physical basis for subsequent state inversion and life prediction. By collecting real-time monitoring data during operation and substituting it into the coupled degradation model for inverse solving, real-time intrinsic state parameters that cannot be directly measured at the current moment can be inverted from the measurable macroscopic response, achieving dynamic synchronization between the model state and the physical entity state. By comparing real-time intrinsic state parameters with initial intrinsic state parameters and calculating the offset and rate of change, a quantitative assessment of the degree of joint performance degradation is achieved. Based on this, the insulation performance degradation level determined and the remaining service life calculated by simulation have a clear physical basis. Thus, based on the integration of manufacturing and installation information and operation monitoring data, individualized condition assessment and life prediction with physical interpretability are realized.

[0008] In some embodiments, in step S3, the electro-thermal-mechanical coupling degradation model includes a material state evolution sub-model, a whole-field distribution calculation sub-model, and a parameter transfer function; The material state evolution sub-model is used to describe the variation of residual stress distribution, interface roughness, and contact pressure with temperature, electric field, and mechanical stress in the initial intrinsic state parameter set. The overall field distribution calculation sub-model is used to calculate the distribution of temperature field, electric field and stress field of the high-voltage cable joint based on its geometric structure, material properties and operating conditions. The parameter transfer function is used to convert the current material state parameters output by the material state evolution sub-model into the material property parameters required by the whole field distribution calculation sub-model.

[0009] By dividing the electro-thermal-mechanical coupled degradation model into three components—a material state evolution sub-model, a whole-field distribution calculation sub-model, and a parameter transfer function—the boundaries and interactions between different functional modules within the model are clearly defined. The material state evolution sub-model focuses on describing the distribution of residual stress, interface roughness, and contact pressure in the initial intrinsic state parameter set as a function of temperature, electric field, and mechanical stress, enabling independent modeling and calculation of the material degradation process based on physical mechanisms. The whole-field distribution calculation sub-model is responsible for calculating the macroscopic distribution of temperature, electric, and stress fields of the high-voltage cable joint based on its geometry, material properties, and operating conditions, providing spatial distribution information for assessing the overall operating state of the joint. The parameter transfer function acts as a bridge connecting the two sub-models, converting the current material state parameters output by the material state evolution sub-model into the material property parameters required by the whole-field distribution calculation sub-model, achieving dynamic coupling between microscopic material property evolution and macroscopic field distribution calculation. This hierarchical modeling and parameter transfer architecture allows the coupled degradation model to maintain clarity of the physical mechanisms while also considering the engineering practicality of macroscopic response calculations.

[0010] In some embodiments, the parameter transfer function specifically includes: converting the current molecular chain orientation degree and residual stress value calculated by the material state evolution sub-model into the thermal conductivity correction coefficient and dielectric constant correction coefficient of the insulating material in the whole field distribution calculation sub-model; and converting the current true contact area ratio calculated by the material state evolution sub-model into the contact resistance coefficient and contact thermal resistance coefficient in the whole field distribution calculation sub-model.

[0011] By specifying the parameter transfer function as the conversion of molecular chain orientation and residual stress values ​​to thermal conductivity and dielectric constant correction coefficients for insulating materials, and the conversion of actual contact area ratio to contact resistance and contact thermal resistance, the quantitative mapping relationship between the output of the material state evolution sub-model and the input of the whole-field distribution calculation sub-model is clarified. Specifically, converting molecular chain orientation and residual stress values ​​to thermal conductivity and dielectric constant correction coefficients allows changes in the thermal conductivity and dielectric properties of insulating materials due to aging during long-term operation to be dynamically reflected in the calculation of the temperature and electric fields. Converting actual contact area ratio to contact resistance and contact thermal resistance allows degradation of the interface contact state due to fretting wear or oxidation to directly affect the calculation results of electrical and heat transfer processes. Through this parameter conversion method, changes in the internal state of the material can be quantitatively transferred to the whole-field distribution calculation sub-model, ensuring a clear correspondence between the data transfers of the various components in the coupled degradation model.

[0012] In some embodiments, step S5, which involves substituting real-time monitoring data into the coupling degradation model for inverse solution, specifically includes the following sub-steps: S51: Construct an objective function with the goal of minimizing the weighted sum of squared residuals between the theoretical temperature distribution and theoretical electric field distribution output by the overall field distribution calculation sub-model and the measured temperature data and partial discharge pulse data in step S4; S52: Take the real-time intrinsic state parameter set as the variable to be optimized and set its value range; S53: Iteratively adjust the variable to be optimized using the adjoint state method or integrated Kalman filter algorithm until the objective function converges, and output the real-time intrinsic state parameter set at the current moment.

[0013] By decomposing the inverse solution process into three sub-steps—constructing the objective function, setting the range of values ​​for optimization variables, and iteratively solving using a specific algorithm—a concrete implementation path is provided for inferring intrinsic state parameters from real-time monitoring data. Specifically, the objective function is constructed with the goal of minimizing the weighted sum of squared residuals between the theoretical temperature distribution, theoretical electric field distribution, and measured temperature and partial discharge pulse data output by the whole-field distribution calculation sub-model. This clarifies the optimization direction of the inverse calculation and gives the model parameter adjustment process a clear convergence objective. Using the real-time intrinsic state parameter set as the variable to be optimized and setting its value range constrains the parameter variation boundaries during the solution process, preventing the inverse results from deviating from engineering reality. Iteratively adjusting the variable to be optimized using the adjoint state method or integrated Kalman filter algorithm until the objective function converges allows for the gradual approximation of the true state using redundant information in the monitoring data. The final output set of real-time intrinsic state parameters represents the estimate with the highest degree of matching to the model under the current monitoring data conditions. This structured inverse solution method enables the coupled degradation model to adapt to changes in monitoring data under different operating conditions, improving the stability and reliability of the inverse results.

[0014] In some embodiments, the real-time monitoring data in step S4 further includes vibration acceleration signals; the objective function in step S51 simultaneously considers the spatial distribution residual of the temperature field, the phase distribution characteristics of the partial discharge pulse, and the spectral energy distribution differences of the vibration signal.

[0015] By further incorporating vibration acceleration signals into real-time monitoring data and simultaneously considering the spatial distribution residuals of the temperature field, the phase distribution characteristics of partial discharge pulses, and the spectral energy distribution differences of vibration signals in the objective function, the information dimensions upon which the inverse solution is based are expanded. The spatial distribution residuals of the temperature field reflect the degree of consistency of the overall thermal state of the joint; the phase distribution characteristics of partial discharge pulses reflect the discharge characteristics of insulation defects; and the spectral energy distribution differences of vibration signals are related to the dynamic changes in the mechanical connection state. Incorporating these three different types of monitoring information with complementary physical meanings into the objective function creates multi-dimensional constraints on the real-time intrinsic state parameter set during the inverse solution process. This helps to distinguish the influence of different physical processes on the changes in intrinsic state parameters, reduces the non-uniqueness of the inversion results, and improves the identification accuracy of the real-time intrinsic state parameter set.

[0016] In some embodiments, step S6, specifically calculating the remaining service life based on the coupling degradation model simulation, includes: Using the current real-time intrinsic state parameter set as the initial value, and the preset future typical load curve and ambient temperature curve as input conditions, the electro-thermal-mechanical coupling degradation model is driven to perform time-domain progressive simulation. During the simulation, the key performance parameters output by the whole field distribution calculation sub-model are monitored in real time. The key performance parameters include at least the maximum field strength, the maximum temperature, and the contact resistance value. When any of the key performance parameters reaches its preset failure threshold, the simulation is stopped, and the time from the start time to the stop time of the simulation is taken as the remaining service life.

[0017] By using the current real-time intrinsic state parameter set as a starting point and pre-set future typical load curves and ambient temperature curves as input conditions, a time-domain progressive simulation of the electro-thermal-mechanical coupling degradation model is driven, clarifying the specific implementation method of remaining service life prediction. Using the real-time intrinsic state parameter set as the simulation starting point ensures that the prediction process is based on the current real state of the joint, rather than an assumed ideal state. Using pre-set future typical load curves and ambient temperature curves as input conditions allows the simulation process to reflect the actual working conditions the joint will experience in future operation. During the simulation, key performance parameters output by the field distribution calculation sub-model are monitored in real time, including maximum field strength, maximum temperature, and contact resistance values. These parameters are directly related to the insulation performance and connection reliability of the joint. The simulation stops when any key performance parameter reaches its pre-set failure threshold, and the time from the start to the stop of the simulation is taken as the remaining service life. This ensures that the service life prediction results directly correspond to specific failure criteria, providing a quantifiable basis for operation and maintenance decisions.

[0018] In some embodiments, step S7 is also included: S7: When the insulation performance degradation level determined in step S6 reaches the preset warning level, based on the abnormal parameter components in the real-time intrinsic state parameter set and combined with the inverse function of the mapping model, the original process parameter deviation range that caused the abnormal parameter components is traced back in reverse, and a process optimization report containing specific improvement suggestions is generated.

[0019] By analyzing the abnormal parameter components in the real-time intrinsic state parameter set and using the inverse function of the mapping model to trace back the original process parameter deviation range that caused the abnormal parameter component when the insulation performance degradation level reaches a preset warning level, a causal relationship between the abnormal operating state and the manufacturing and installation process is established. The abnormal parameter components are core indicators reflecting the current performance degradation, and their numerical changes directly point to the degree of degradation of a specific physical mechanism. By tracing back using the inverse function of the mapping model, abnormal changes in the intrinsic state parameter space can be mapped back to the original process parameter space, allowing the calculation of the process parameter deviation range that may have caused the abnormal change. Generating a process optimization report with specific improvement suggestions extends the monitoring results beyond the state assessment itself to the process improvement stage, providing data support for continuous optimization of the production and installation process. This reverse tracing mechanism from operational abnormalities to process deviations gives the monitoring system diagnostic capabilities, enabling it to feed back accumulated experience to the upstream manufacturing process.

[0020] In some embodiments, step S7 specifically includes: if the current contact resistance increase in the real-time intrinsic state parameter set exceeds a preset threshold, then the corresponding assembly torque deviation value and torque application rate deviation value are calculated through the contact mechanics inverse function in the mapping model, and the bolt position where the deviation occurs and the suggested correction value are indicated in the process optimization report.

[0021] By focusing on the specific scenario of abnormal contact resistance increases, the execution method of reverse tracing was further clarified. When the current contact resistance increase in the real-time intrinsic state parameter set exceeds a preset threshold, the corresponding assembly torque deviation and torque application rate deviation are calculated using the inverse contact mechanics function in the mapping model. The process optimization report then indicates the bolt location where the deviation occurred and the suggested correction value. The inverse contact mechanics function establishes a quantitative mapping relationship between the contact resistance increase and the original assembly parameters, allowing abnormal changes in contact resistance to be accurately attributed to specific assembly parameter deviations. The calculated torque and application rate deviation values ​​provide a quantitative basis for process improvement, while the bolt location indication enables maintenance personnel to precisely pinpoint the source of the problem. The proposed correction values ​​further enhance the operability of the process optimization report, enabling monitoring results to directly guide on-site operations.

[0022] In some embodiments, in step S2, the assembly parameters include the torque timing curve and the rotation timing curve during the bolt tightening process; the mapping model is established based on contact mechanics and is used to calculate the pressure distribution of the contact interface and the actual contact area of ​​the interface according to the torque timing curve and the rotation timing curve.

[0023] By specifying assembly parameters as torque and rotation time-series curves during bolt tightening, and employing a mapping model based on contact mechanics to calculate the pressure distribution and actual contact area at the contact interface, a quantitative transformation path from raw assembly data to the interface contact state is clarified. The torque and rotation time-series curves record complete dynamic information of the tightening process, providing a more comprehensive reflection of the bolt preload establishment process and potential abnormal conditions compared to a single final torque. The contact mechanics-based mapping model utilizes this time-series data, combined with bolt specifications and friction conditions, to calculate the pressure distribution and actual contact area at the contact interface. These two parameters are directly related to the initial state of contact resistance and contact thermal resistance. In this way, the mechanical assembly data during the installation phase is transformed into initial intrinsic state parameters with clear physical meaning, providing an initial basis for the evolutionary analysis of interface performance in the subsequent coupled degradation model, and also providing data support for the quantitative assessment of installation quality and anomaly tracing.

[0024] In some embodiments, in step S2, the process parameters include the die head temperature time-series curve, the screw speed time-series curve, and the extrusion pressure time-series curve during the insulation extrusion process; the mapping model is established based on material rheology and is used to calculate the residual stress distribution of the insulation layer according to the die head temperature time-series curve, the screw speed time-series curve, and the extrusion pressure time-series curve.

[0025] By specifying process parameters as time-series curves of die head temperature, screw speed, and extrusion pressure during the insulation extrusion process, and employing a mapping model based on materials rheology to calculate the residual stress distribution of the insulation layer, the transformation path from raw process data to the initial state of the material was clarified. The temporal variations of die head temperature, screw speed, and extrusion pressure directly affect the material's flow behavior, cooling rate, and molecular chain orientation during extrusion, which collectively determine the formation of residual stress within the insulation layer. The materials rheology-based mapping model can describe the rheological properties of the material under high temperature and shear conditions, as well as the stress evolution during cooling and solidification, transforming the raw process time-series data into quantitative results of residual stress distribution. As one of the initial intrinsic state parameters, the residual stress of the insulation layer is a crucial factor influencing the subsequent electrothermal aging behavior of the material. The residual stress distribution obtained in this way has higher individual specificity and accuracy compared to empirical estimates.

[0026] In some embodiments, step S6, determining the insulation performance degradation level specifically includes: Calculate the relative degradation of the current offset of each parameter relative to its initial value; Calculate the average degradation rate of each parameter within a preset time window; Using the relative degradation degree and the average degradation rate as input features, a pre-trained health status classification model is invoked to output the insulation performance degradation level of the high-voltage cable joint.

[0027] By specifying the process of determining the insulation performance degradation level as calculating the relative degradation degree, calculating the average degradation rate, constructing a health state feature vector, and calling the health state classification model to output the level, the quantitative path from parameter offset to degradation level is clarified. The relative degradation degree reflects the degree of each parameter's current offset relative to its initial value, embodying the cumulative effect of performance degradation; the average degradation rate reflects the speed trend of performance degradation. Combining these two factors provides a more comprehensive description of the degradation state. A health state feature vector is constructed using the relative degradation degree and average degradation rate as input features, integrating the degradation information of multiple parameters into a unified data expression. Calling the pre-trained health state classification model to output the insulation performance degradation level provides a data-driven objective basis for the degradation level determination process. The classification model can establish a mapping relationship between the feature vector and the level based on historical fault cases and simulation data. This quantitative evaluation method makes the determination of the insulation performance degradation level repeatable and consistent, reducing the uncertainty caused by subjective judgment.

[0028] In some embodiments, step S8 is also included: S8: Establish a digital profile of the high-voltage cable joint based on the identification code. The digital profile includes at least the initial intrinsic state parameter set, the historical sets of real-time intrinsic state parameters, the historical records of insulation performance degradation levels, and the remaining service life prediction curve, and provides a visual query interface.

[0029] By establishing digital archives for high-voltage cable joints based on identification codes, and incorporating initial intrinsic state parameter sets, historical real-time intrinsic state parameter sets, historical insulation performance degradation levels, and remaining service life prediction curves, a visual query interface is provided, enabling centralized management and intuitive presentation of monitoring data throughout the entire lifecycle. The identification code serves as a unique index throughout the entire process, ensuring that data generated at different stages can be accurately attributed to the same joint, avoiding information misalignment or loss. The initial intrinsic state parameter set records the initial performance characteristics of the joint before commissioning, the historical real-time intrinsic state parameter sets reflect the dynamic evolution trajectory of performance during operation, the historical insulation performance degradation level records the changes in health status, and the remaining service life prediction curve provides a forecast of future trends. Through the visual query interface, maintenance personnel can intuitively view the above information graphically, compare parameter changes at different times, analyze the evolutionary patterns of performance degradation, and provide data support for maintenance decisions.

[0030] A second aspect of this application provides an online monitoring system for the production cycle of high-voltage cable joints, the system comprising: The identification code management unit is used to generate and assign identification codes to high-voltage cable joints, and to establish the association between the identification codes and the full-cycle data; The process data acquisition unit is used to acquire the timing data of the insulation extrusion process parameters and the timing data of the bolt fastening assembly parameters associated with the identification code from the extrusion production line control system and intelligent assembly tools. The parameter parsing and modeling unit has the mapping model built in, which is used to calculate and output the initial intrinsic state parameter set based on the process parameter time series data and the assembly parameter time series data, and to construct the electro-thermal-mechanical coupling degradation model based on the initial intrinsic state parameter set. The operation monitoring data acquisition unit is deployed at the high-voltage cable joint site to collect temperature, partial discharge pulse, load current and vibration acceleration data associated with the identification code in real time; The state inversion calculation unit is connected to the operation monitoring data acquisition unit and the parameter analysis and modeling unit. It is used to drive the electro-thermal-mechanical coupling degradation model to perform inverse solution with the real-time monitoring data as input, and calculate and output the real-time intrinsic state parameter set. The health assessment and life prediction unit is used to calculate the current offset and rate of change of each parameter based on the initial intrinsic state parameter set and the real-time intrinsic state parameter set, determine the insulation performance degradation level, and simulate and calculate the remaining service life. The source tracing diagnosis and process feedback unit is used to generate and output a process optimization report when the insulation performance degradation level reaches the warning level, based on the abnormal parameter components in the real-time intrinsic state parameter set and combined with the inverse function of the mapping model. The data storage and visualization unit is used to store the full-cycle data associated with the identification code and to provide a visual query interface for digital archives.

[0031] By integrating an identification code management unit, a process data acquisition unit, a parameter analysis and modeling unit, an operation monitoring data acquisition unit, a state inversion calculation unit, a health assessment and life prediction unit, a traceability diagnosis and process feedback unit, and a data storage and visualization unit, an online monitoring and life prediction system covering the entire life cycle of high-voltage cable joints has been constructed. The identification code management unit establishes a unique identifier for each joint and associates it with full-cycle data, laying the foundation for data traceability. The process data acquisition unit obtains time-series process parameters and assembly parameters associated with the identification code from the extrusion production line control system and intelligent assembly tools, ensuring accurate acquisition of front-end data. The parameter analysis and modeling unit has a built-in mapping model that transforms raw process data into an initial intrinsic state parameter set and constructs a coupled degradation model, realizing the transformation of manufacturing information into a simulation model. The operation monitoring data acquisition unit is deployed on-site to collect real-time temperature, partial discharge, load current, and vibration acceleration data, providing input for state inversion. The state inversion calculation unit uses real-time monitoring data to drive the coupled degradation model to perform inverse solving, outputting the real-time intrinsic state parameter set at the current moment. The health assessment and lifespan prediction unit determines the insulation performance degradation level and simulates and calculates the remaining service life based on the comparison of initial and real-time intrinsic state parameters. The source tracing and process feedback unit, when the degradation level reaches the warning level, uses the inverse function of the mapping model to trace process parameter deviations and generates an optimization report. The data storage and visualization unit centrally stores full-cycle data and provides a visual query interface. Through the collaborative work of these units, the system achieves a complete functional closed loop from data acquisition, state inversion, health assessment to source tracing and diagnosis. Attached Figure Description

[0032] Figure 1 A schematic flowchart illustrating an online monitoring method for the production cycle of high-voltage cable joints provided in this application embodiment; Figure 2 A schematic diagram of the framework of an online monitoring system for the production cycle of high-voltage cable joints provided in this application embodiment; Figure 3 A schematic diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0033] 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 and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0034] In existing high-voltage cable joint condition monitoring technologies, the focus is typically on the operational phase after commissioning. Real-time data is collected by deploying devices such as temperature sensors and partial discharge sensors, and fault warnings are issued based on preset thresholds or data-driven models. However, these methods merely store process parameters from the manufacturing process and mechanical assembly data from on-site installation as static records, lacking an effective correlation with dynamic monitoring data during operation. This makes it difficult to trace anomalies observed during operation back to specific manufacturing or installation stages, and also fails to assess the impact of pre-existing initial performance differences on subsequent service life. Furthermore, existing life prediction models often employ statistical methods, lacking a clear physical correlation between input parameters and output results, making it difficult to explain the inherent laws of performance degradation. The individual specificity and confidence level of the prediction results need improvement.

[0035] Analysis revealed that the root cause of the aforementioned problems lies in the disconnect between manufacturing and installation information and operational monitoring data, as well as the lack of individual information during model construction. Process parameters in the production stage, such as the temporal variations of extrusion temperature, screw speed, and extrusion pressure, and assembly parameters in the installation stage, such as the dynamic curves of torque and rotation angle, actually contain crucial information affecting the long-term performance of the joint. This includes, for example, the residual stress distribution within the insulation material, the micro-roughness of the semi-conductive layer interface, and the pressure distribution at the contact interface. However, this information was not effectively extracted and integrated into the subsequent condition assessment model. Furthermore, temperature and partial discharge data collected during the operation phase only reflect the current apparent state and cannot be effectively compared with the initial performance characteristics of the joint, resulting in a lack of reliable physical basis for condition assessment and life prediction.

[0036] Based on the above analysis, this application proposes a monitoring approach covering the entire lifecycle of high-voltage cable joints. First, by establishing a unique identifier for each joint, process parameters and assembly parameters from the production and installation stages are correlated and collected, establishing a foundation for data traceability. Second, using a mapping model based on material rheology and contact mechanics, these raw time-series process data are transformed into initial intrinsic state parameters with clear physical meaning, including the residual stress distribution of the insulation layer, the interface roughness of the semiconductive layer, and the pressure distribution at the contact interface. This allows the subsequently established coupled degradation model to reflect the unique characteristics of each individual joint from the outset. Furthermore, by collecting real-time monitoring data during the operational phase and driving the electro-thermal-mechanical coupled degradation model for inverse solution, it is possible to deduce real-time intrinsic state parameters that cannot be directly measured at the current moment from measurable macroscopic responses such as temperature fields and partial discharge signals. Then, by comparing the differences between the initial and current states, a physically based degradation level determination and remaining life prediction can be achieved, while also providing data support for tracing and diagnosing abnormal states.

[0037] Please refer to Figure 1 , Figure 1 This application provides a method for online monitoring of the production cycle of high-voltage cable joints, comprising the following steps: S1: Configure an identification code for the high-voltage cable joint, and collect the process parameters of the joint in the insulation extrusion process and the assembly parameters in the bolt fastening process based on the identification code; S2: Input the process parameters and assembly parameters into the pre-established mapping model, and output the initial intrinsic state parameter set of the joint from the mapping model. The initial intrinsic state parameter set includes the residual stress distribution of the insulation layer and the interface roughness of the semiconductive layer calculated from the process parameters, as well as the contact interface pressure distribution calculated from the assembly parameters. S3: Establish an electro-thermal-mechanical coupled degradation model for high-voltage cable joints. The coupled degradation model uses the initial intrinsic state parameter set as the initial boundary condition and integrates the electrothermal aging sub-model of insulation material and the oxidation resistance-increasing sub-model of contact interface. S4: Collect real-time monitoring data of the high-voltage cable joint under operating conditions. The real-time monitoring data shall include at least the joint body temperature, partial discharge pulse, and load current. S5: Substitute the real-time monitoring data into the coupled degradation model for inverse solution, and calculate the real-time intrinsic state parameter set at the current moment through an iterative algorithm. The real-time intrinsic state parameter set includes the current dielectric performance attenuation of the insulation layer and the current contact resistance increase. S6: Compare each parameter in the real-time intrinsic state parameter set with the corresponding parameters in the initial intrinsic state parameter set one by one, and calculate the current offset of each parameter; based on the calculated values ​​of the current offset at multiple consecutive sampling times, fit the change rate of each parameter; based on the current offset and change rate, determine the insulation performance degradation level of the high-voltage cable joint, and simulate and calculate its remaining service life based on the coupled degradation model.

[0038] It is understandable that the identification code refers to the coded identifier assigned to a high-voltage cable joint for identification throughout its entire lifecycle. This can be implemented using QR codes, RFID tags, or numerical codes, and its function is to link data generated at each stage of production, installation, and operation to the same joint object. Process parameters refer to data collected during the insulation extrusion process that reflects the manufacturing process status. Examples include curves showing the change in die head temperature over time, screw speed over time, and extrusion pressure over time. These parameters reflect the thermal history and flow state of the material during extrusion molding. Assembly parameters refer to data collected during the bolt tightening process that reflects the installation process status. Examples include curves showing the change in torque over time and rotation angle over time. These parameters reflect the establishment process of bolt preload and the stress state of the interface. The mapping model is a pre-established computational model used to convert process parameters and assembly parameters into initial intrinsic state parameters. This model can be constructed based on theories such as material rheology, curing kinetics, or contact mechanics. The initial intrinsic state parameter set refers to a set of parameters output by the mapping model that describe the initial performance characteristics of the joint. These parameters may include the residual stress distribution of the insulation layer, the interface roughness of the semiconducting layer, and the pressure distribution at the contact interface. Specifically, the residual stress distribution of the insulation layer reflects the internal stress state of the material during extrusion molding and cooling curing; the interface roughness of the semiconducting layer reflects the microscopic morphology of the interface between the insulation and semiconducting layers; and the pressure distribution at the contact interface reflects the pressure space distribution formed at the connection interface after fastening.

[0039] The electro-thermal-mechanical coupled degradation model is a simulation model that can simultaneously describe the interaction of electric field, temperature field, and stress field and their impact on joint performance. This model uses an initial intrinsic state parameter set as initial boundary conditions and integrates an electrothermal aging sub-model for insulating materials and an oxidation resistance-increasing sub-model for the contact interface. The electrothermal aging sub-model for insulating materials describes the performance degradation of insulating materials under the combined effects of electric field and temperature, while the oxidation resistance-increasing sub-model for the contact interface describes the increase in contact resistance due to oxidation or fretting wear. Real-time monitoring data refers to data reflecting the operating status collected after the high-voltage cable joint is put into operation, which may include joint body temperature, partial discharge pulses, and load current. The joint body temperature can be obtained through distributed optical fibers or point-type temperature sensors, the partial discharge pulses can be obtained through high-frequency current transformers or capacitively coupled sensors, and the load current can be obtained through current transformers. Inverse solving refers to the process of substituting real-time monitoring data into the coupled degradation model and deriving the internal state parameters at the current moment through reverse calculation. Iterative algorithms refer to a class of computational methods that repeatedly correct the parameters to be optimized during the inverse solving process to make the model output approximate the measured data. The real-time intrinsic state parameter set refers to the parameters reflecting the internal state of the joint at the current moment, obtained through inverse calculation. These parameters can include the current dielectric performance degradation of the insulation layer and the current increase in contact resistance. The dielectric performance degradation reflects the degree of change in dielectric properties due to aging of the insulation material, while the increase in contact resistance reflects the increase in resistance at the connection interface due to degradation. The current offset refers to the difference between each parameter in the real-time intrinsic state parameter set and the corresponding parameter in the initial intrinsic state parameter set. The rate of change refers to how quickly the current offset changes over time, and can be obtained by fitting the offsets from multiple consecutive sampling moments. The insulation performance degradation level is a classification of the joint insulation state based on the current offset and the rate of change. The remaining service life is the length of time from the current moment to the failure moment, calculated based on the coupled degradation model simulation.

[0040] By configuring identification codes for high-voltage cable joints and associating these codes with process and assembly parameters collected during production and installation, a data traceability foundation spanning the entire lifecycle of the joint was established. This ensures that all subsequent monitoring and analysis correspond to the individual historical information of a specific joint. By inputting process and assembly parameters into a mapping model and outputting an initial intrinsic state parameter set containing residual stress distribution in the insulation layer, interface roughness of the semiconductive layer, and pressure distribution at the contact interface, discrete manufacturing process information was transformed into physically meaningful initial conditions for the model. This allowed the subsequently established coupled degradation model to carry the unique characteristics of the individual joint from the outset. By establishing an electro-thermal-mechanical coupled degradation model and using the initial intrinsic state parameter set as initial boundary conditions, a simulation framework capable of reflecting multi-physics interactions and matching the initial state of the individual joint was constructed, providing a physical basis for subsequent state inversion and life prediction. By collecting real-time monitoring data during operation and substituting it into the coupled degradation model for inverse solving, real-time intrinsic state parameters that cannot be directly measured at the current moment can be inverted from the measurable macroscopic response, achieving dynamic synchronization between the model state and the physical entity state. By comparing real-time intrinsic state parameters with initial intrinsic state parameters and calculating the offset and rate of change, a quantitative assessment of the degree of joint performance degradation is achieved. Based on this, the insulation performance degradation level determined and the remaining service life calculated by simulation have a clear physical basis. Thus, based on the integration of manufacturing and installation information and operation monitoring data, individualized condition assessment and life prediction with physical interpretability are realized.

[0041] In some embodiments disclosed in this application, in step S3, the electro-thermal-mechanical coupling degradation model includes a material state evolution sub-model, a whole-field distribution calculation sub-model, and a parameter transfer function; The material state evolution sub-model is used to describe the variation of residual stress distribution, interface roughness, and contact pressure with temperature, electric field, and mechanical stress in the initial intrinsic state parameter set. The whole field distribution calculation sub-model is used to calculate the distribution of temperature field, electric field and stress field of high-voltage cable joints based on their geometric structure, material properties and operating conditions. The parameter transfer function is used to convert the current material state parameters output by the material state evolution sub-model into the material property parameters required by the whole-field distribution calculation sub-model.

[0042] It is understandable that the electro-thermal-mechanical coupled degradation model specifically comprises three components: a material state evolution sub-model, a whole-field distribution calculation sub-model, and a parameter transfer function. The material state evolution sub-model describes the distribution of residual stress, interface roughness, and contact pressure in the initial intrinsic state parameter set as a function of temperature, electric field, and mechanical stress. It is understood that the material state evolution sub-model focuses on the degradation process of material properties over time. Its inputs include external load conditions such as temperature, electric field strength, and mechanical stress, and its output is the change in the internal state of the material. For example, this sub-model can be based on material aging kinetics; for instance, the Arrhenius equation can be used to describe the thermal activation process, an inverse power model can be used to describe the electrical aging rate, or other empirical formulas or semi-empirical relationships can be used depending on the specific material characteristics. The construction of the material state evolution sub-model needs to consider the physicochemical properties of the insulating material, such as the molecular chain breakage and cross-linking reactions of polymer materials, and the interfacial bonding state between the filler and the matrix. These factors collectively determine the degradation law of material properties.

[0043] The overall field distribution calculation sub-model is used to calculate the distribution of temperature, electric, and stress fields of a high-voltage cable joint based on its geometry, material properties, and operating conditions. This sub-model can be implemented using numerical methods, such as the finite element method, finite volume method, or boundary element method. Geometric parameters can be derived from design drawings or obtained through 3D scanning measurements. Material properties may include thermal conductivity, electrical conductivity, elastic modulus, and coefficient of thermal expansion, which may change with temperature and aging. Operating conditions may include boundary conditions such as load current, ambient temperature, and voltage level. By solving the electric field control equations, heat conduction equations, and solid mechanics equilibrium equations, the overall field distribution calculation sub-model can output the temperature, electric field strength, and stress values ​​at various locations within the joint, forming a spatial distribution result. The parameter transfer function is used to convert the current material state parameters output by the material state evolution sub-model into the material property parameters required by the overall field distribution calculation sub-model. Understandably, parameters output by the material state evolution sub-model, such as molecular chain orientation, residual stress, and actual contact area ratio, cannot be directly used as material property inputs for field distribution calculations and need to be transformed through a parameter transfer function. For example, this transfer function can be established based on homogenization theory, adjusting microscopic material state parameters to macroscopic material properties. For instance, the anisotropy coefficient of thermal conductivity can be calculated based on molecular chain orientation, the dielectric constant of the material can be corrected based on residual stress, and the contact resistance coefficient and contact thermal resistance coefficient can be calculated based on the actual contact area ratio. The parameter transfer function can also be calibrated based on numerical simulation results using the representative volume element method, or using empirical relationships obtained from experimental testing.

[0044] This embodiment of the application divides the electro-thermal-mechanical coupled degradation model into three components: a material state evolution sub-model, a whole-field distribution calculation sub-model, and a parameter transfer function, clarifying the boundaries and interactions between different functional modules within the model. The material state evolution sub-model focuses on describing the distribution of residual stress, interface roughness, and contact pressure in the initial intrinsic state parameter set as a function of temperature, electric field, and mechanical stress, enabling independent modeling and calculation of the material performance degradation process based on physical mechanisms. The whole-field distribution calculation sub-model is responsible for calculating the macroscopic distribution of the temperature, electric, and stress fields of the high-voltage cable joint based on its geometry, material properties, and operating conditions, providing spatial distribution information for evaluating the overall operating state of the joint. The parameter transfer function acts as a bridge connecting the two sub-models, converting the current material state parameters output by the material state evolution sub-model into the material property parameters required by the whole-field distribution calculation sub-model, achieving dynamic coupling between microscopic material performance evolution and macroscopic field distribution calculation. This hierarchical modeling and parameter transfer architecture allows the coupled degradation model to maintain clarity of the physical mechanisms while also considering the engineering practicality of macroscopic response calculations.

[0045] In some embodiments disclosed in this application, the parameter transfer function specifically includes: converting the current molecular chain orientation degree and residual stress value calculated by the material state evolution sub-model into the thermal conductivity correction coefficient and dielectric constant correction coefficient of the insulating material in the whole field distribution calculation sub-model; and converting the current true contact area ratio calculated by the material state evolution sub-model into the contact resistance coefficient and contact thermal resistance coefficient in the whole field distribution calculation sub-model.

[0046] It is understandable that the parameter transfer function specifically includes converting the current molecular chain orientation and residual stress values ​​calculated by the material state evolution sub-model into the thermal conductivity correction coefficient and dielectric constant correction coefficient of the insulating material in the whole-field distribution calculation sub-model; and converting the current true contact area ratio calculated by the material state evolution sub-model into the contact resistance coefficient and contact thermal resistance coefficient in the whole-field distribution calculation sub-model. Molecular chain orientation refers to the degree to which the polymer chains of the insulating material align along the flow direction during extrusion molding. This parameter can be calculated using a rheological model combined with extrusion process parameters, or it can be calibrated using experimental methods such as wide-angle X-ray diffraction. Residual stress refers to the magnitude of the stress remaining inside the material after cooling and solidification, and its distribution is affected by factors such as extrusion temperature and cooling rate. It is understandable that molecular chain orientation and residual stress will change the internal microstructure of the material, thereby affecting its thermal conductivity and dielectric properties. For example, when the molecular chains are highly oriented along a certain direction, the thermal conductivity in that direction may be higher than that in the perpendicular direction, forming anisotropic thermal conductivity. The presence of residual stress may change the polarization behavior of the material, causing a shift in the dielectric constant. The thermal conductivity correction factor and the dielectric constant correction factor are parameters used to quantify these effects. They are used to adjust the basic thermal conductivity and basic dielectric constant of the insulating material so that the whole field distribution calculation sub-model can reflect the changes in macroscopic performance caused by changes in the microscopic state of the material.

[0047] The true contact area ratio refers to the ratio of the actual contact area to the nominal contact area at the connection interface. This parameter can be calculated using a contact mechanics model combined with assembly parameters, or it can be calibrated using surface topography measurement data. The contact interface exhibits an uneven morphology at the microscale, with actual contact occurring only at the tips of the micro-protrusions; therefore, the true contact area ratio is typically much less than 1. During operation, due to fretting wear and oxidation, the state of the contact interface changes, leading to an increase in contact resistance and contact thermal resistance. The contact resistivity coefficient describes the influence of the contact state on electrical conductivity, and its value is negatively correlated with the true contact area ratio. The contact thermal resistance coefficient describes the influence of the contact state on thermal conductivity, and its value is also related to the true contact area ratio. By converting the true contact area ratio into contact resistivity and contact thermal resistance coefficients, the overall field distribution calculation sub-model can accurately reflect the influence of the interface contact state on the electric field distribution and temperature field distribution.

[0048] This application's embodiments specify the parameter transfer function as the conversion of molecular chain orientation and residual stress values ​​to thermal conductivity and dielectric constant correction coefficients for insulating materials, and the conversion of actual contact area ratio to contact resistance coefficient and contact thermal resistance coefficient, thus clarifying the quantitative mapping relationship between the output of the material state evolution sub-model and the input of the whole-field distribution calculation sub-model. Specifically, converting molecular chain orientation and residual stress values ​​to thermal conductivity and dielectric constant correction coefficients allows changes in the thermal conductivity and dielectric properties of the insulating material due to aging during long-term operation to be dynamically reflected in the calculation of the temperature and electric fields. Converting actual contact area ratio to contact resistance coefficient and contact thermal resistance coefficient ensures that degradation of the interface contact state due to fretting wear or oxidation directly affects the calculation results of electrical and heat transfer processes. Through this parameter conversion method, changes in the internal state of the material can be quantitatively transferred to the whole-field distribution calculation sub-model, ensuring a clear correspondence between the data transfers of the various components in the coupled degradation model.

[0049] In some embodiments disclosed in this application, step S5 involves substituting real-time monitoring data into the coupling degradation model for inverse solution, specifically including the following sub-steps: S51: Construct an objective function with the goal of minimizing the weighted sum of squared residuals between the theoretical temperature distribution and theoretical electric field distribution output by the whole field distribution calculation sub-model and the measured temperature data and partial discharge pulse data in step S4; S52: Use the real-time intrinsic state parameter set as the variable to be optimized and set its value range; S53: Iteratively adjust the variables to be optimized using the adjoint state method or integrated Kalman filter algorithm until the objective function converges, and output the real-time intrinsic state parameter set at the current moment.

[0050] Understandably, in step S51, the objective function is constructed with the goal of minimizing the weighted sum of squared residuals between the theoretical temperature distribution and theoretical electric field distribution output by the whole-field distribution calculation sub-model and the measured temperature data and partial discharge pulse data from step S4. The objective function is a mathematical expression used in the reverse solution to quantify the difference between the model output and the measured data. The theoretical temperature distribution and theoretical electric field distribution are calculated by the whole-field distribution calculation sub-model based on the current parameters to be optimized, while the measured temperature data and partial discharge pulse data are acquired by field sensors. The weighted sum of squared residuals refers to squaring the difference between the processed theoretical value and the measured value at each measurement point, multiplying it by the corresponding weight coefficient, and then summing the results. For example, the weight coefficients can be allocated based on the sensor's measurement accuracy, with sensors having higher measurement accuracy corresponding to larger weights, or they can be allocated based on parameter sensitivity, with measurement points having higher sensitivity to the parameters to be identified corresponding to larger weights. Understandably, the form of the objective function is not limited to the weighted sum of squared residuals; it can also use absolute error, relative error, or other statistical indicators that can quantify the differences.

[0051] In step S52, the real-time intrinsic state parameter set is used as the variable to be optimized, and its value range is set. The variable to be optimized refers to the unknown parameters that need to be adjusted during the reverse solution process, that is, the real-time intrinsic state parameters that need to be identified at the current moment, which may include the current dielectric performance degradation of the insulation layer and the current increase in contact resistance. The value range refers to the range of values ​​that these variables to be optimized may exist in a physical sense. For example, the dielectric performance degradation should not be lower than zero or exceed the limit value when the material fails completely, and the increase in contact resistance should not be lower than the initial value or exceed the theoretical value of the open circuit state. The value range can be set based on material characteristics, historical experience, or theoretical calculations, or it can be set according to statistical data of similar joints. By setting the value range, the optimization process can be constrained to proceed within a reasonable physical space, avoiding the solution results from deviating from engineering reality.

[0052] In step S53, the adjoint state method or ensemble Kalman filter algorithm is used to iteratively adjust the variables to be optimized until the objective function converges, and the real-time intrinsic state parameter set at the current moment is output. The adjoint state method is an optimization algorithm based on gradient information. It obtains the gradient of the objective function on the variables to be optimized by solving the adjoint equation, thereby guiding the direction of parameter adjustment. The ensemble Kalman filter is a probabilistic recursive filtering algorithm that continuously corrects the variables to be optimized through two steps: state prediction and observation update. It is suitable for processing noisy monitoring data. Iterative adjustment refers to repeatedly executing the process of "calculating the objective function - determining whether convergence has occurred - adjusting the variables to be optimized" until the objective function value is less than a preset threshold or the change tends to stabilize. Convergence of the objective function means that the difference between the model output corresponding to the current variable to be optimized and the measured data has reached an acceptable range. The value of the variable to be optimized obtained at this time is the result of the reverse solution. It is understood that in addition to the adjoint state method and ensemble Kalman filter, other optimization algorithms can also be used, such as gradient descent, particle swarm optimization, and genetic algorithms. The specific choice can be determined according to the problem size and computational resources.

[0053] This application provides a concrete implementation path for inferring intrinsic state parameters from real-time monitoring data by decomposing the inverse solution process into three sub-steps: constructing an objective function, setting the range of values ​​for optimization variables, and iteratively solving using a specific algorithm. Specifically, the objective function is constructed with the goal of minimizing the weighted sum of squared residuals between the theoretical temperature distribution, theoretical electric field distribution, and measured temperature and partial discharge pulse data output by the whole-field distribution calculation sub-model. This clarifies the optimization direction of the inverse calculation and gives the model parameter adjustment process a clear convergence objective. Using the real-time intrinsic state parameter set as the variable to be optimized and setting its value range constrains the parameter variation boundaries during the solution process, preventing the inverse results from deviating from engineering reality. Iteratively adjusting the variable to be optimized using the adjoint state method or integrated Kalman filter algorithm until the objective function converges allows for the gradual approximation of the true state using redundant information in the monitoring data. The final output set of real-time intrinsic state parameters represents the estimation result with the highest degree of matching with the model under the current monitoring data conditions. This structured inverse solution method enables the coupled degradation model to adapt to changes in monitoring data under different operating conditions, improving the stability and reliability of the inverse results.

[0054] In some embodiments disclosed in this application, the real-time monitoring data in step S4 also includes vibration acceleration signals; in step S51, the objective function simultaneously considers the spatial distribution residual of the temperature field, the phase distribution characteristics of the partial discharge pulse, and the spectral energy distribution differences of the vibration signal.

[0055] Understandably, the real-time monitoring data collected in step S4 also includes vibration acceleration signals. Vibration acceleration signals refer to the physical quantity of mechanical vibration in the joint area changing over time, collected by an accelerometer; their unit is typically meters per second squared. Vibration acceleration signals can reflect the mechanical state of the joint's interior and connection points. For example, phenomena such as bolt loosening, interface fretting, and structural resonance will exhibit specific frequency characteristics in the vibration signal. For instance, vibration acceleration signals can be collected using piezoelectric accelerometers or microelectromechanical systems (MEMS) accelerometers. The sensor can be installed on the joint housing, connecting flange, or adjacent cable body. The sampling frequency setting needs to consider the frequency range of the vibration signal. For low-to-medium frequency vibrations caused by mechanical loosening, the sampling frequency can be set in the kilohertz range; for high-frequency vibrations caused by fretting wear, the sampling frequency may need to reach tens of kilohertz.

[0056] The objective function constructed in step S51 simultaneously considers the spatial distribution residual of the temperature field, the phase distribution characteristics of partial discharge pulses, and the spectral energy distribution differences of vibration signals. The spatial distribution residual of the temperature field refers to the difference between the theoretical temperature value output by the overall field distribution calculation sub-model and the measured temperature value at various spatial locations. It can be understood that the temperature distribution reflects the heating and cooling state of the joint and is directly related to the dielectric loss and contact resistance of the insulating material. The phase distribution characteristics of partial discharge pulses refer to the difference between the theoretically calculated phase distribution of partial discharge pulses and the actual acquired phase distribution. The phase distribution of partial discharge pulses can reflect the type and severity of insulation defects; for example, internal air gap discharge, surface discharge, and corona discharge exhibit different phase characteristics. The spectral energy distribution differences of vibration signals refer to the differences in energy distribution across different frequency bands between the theoretically calculated vibration spectrum and the actual acquired vibration spectrum. Spectral energy distribution can be obtained through Fourier transform or wavelet transform and can reflect the vibration intensity of different frequency components. For example, an increase in energy at a specific frequency band may be related to bolt loosening, while an increase in energy at high frequencies may be related to interface micro-motion. Incorporating these three differences into the objective function means that the reverse solution process requires simultaneously fitting monitoring data from three dimensions: temperature field, partial discharge characteristics, and vibration spectrum. For example, the objective function can be expressed as a superposition of three weighted residual squares: the temperature field residual term multiplied by a temperature weighting coefficient, the partial discharge phase residual term multiplied by a partial discharge weighting coefficient, and the vibration spectrum residual term multiplied by a vibration weighting coefficient. The weighting coefficients can be adjusted based on the measurement accuracy of each monitored quantity and its sensitivity to intrinsic state parameters. For instance, when the vibration signal is highly sensitive to changes in contact state, the weight of the vibration term can be appropriately increased.

[0057] This embodiment of the application expands the information dimensions upon which the inverse solution is based by further incorporating vibration acceleration signals into the real-time monitoring data and simultaneously considering the spatial distribution residuals of the temperature field, the phase distribution characteristics of partial discharge pulses, and the spectral energy distribution differences of vibration signals in the objective function. The spatial distribution residuals of the temperature field reflect the degree of consistency of the overall thermal state of the joint; the phase distribution characteristics of partial discharge pulses reflect the discharge characteristics of insulation defects; and the spectral energy distribution differences of vibration signals are related to the dynamic changes in the mechanical connection state. Incorporating these three different types of monitoring information with complementary physical meanings into the objective function creates multi-dimensional constraints on the real-time intrinsic state parameter set during the inverse solution process. This helps to distinguish the influence of different physical processes on the changes in intrinsic state parameters, reduces the non-uniqueness of the inversion results, and improves the identification accuracy of the real-time intrinsic state parameter set.

[0058] In some embodiments disclosed in this application, step S6, which involves simulating and calculating the remaining useful life based on the coupling degradation model, specifically includes: Using the current real-time intrinsic state parameter set as the initial value, and the preset future typical load curve and ambient temperature curve as input conditions, the electro-thermal-mechanical coupling degradation model is driven to perform time-domain progressive simulation. During the simulation, the key performance parameters output by the whole field distribution calculation sub-model are monitored in real time. The key performance parameters include at least the maximum field strength, the maximum temperature and the contact resistance value. When any key performance parameter reaches its preset failure threshold, the simulation stops, and the time from the start time to the stop time of the simulation is taken as the remaining service life.

[0059] This can be understood as "using the current real-time intrinsic state parameter set as the initial value" meaning that the values ​​in the current real-time intrinsic state parameter set obtained through reverse calculation in step S5 are used as the starting calculation state for subsequent time-domain simulation. This means that the real-time intrinsic state parameter set reflects the actual internal state of the joint at the current moment, including the attenuation of the dielectric properties of the insulation layer and the increase in contact resistance. Using this as the initial value for simulation calculation allows the prediction process to be based on the actual state of the joint rather than idealized assumptions. "Using the preset future typical load curve and ambient temperature curve as input conditions" refers to the external load conditions that need to be applied during the simulation process. The load curve reflects the change of line current over time, and the ambient temperature curve reflects the change of temperature of the medium surrounding the joint. For example, the future typical load curve can be predicted based on the statistical laws of historical load data or set according to the power grid dispatch plan; the ambient temperature curve can adopt the seasonal average temperature change law or be set according to weather forecast data.

[0060] "Time-domain progressive simulation using a driven electro-thermal-mechanical coupling degradation model" refers to treating time as an independent variable and progressively calculating the evolution of joint performance over time according to a certain time step. Time-domain progressive simulation can employ explicit or implicit time integration methods. Within each time step, the material state evolution sub-model is sequentially called to update the material state, material properties are transformed through parameter transfer functions, and then the temperature, electric, and stress field distributions are recalculated by the whole-field distribution calculation sub-model. This process is repeated until a stopping condition is met. During the simulation, key performance parameters output by the whole-field distribution calculation sub-model are monitored in real time. These parameters are macroscopic indicators that directly reflect whether the joint has reached a failure state. For example, key performance parameters include at least the maximum electric field strength, maximum temperature, and contact resistance. The maximum electric field strength refers to the maximum value of the electric field strength inside the joint, usually occurring in the insulation layer or interface region; excessively high field strength may lead to insulation breakdown. The maximum temperature refers to the maximum value of the temperature inside the joint; excessively high temperature may accelerate material aging or even trigger thermal runaway. The contact resistance reflects the conductivity of the connection interface; excessive resistance may lead to localized overheating and energy loss. The statement "Stop simulation when any key performance parameter reaches its preset failure threshold" refers to a critical value determined based on material properties, design standards, and operational experience. When a key performance parameter exceeds this value, the joint is considered unsafe to operate. For example, the failure threshold for the highest electric field strength can be set based on the breakdown electric field strength of the insulation material, considering a certain safety margin; the failure threshold for the highest temperature can be set based on the long-term temperature resistance rating of the insulation material; and the failure threshold for contact resistance can be set based on the design requirements of the connection interface and operational experience. It is understood that the failure thresholds for different key performance parameters may differ and may be adjusted as operating conditions change. The statement "The time from the start to the stop of the simulation is the remaining service life" refers to the current moment when the simulation begins, and the stop moment is the moment when any key performance parameter first reaches its failure threshold. The time interval between these two moments is the predicted remaining service life. It is understood that the predicted remaining service life is closely related to the input future load spectrum; different load curves and ambient temperature curves will lead to different prediction results. In practical applications, various possible future operating scenarios can be used for simulation to obtain the probability distribution or interval estimate of the remaining service life.

[0061] This application embodiment clarifies the specific implementation method of remaining lifetime prediction by using the real-time intrinsic state parameter set at the current moment as the starting point and the preset future typical load curve and ambient temperature curve as input conditions to drive the electro-thermal-mechanical coupling degradation model to perform time-domain progressive simulation. Using the real-time intrinsic state parameter set as the simulation starting point ensures that the prediction process is based on the current real state of the joint, rather than an assumed ideal state. Using the preset future typical load curve and ambient temperature curve as input conditions allows the simulation process to reflect the actual working conditions the joint will experience in future operation. During the simulation, key performance parameters output by the whole-field distribution calculation sub-model are monitored in real time, including the maximum field strength, maximum temperature, and contact resistance value. These parameters are directly related to the insulation performance and connection reliability of the joint. The simulation stops when any key performance parameter reaches its preset failure threshold, and the time length from the start time to the stop time of the simulation is taken as the remaining lifetime. This ensures that the lifetime prediction results directly correspond to specific failure criteria, providing a quantifiable basis for operation and maintenance decisions.

[0062] In some embodiments disclosed in this application, step S7 is also included: S7: When the insulation performance degradation level determined in step S6 reaches the preset warning level, based on the abnormal parameter components in the real-time intrinsic state parameter set and combined with the inverse function of the mapping model, the original process parameter deviation range that caused the abnormal parameter components is traced back in reverse, and a process optimization report containing specific improvement suggestions is generated.

[0063] It is understandable that "the insulation performance degradation level determined in step S6 reaches the preset warning level" means that the degradation level calculated in step S6 has exceeded or equaled the preset threshold requiring attention or intervention. The warning level can be set according to the operation and maintenance strategy; for example, it can be divided into multiple levels such as attention level, warning level, and danger level, or a single unified warning threshold can be set. It is understandable that the setting of the warning level needs to consider factors such as the criticality of the joint, the operating environment, and historical statistical data; different classification standards can be adopted for different application scenarios. "Based on abnormal parameter components in the real-time intrinsic state parameter set" means identifying parameter indicators that deviate from the normal range from the real-time intrinsic state parameter set calculated in step S5. The judgment of abnormal parameter components can be based on the comparison results with the initial intrinsic state parameter set, or on the comparison with statistical data of similar joints. For example, when the current contact resistance increase exceeds a certain threshold, it can be considered an abnormal component; when the current dielectric performance degradation of the insulation layer exceeds a certain limit, it can also be considered an abnormal component. An abnormal parameter component can be a single parameter or a combination of multiple parameters, depending on the actual degradation mode. "Reverse tracing using the inverse function of the mapping model" refers to using the inverse operation relationship of the mapping model established in step S2 to map abnormal changes in the intrinsic state parameter space back to the original process parameter space. The inverse function of the mapping model mathematically represents a mapping relationship from output to input. However, since mapping models are usually complex and may have multiple solutions, reverse tracing often requires combining optimization algorithms or probabilistic statistical methods. For example, a Bayesian inversion method can be used to obtain the most probable range of process parameter deviations by calculating the posterior probability distribution of the original process parameters under given abnormal intrinsic state parameters.

[0064] "The original process parameter deviation range causing the abnormal parameter component" refers to the range of deviations between the process parameters in the production or installation process that are causally related to the current abnormal state and the design or standard values, obtained through reverse tracing calculations. The process parameter deviation range can include numerical deviations of specific parameters, such as an extrusion temperature that is 20 degrees Celsius higher or a bolt tightening torque that is 5 Nm lower, or deviations in the temporal characteristics of the process, such as a slow torque increase rate or excessive pressure fluctuations. It is understood that determining the deviation range requires consideration of the uncertainties in the reverse tracing process and is usually given as an interval estimate rather than a single, definitive value. "Generating a process optimization report containing specific improvement suggestions" refers to outputting the results obtained from reverse tracing in report form, containing improvement suggestions for reference in the production or installation process. Improvement suggestions can be based on the deviation range and, for example, can include adjusting the extrusion temperature setpoint, optimizing the screw speed curve, increasing the frequency of torque verification for specific bolts, and correcting the torque application rate. The process optimization report can be in the form of a text report, data file, or visualization chart, and different formats can be used depending on the application scenario. The target users of the report can be process engineers on the production line or operators on the installation site, and the recommendations should be actionable.

[0065] This application embodiment establishes a causal relationship between operational anomalies and manufacturing / installation processes by, when the insulation performance degradation level reaches a preset warning level, using abnormal parameter components in the real-time intrinsic state parameter set and the inverse function of a mapping model to trace back the original process parameter deviation range leading to the abnormal parameter component, and generating a process optimization report containing specific improvement suggestions. The abnormal parameter component is a core indicator reflecting current performance degradation, and its numerical changes directly point to the degree of degradation of a specific physical mechanism. By tracing back using the inverse function of the mapping model, abnormal changes in the intrinsic state parameter space can be mapped back to the original process parameter space, allowing the calculation of the process parameter deviation range that may have caused the abnormal change. Generating a process optimization report with specific improvement suggestions extends the monitoring results beyond the state assessment itself to the process improvement stage, providing data support for continuous optimization of the production and installation process. This reverse tracing mechanism from operational anomalies to process deviations gives the monitoring system diagnostic capabilities, enabling it to feed back accumulated experience from operation to the front-end manufacturing process.

[0066] In some embodiments disclosed in this application, step S7, reverse tracing specifically includes: if the current contact resistance increase in the real-time intrinsic state parameter set exceeds a preset threshold, the corresponding assembly torque deviation value and torque application rate deviation value are calculated through the inverse function of contact mechanics in the mapping model, and the bolt position where the deviation occurs and the suggested correction value are indicated in the process optimization report.

[0067] It is understandable that "the current contact resistance increase in the real-time intrinsic state parameter set exceeds the preset threshold" means that the increase in contact resistance, calculated in step S5 and reflecting the current contact interface state, has exceeded the preset threshold of concern. The preset threshold can be set according to the joint's design requirements and operational experience. For example, it can be set by referring to historical statistical data of similar joints, or it can be determined based on the sensitivity analysis of the contact resistance's influence on the temperature field. It is understandable that the increase in contact resistance is a key indicator reflecting the degradation of the connection interface state, and its increase may originate from various physical mechanisms such as fretting wear, oxidation corrosion, and stress relaxation. "Through the inverse function of contact mechanics in the mapping model" means using the inverse operation relationship of the contact mechanics-based mapping model established in step S2 to map the intrinsic state parameter of contact resistance increase back to the original assembly parameter space. Mathematically, the inverse function of contact mechanics represents the inverse mapping relationship from contact state parameters to assembly process parameters. Its establishment requires a comprehensive analysis based on Hertzian contact theory, elastoplastic deformation models, and interface conductivity mechanisms. For example, the inverse function can be obtained by numerically inverting the forward mapping model, or it can be obtained by fitting using data-driven methods such as neural networks. Understandably, due to the complexity and nonlinearity of the contact process, the inversion results are usually given in the form of a probability distribution or a range.

[0068] "Calculating the corresponding assembly torque deviation and torque application rate deviation" refers to deducing the specific values ​​of the original assembly parameters that deviate from the design values, causing the current increase in contact resistance, through the inverse function of contact mechanics. The assembly torque deviation reflects the difference between the applied torque during tightening and the standard value. Insufficient torque may lead to insufficient interface pressure, while excessive torque may cause thread damage or overstress. The torque application rate deviation reflects the difference between the rate of torque increase during tightening and the standard process requirements. An excessively fast application rate may lead to unstable friction coefficients, while an excessively slow application rate may affect assembly efficiency. For example, the inversion calculation may conclude that the torque of a bolt is 5 to 8 Nm lower than the standard value, and the application rate is within 0.2 seconds faster than the standard value. "Indicating the bolt location where the deviation occurs and the suggested correction value in the process optimization report" means outputting the results obtained from the inversion in an operable form. Bolt location indication can be achieved through bolt numbers, spatial coordinates, or schematic diagrams, enabling maintenance personnel to accurately locate the tightening points requiring attention. Suggested correction values ​​may include the target torque range, the suggested application rate, and information on adjacent bolts that may require re-inspection. For example, the report may include the statement that “the tightening torque of bolt #3 is 5 to 8 Nm too low. It is recommended to retighten it to the standard torque of 120 Nm during the next maintenance, and to be careful to control the application rate within 0.5 to 0.8 seconds.”

[0069] This application embodiment further clarifies the reverse tracing execution method by addressing the specific scenario of abnormal contact resistance increases. When the current contact resistance increase in the real-time intrinsic state parameter set exceeds a preset threshold, the corresponding assembly torque deviation value and torque application rate deviation value are calculated using the inverse contact mechanics function in the mapping model. The process optimization report then indicates the bolt location where the deviation occurred and the suggested correction value. The inverse contact mechanics function establishes a quantitative mapping relationship between the contact resistance increase and the original assembly parameters, allowing abnormal changes in contact resistance to be accurately attributed to specific assembly parameter deviations. The calculated torque deviation value and application rate deviation value provide a quantitative basis for process improvement, while the bolt location indication enables maintenance personnel to precisely pinpoint the source of the problem. The proposed correction value further enhances the operability of the process optimization report, enabling monitoring results to directly guide on-site operations.

[0070] In some embodiments disclosed in this application, in step S2, the assembly parameters include the torque time-series curve and the rotation angle time-series curve during the bolt tightening process; the mapping model is established based on contact mechanics and is used to calculate the pressure distribution of the contact interface and the actual contact area of ​​the interface according to the torque time-series curve and the rotation angle time-series curve.

[0071] It is understandable that assembly parameters include torque and rotation time-series curves during bolt tightening. The torque time-series curve is a continuous record of the torque value over time from the start of tightening to completion, reflecting the establishment and fluctuation of torque over time. The rotation time-series curve is a continuous record of the bolt angle over time from the start of tightening to completion, reflecting the change in rotation angle over time. It is understandable that using time-series curves, rather than a single final torque or final angle, allows for a more complete recording of the dynamic characteristics of the tightening process. Information such as the rate of torque increase, the presence of sudden torque changes, and the correspondence between rotation angle and torque can all be extracted from the time-series curve. For example, the torque time-series curve can be acquired in real-time using the torque sensor and timer built into the intelligent torque wrench, with a sampling frequency set between 100 Hz and 1000 Hz; the rotation time-series curve can be acquired synchronously using an encoder or angle sensor. "The mapping model is based on contact mechanics" means that the model is constructed using the fundamental theories of contact mechanics as its mathematical foundation. Contact mechanics studies the stress, deformation, and contact area distribution that occur when two solid surfaces come into contact. Its theoretical foundations include Hertzian contact theory, elastoplastic mechanics, and tribology. A mapping model based on contact mechanics can describe how torque is converted into axial preload during bolt tightening, and how this axial preload forms a pressure distribution at the contact interface. For example, the model may include the torque-preload relationship for threaded pairs, the pressure distribution calculation formula for flange interfaces, and the relationship between microscopic contact area and nominal pressure. Understandably, due to the complexity of the contact process, mapping models typically require parameter correction through finite element simulation or experimental calibration.

[0072] "Based on the torque and rotation time-series curves," the pressure distribution and actual contact area of ​​the interface are calculated. The pressure distribution refers to the magnitude of the normal pressure exerted at various locations on the connection interface. Since the interface is not perfectly flat, the pressure distribution is usually non-uniform, with higher pressure near the bolt holes and lower pressure further away. The actual contact area refers to the total area of ​​the microscopic protrusions that actually make contact on the interface. Compared to the nominal contact area, the actual contact area is usually much smaller, and its size directly affects the values ​​of contact resistance and contact thermal resistance. For example, the calculation process may include: first, estimating the axial preload generated by the bolt based on the torque and rotation time-series curves, combined with thread parameters and the coefficient of friction; then, calculating the nominal pressure distribution on the interface based on the axial preload, the geometry of the connector, and its stiffness characteristics; finally, converting the nominal pressure distribution into a spatial distribution of the actual contact area based on surface roughness parameters and a contact mechanics model.

[0073] This application embodiment concretizes assembly parameters into torque and rotation time-series curves during bolt tightening, and uses a mapping model based on contact mechanics to calculate the pressure distribution and actual contact area of ​​the contact interface, clarifying the quantitative transformation path from original assembly data to interface contact state. The torque and rotation time-series curves record complete dynamic information of the tightening process, and compared to a single final torque, they can more comprehensively reflect the establishment process of bolt preload and potential abnormal working conditions. The mapping model based on contact mechanics uses this time-series data, combined with bolt specifications, friction conditions, and other factors, to calculate the pressure distribution and actual contact area of ​​the contact interface. These two parameters are directly related to the initial state of contact resistance and contact thermal resistance. In this way, the mechanical assembly data during the installation phase is transformed into initial intrinsic state parameters with clear physical meaning, providing an initial basis for the evolution analysis of interface performance in the subsequent coupled degradation model, and also providing data support for the quantitative assessment of installation quality and anomaly tracing.

[0074] In some embodiments disclosed in this application, in step S2, the process parameters include the die head temperature time-series curve, the screw speed time-series curve, and the extrusion pressure time-series curve during the insulation extrusion process; the mapping model is established based on material rheology and is used to calculate the residual stress distribution of the insulation layer according to the die head temperature time-series curve, the screw speed time-series curve, and the extrusion pressure time-series curve.

[0075] It is understandable that the process parameters include the die head temperature time-series curve, screw speed time-series curve, and extrusion pressure time-series curve during the insulation extrusion process. The die head temperature time-series curve is a continuous record of the temperature change at the extruder head over time, reflecting the thermal history of the insulating material during extrusion molding. The screw speed time-series curve is a continuous record of the screw speed change over time, reflecting the shear rate and residence time experienced by the material within the barrel. The extrusion pressure time-series curve is a continuous record of the melt pressure change at the die head over time, reflecting the material flow resistance and extrusion stability. It is understandable that using time-series curves instead of a single process setpoint allows for a more complete recording of the dynamic characteristics of the extrusion process; for example, information such as temperature fluctuations, screw speed stability, and pressure pulsations can all be extracted from the time-series curves. For example, this time-series data can be recorded in real-time by the extrusion production line control system at a sampling frequency of several to tens of times per second and stored in association with the connector's identification code.

[0076] "The mapping model is based on materials rheology," meaning that the model is constructed using the fundamental theories of materials rheology as its mathematical foundation. Materials rheology studies the flow and deformation of materials under force. For polymeric insulating materials, their rheological behavior directly affects the molecular chain orientation, residual stress formation, and interfacial bonding state during extrusion molding. A mapping model based on materials rheology can describe how temperature, shear rate, and pressure affect the material's flow characteristics and curing behavior during extrusion. For example, the model may include constitutive equations describing material viscosity, Arrhenius equations considering temperature dependence, and viscoelastic models describing molecular chain relaxation processes. Understandably, since insulating materials are typically complex polymer systems, the establishment of the mapping model requires parameter fitting using rheological experimental data, and numerical simulation methods can also be used for auxiliary verification. Residual stress in the insulating layer refers to the stress state remaining inside the material during extrusion molding and subsequent cooling and curing due to factors such as uneven thermal shrinkage and hindered molecular chain orientation relaxation. Residual stress distribution refers to the magnitude and direction of residual stress at different locations in the insulating layer, typically exhibiting a non-uniform distribution along the wall thickness direction and along the circumferential direction. For example, the calculation process may include: first, calculating the flow velocity field and shear stress field of the melt in the die head based on the temperature time series curve and the screw speed time series curve, combined with the rheological properties of the material; then, determining the pressure boundary conditions at the die orifice based on the extrusion pressure time series curve; and finally, simulating the transformation process of the material from molten to solid state through thermo-mechanical coupling analysis during the cooling and solidification process, and calculating the stress distribution that ultimately remains inside the insulating layer.

[0077] This application embodiment specifies the process parameters as the time-series curves of die head temperature, screw speed, and extrusion pressure during the insulation extrusion process. A mapping model based on materials rheology is used to calculate the residual stress distribution of the insulation layer, clarifying the transformation path from the original process data to the initial state of the material. The temporal changes in die head temperature, screw speed, and extrusion pressure directly affect the material's flow behavior, cooling rate, and molecular chain orientation during extrusion. These factors collectively determine the formation of residual stress within the insulation layer. The mapping model based on materials rheology can describe the rheological properties of the material under high temperature and shear conditions, as well as the stress evolution during cooling and solidification, transforming the original process time-series data into quantitative results of residual stress distribution. As one of the initial intrinsic state parameters, the residual stress of the insulation layer is a crucial factor affecting the subsequent electrothermal aging behavior of the material. The residual stress distribution obtained in this way has higher individual specificity and accuracy compared to empirical estimates.

[0078] In some embodiments disclosed in this application, step S6, determining the insulation performance degradation level specifically includes: Calculate the relative degradation of the current offset of each parameter relative to its initial value; Calculate the average degradation rate of each parameter within a preset time window; Using relative degradation degree and average degradation rate as input features, a pre-trained health status classification model is invoked to output the insulation performance degradation level of the high-voltage cable joint.

[0079] It is understandable that "the relative degradation degree of the current offset of each parameter relative to its initial value" refers to the dimensionless ratio obtained by dividing the difference between the real-time intrinsic state parameter calculated in step S6 and the initial intrinsic state parameter by the value of the initial intrinsic state parameter. It is also understandable that the relative degradation degree eliminates the differences in dimensions and orders of magnitude between different parameters, allowing parameters with different physical units, such as the attenuation of the dielectric properties of the insulating layer and the increase in contact resistance, to be compared and integrated on the same scale. For example, if the initial value of a parameter is A0 and the current offset is ΔA, then the relative degradation degree can be expressed as ΔA / A0, or as a percentage of ΔA / A0×100%. The relative degradation degree reflects the cumulative change of the parameter from its initial state to its current state; a larger value indicates more severe degradation. "The average degradation rate of each parameter within a preset time window" refers to performing time-series analysis on the current offset calculated at multiple consecutive sampling times to fit the average change in offset per unit time. The preset time window can be set according to monitoring needs and parameter change characteristics. For example, it can be set to the most recent week, the most recent month, or the most recent quarter, and the window length can be dynamically adjusted according to the rate of parameter change. The average degradation rate can be calculated by linear fitting to find the slope, or by time series analysis methods such as exponential smoothing or moving average. For example, if an offset sequence ΔA1, ΔA2...ΔAn is obtained at n time points within the time window, with a corresponding time interval of Δt, the average degradation rate can be calculated as the difference between ΔAn and ΔA1 divided by the total time, or the slope can be obtained by least squares fitting. The average degradation rate reflects the trend of parameter change; the larger the value, the faster the performance degradation.

[0080] "Using relative degradation degree and average degradation rate as input features" means combining the calculated relative degradation degree and average degradation rate of each parameter into a multi-dimensional feature vector, which serves as the input to the subsequent classification model. The dimensionality of the input features is determined by the number of selected parameters. For example, if two parameters are selected: the attenuation of dielectric properties of the insulation layer and the increase in contact resistance, the input features can include the relative degradation degree and average degradation rate of these two parameters, for a total of four dimensions. It is understandable that the selection and combination of input features can be adjusted according to actual needs, and other derived features such as acceleration and volatility can also be considered. "Calling a pre-trained health status classification model" means using a machine learning model that has been trained on historical data to classify and predict the new input feature vector. The health status classification model can be any of the supervised learning algorithms such as support vector machines, random forests, gradient boosting trees, and neural networks. Model training relies on labeled historical data, where the labels are known insulation performance degradation levels, and the features are the relative degradation degree and average degradation rate at the corresponding time. For example, training data can come from historical fault cases, accelerated aging test data, or high-fidelity simulation data. The degradation level corresponding to each sample is determined through manual annotation or fault records. The model training process involves finding the optimal mapping function from the feature space to the level label. "The insulation performance degradation level of the output high-voltage cable joint" refers to the degradation level category of the joint calculated by the classification model based on the input feature vector. The degradation level can be divided according to operation and maintenance needs. For example, it can be divided into four levels: normal, slight degradation, moderate degradation, and severe degradation, or into four levels: normal, attention, warning, and danger. Each level corresponds to a different maintenance strategy. For example, the normal level can continue routine monitoring, the attention level requires increased monitoring frequency, the warning level requires planned maintenance, and the danger level requires immediate shutdown.

[0081] This application embodiment concretizes the process of determining the insulation performance degradation level by calculating the relative degradation degree, calculating the average degradation rate, constructing a health state feature vector, and calling the health state classification model to output the level, thus clarifying the quantitative path from parameter offset to degradation level. The relative degradation degree reflects the degree of each parameter's current offset relative to its initial value, reflecting the cumulative effect of performance degradation; the average degradation rate reflects the speed trend of performance degradation. Combining the two provides a more comprehensive description of the degradation state. A health state feature vector is constructed using the relative degradation degree and average degradation rate as input features, integrating the degradation information of multiple parameters into a unified data expression form. The pre-trained health state classification model is called to output the insulation performance degradation level, giving the degradation level determination process a data-driven objective basis. The classification model can establish a mapping relationship between the feature vector and the level based on historical fault cases and simulation data. This quantitative evaluation method makes the determination process of insulation performance degradation level repeatable and consistent, reducing the uncertainty caused by subjective judgment.

[0082] In some embodiments disclosed in this application, step S8 is also included: S8: Establish a digital archive of high-voltage cable joints based on the identification code. The digital archive includes at least the initial intrinsic state parameter set, the historical intrinsic state parameter sets of each real time, the historical record of insulation performance degradation level, and the remaining service life prediction curve, and provides a visual query interface.

[0083] It is understandable that "establishing a digital archive of a high-voltage cable joint based on the identification code" refers to organizing various types of data related to the joint according to a certain structure, using the identification code configured in step S1 as a unique index, to form a centrally stored information set. The digital archive can be understood as a summary of various types of data generated throughout the entire lifecycle of the joint, from production to operation. Its organization can adopt any of the following forms: relational database, time-series database, or distributed file system. It is understood that the identification code, as a consistent index key, ensures that data of different stages and types can be accurately attributed to the same joint object, avoiding information misalignment or loss. For example, the digital archive can be stored on a cloud server or deployed in a local data center storage system. The digital archive includes at least the initial intrinsic state parameter set, the historical sets of real-time intrinsic state parameters, the historical record of insulation performance degradation levels, and the remaining service life prediction curve. The initial intrinsic state parameter set refers to a set of parameters output by the mapping model in step S2 that reflects the initial performance characteristics of the joint, including the residual stress distribution of the insulation layer, the interface roughness of the semiconducting layer, and the pressure distribution of the contact interface. These parameters record the initial state of the joint before commissioning. The historical intrinsic state parameter set refers to the record set of real-time intrinsic state parameters calculated in step S5 at different times through reverse solving. The results of each calculation are stored in chronological order, reflecting the performance evolution trajectory of the connector during operation. It is understandable that the storage frequency of the historical intrinsic state parameter set can be set according to monitoring needs; it can be stored once per minute, once per hour, or once per day, depending on the data volume and application scenario.

[0084] The insulation performance degradation level history record refers to the time-series record of the insulation performance degradation level determined in step S6 at different times, reflecting the change process of the joint's health status. The degradation level history record can correspond to the timestamp of the real-time intrinsic state parameter set, facilitating the tracing of the correlation between the state and the level at a specific moment. The remaining service life prediction curve refers to the trend line of the remaining service life changing over time based on the simulation calculation in step S6. It can be a two-dimensional curve with time as the horizontal axis and remaining service life as the vertical axis, or it can be a sequence of remaining service life estimates generated at different prediction time points. For example, the remaining service life prediction curve can include a single prediction curve based on the current operating condition, or it can include prediction intervals or probability distribution bands considering different operating conditions. "Providing a visual query interface" refers to presenting the data stored in the digital archive to maintenance personnel in an intuitive way through a graphical user interface. The visual query interface can use a web-based browser interface or a dedicated client application. Interface functions can include data query, trend analysis, comparison display, report generation, etc. For example, users can enter an identification code or select a specific connector through the interface to bring up the connector's digital profile homepage. The homepage can display a summary of the current health status, the predicted remaining lifespan, and recent trend graphs of key parameters. Users can also choose to view historical data to compare parameter changes over different periods and analyze the evolution of performance degradation.

[0085] This application embodiment establishes a digital archive for high-voltage cable joints based on identification codes, incorporating initial intrinsic state parameter sets, historical real-time intrinsic state parameter sets, historical insulation performance degradation levels, and remaining service life prediction curves. A visual query interface is also provided, enabling centralized management and intuitive presentation of monitoring data throughout the entire lifecycle. The identification code serves as a unique index throughout the entire process, ensuring that data generated at different stages can be accurately attributed to the same joint, avoiding information misalignment or loss. The initial intrinsic state parameter set records the initial performance characteristics of the joint before commissioning; the historical real-time intrinsic state parameter sets reflect the dynamic evolution trajectory of performance during operation; the historical insulation performance degradation level records the changes in health status; and the remaining service life prediction curve provides a prediction of future trends. Through the visual query interface, maintenance personnel can intuitively view the above information graphically, compare parameter changes at different times, analyze the evolutionary patterns of performance degradation, and provide data support for maintenance decisions.

[0086] Please refer to Figure 2. Figure 2 shows an embodiment of this application providing an online monitoring system for the production cycle of high-voltage cable joints. The system includes: The identification code management unit 21 is used to generate and assign identification codes to high-voltage cable joints and establish the association between the identification codes and the full-cycle data. The process data acquisition unit 22 is used to acquire the timing data of insulation extrusion process parameters and bolt fastening assembly parameters associated with the identification code from the extrusion production line control system and intelligent assembly tools. The parameter parsing and modeling unit 23 has a built-in mapping model, which is used to calculate and output the initial intrinsic state parameter set based on the process parameter time series data and assembly parameter time series data, and to construct an electro-thermal-mechanical coupling degradation model based on the initial intrinsic state parameter set. The operation monitoring data acquisition unit 24 is deployed at the high-voltage cable joint site to collect temperature, partial discharge pulse, load current and vibration acceleration data associated with the identification code in real time; The state inversion calculation unit 25 is connected to the operation monitoring data acquisition unit 24 and the parameter analysis and modeling unit 23. It is used to drive the electro-thermal-mechanical coupling degradation model to perform inverse solution with real-time monitoring data as input, and calculate and output the real-time intrinsic state parameter set. The health assessment and life prediction unit 26 is used to calculate the current offset and rate of change of each parameter based on the initial intrinsic state parameter set and the real-time intrinsic state parameter set, determine the insulation performance degradation level, and simulate and calculate the remaining service life. The source tracing diagnosis and process feedback unit 27 is used to perform reverse tracing based on the abnormal parameter components in the real-time intrinsic state parameter set and the inverse function of the mapping model when the insulation performance degradation level reaches the warning level, and generate and output a process optimization report. The data storage and visualization unit 28 is used to store full-cycle data associated with the identification code and to provide a visual query interface for digital archives.

[0087] Understandably, the identification code management unit 21 is used to generate and assign identification codes to high-voltage cable joints and establish the association between the identification codes and the full-cycle data. The identification codes can be generated using a random number generation algorithm or a coding rule based on timestamps and serial numbers, ensuring that each joint obtains a unique identifier. Identification codes can be assigned physically to the joint body at the beginning of the production process through methods such as printing labels, affixing RFID electronic tags, or laser engraving. Establishing the association means storing subsequently collected production data, installation data, and operational data using this identification code as an index, enabling data generated at different stages to be quickly retrieved and integrated through the identification code. Understandably, the identification code management unit 21 is the foundation of the entire system's data traceability, and its stability and uniqueness directly affect the reliability of the full-cycle data association. The process data acquisition unit 22 is used to collect time-series data of insulation extrusion process parameters and bolt tightening assembly parameters associated with the identification codes from the extrusion production line control system and intelligent assembly tools. The extrusion production line control system typically employs a programmable logic controller (PLC) or an industrial computer. The process data acquisition unit 22 can read data from these devices via communication protocols such as industrial Ethernet, OPC unified architecture, or Modbus. Intelligent assembly tools include intelligent torque wrenches, assembly robots, etc., and these devices typically possess data recording and communication capabilities. For example, the process data acquisition unit 22 can be deployed at the edge of the production workshop, aggregating data from multiple devices through an industrial gateway, and storing it in association with identification codes after time synchronization and format standardization.

[0088] The parameter analysis and modeling unit 23 has a built-in mapping model used to calculate and output the initial intrinsic state parameter set based on the time-series data of process parameters and assembly parameters, and to construct an electro-thermal-mechanical coupling degradation model based on the initial intrinsic state parameter set. The mapping model can be implemented using commercial numerical calculation software or a self-developed solver. The process of calculating the initial intrinsic state parameter set may require calling various calculation methods such as finite element analysis, boundary element method, or analytical formulas. Constructing the coupling degradation model requires generating a computational mesh based on the geometry and material properties of the joint, setting initial boundary conditions, and integrating the material state evolution sub-model, the whole-field distribution calculation sub-model, and the parameter transfer function into a unified simulation framework. Understandably, the parameter analysis and modeling unit 23 has high requirements for computing resources and can be implemented using high-performance workstations or cloud computing resources. The operation monitoring data acquisition unit 24 is deployed at the high-voltage cable joint site to collect temperature, partial discharge pulse, load current, and vibration acceleration data associated with the identification code in real time. This unit 24 typically consists of sensors, signal conditioning circuits, data acquisition modules, and communication modules. Temperature sensors can be thermocouples, resistance temperature detectors (RTDs), or fiber Bragg grating sensors; partial discharge sensors can be high-frequency current transformers or capacitively coupled sensors; current sensors can be Hall effect sensors or Rogowski coils; and vibration sensors can be piezoelectric accelerometers or microelectromechanical system (MEMS) accelerometers. The signal conditioning circuit filters, amplifies, and converts the raw signal from analog to digital. The data acquisition module collects data at a preset sampling frequency and adds timestamps. The communication module uploads the data to the data center via wired or wireless means. For example, the sampling frequency can be set according to the signal type; temperature data can be collected once per minute, while partial discharge and vibration data may require sampling rates of thousands to tens of thousands of times per second.

[0089] The state inversion calculation unit 25 is connected to the operation monitoring data acquisition unit 24 and the parameter analysis and modeling unit 23. It uses real-time monitoring data as input to drive the electro-thermal-mechanical coupling degradation model to perform inverse solving, calculating and outputting the real-time intrinsic state parameter set. This unit needs to run the inverse solving algorithm, including constructing the objective function, setting the range of optimization variables, and using algorithms such as the adjoint state method or integrated Kalman filter for iterative calculation. The inverse solving process has certain requirements for computing resources, but also requires high response time. Therefore, the state inversion calculation unit 25 can be deployed on a field edge computing gateway to quickly complete the inversion calculation after data acquisition and upload the results to the cloud or data center. Understandably, the frequency of inverse solving can be set according to monitoring needs, either once per minute or once per hour, depending on the rate of parameter change and the limitations of computing resources. The health assessment and life prediction unit 26 is used to calculate the current offset and rate of change of each parameter based on the initial intrinsic state parameter set and the real-time intrinsic state parameter set, determine the insulation performance degradation level, and simulate the remaining service life. This unit first performs parameter comparison to calculate the offset and rate of change. Then, it determines the degradation level according to preset rules or classification models. Finally, it performs time-domain progressive simulation to predict the remaining lifetime, starting from the current state. The remaining lifetime simulation requires calling the coupled degradation model for long-term calculations, which involves a large amount of computation. This can be achieved using high-performance cloud computing resources, and the prediction results can be updated periodically as needed.

[0090] The traceability diagnosis and process feedback unit 27 is used to generate and output a process optimization report when the insulation performance degradation level reaches the warning level, based on the abnormal parameter components in the real-time intrinsic state parameter set and combined with the inverse function of the mapping model for reverse tracing. This unit first identifies abnormal parameter components, such as when the contact resistance increases beyond a threshold. Then, it calls the inverse function of the mapping model for reverse tracing, which can be obtained by numerically inverting the forward mapping model or by fitting it with a neural network. Finally, it generates a process optimization report based on the inversion results, which includes the specific stage where the deviation occurred, the range of the deviation value, and suggested corrective measures. For example, the report can be pushed to responsible personnel in the production or installation stages via email, SMS, or system messages. The data storage and visualization unit 28 is used to store full-cycle data associated with the identification code and provides a visual query interface for the digital archive. This unit uses a database system to store structured data and a file system or object storage to store unstructured data such as raw waveforms. The visual query interface is implemented based on Web technology, supporting the retrieval of joint information by identification code and displaying initial intrinsic state parameters, historical curves of real-time intrinsic state parameters, degradation level trends, and remaining life prediction results in chart form. The interface also provides a comparative analysis function, which supports selecting data from multiple connectors for horizontal comparison, or selecting different time periods for vertical comparison of the same connector.

[0091] This embodiment integrates an identification code management unit 21, a process data acquisition unit 22, a parameter analysis and modeling unit 23, an operation monitoring data acquisition unit 24, a state inversion calculation unit 25, a health assessment and life prediction unit 26, a traceability diagnosis and process feedback unit 27, and a data storage and visualization unit 28 to construct an online monitoring and life prediction system covering the entire life cycle of high-voltage cable joints. The identification code management unit 21 establishes a unique identifier for each joint and associates it with full-cycle data, laying the foundation for data traceability. The process data acquisition unit 22 obtains the time-series process parameters and assembly parameters associated with the identification code from the extrusion production line control system and intelligent assembly tools, ensuring accurate acquisition of front-end data. The parameter analysis and modeling unit 23 incorporates a mapping model, transforming the original process data into an initial intrinsic state parameter set and constructing a coupled degradation model, realizing the transformation of manufacturing information into a simulation model. The operation monitoring data acquisition unit 24 is deployed on-site, acquiring real-time data on temperature, partial discharge, load current, and vibration acceleration, providing input for state inversion. The state inversion calculation unit 25 uses real-time monitoring data to drive the coupled degradation model for inverse solving, outputting the real-time intrinsic state parameter set at the current moment. The health assessment and lifetime prediction unit 26 determines the insulation performance degradation level and simulates the remaining service life based on the comparison of initial and real-time intrinsic state parameters. The source tracing diagnosis and process feedback unit 27, when the degradation level reaches the warning level, uses the inverse function of the mapping model to trace process parameter deviations in reverse and generate an optimization report. The data storage and visualization unit 28 centrally stores the entire lifecycle data and provides a visual query interface. Through the collaborative work of these units, the system achieves a complete functional closed loop from data acquisition, state inversion, health assessment to source tracing diagnosis.

[0092] Please refer to 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 memory 304 stores computer programs, including program instructions. The processors 301 execute the program instructions stored in the memory 304. Specifically, the processors 301 are configured to invoke the program instructions to execute the online monitoring method for the production cycle of the high-voltage cable connector.

[0093] 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. A general-purpose processor may be a microprocessor or any conventional processor.

[0094] 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.

[0095] 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.

[0096] 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 online monitoring method for the production cycle of high-voltage cable joints 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.

[0097] In another embodiment of this application, an electronic device is provided. The electronic device 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 online monitoring method for the production cycle of high-voltage cable joints described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in an electronic device, 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. Computer-readable media can include any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0098] 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.

[0099] 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.

[0100] 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.

[0101] 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.

[0102] 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.

[0103] 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.

[0104] 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 online monitoring of the production cycle of high-voltage cable joints, characterized in that, Includes the following steps: S1: Configure an identification code for the high-voltage cable joint, and collect the process parameters of the joint in the insulation extrusion process and the assembly parameters in the bolt fastening process based on the identification code; S2: Input the process parameters and assembly parameters into a pre-established mapping model, and output the initial intrinsic state parameter set of the joint from the mapping model. The initial intrinsic state parameter set includes the residual stress distribution of the insulating layer and the interface roughness of the semiconductive layer calculated from the process parameters, as well as the contact interface pressure distribution calculated from the assembly parameters. S3: Establish an electro-thermal-mechanical coupled degradation model for the high-voltage cable joint. The coupled degradation model uses the initial intrinsic state parameter set as the initial boundary condition and integrates the electrothermal aging sub-model of the insulating material and the oxidation resistance-increasing sub-model of the contact interface. S4: Collect real-time monitoring data of the high-voltage cable joint under operating conditions. The real-time monitoring data includes at least the joint body temperature, partial discharge pulse, and load current. S5: Substitute the real-time monitoring data into the coupling degradation model for inverse solution, and calculate the real-time intrinsic state parameter set at the current moment through an iterative algorithm. The real-time intrinsic state parameter set includes the current dielectric performance attenuation of the insulation layer and the current contact resistance increase. S6: Compare each parameter in the real-time intrinsic state parameter set with each parameter in the initial intrinsic state parameter set one by one, and calculate the current offset of each parameter. Based on the calculated values ​​of the current offset at multiple consecutive sampling times, the rate of change of each parameter is fitted; based on the current offset and the rate of change, the insulation performance degradation level of the high-voltage cable joint is determined, and its remaining service life is calculated based on the coupled degradation model.

2. The method according to claim 1, characterized in that, In step S3, the electro-thermal-mechanical coupled degradation model includes a material state evolution sub-model, a whole-field distribution calculation sub-model, and a parameter transfer function; The material state evolution sub-model is used to describe the variation of residual stress distribution, interface roughness, and contact pressure with temperature, electric field, and mechanical stress in the initial intrinsic state parameter set. The overall field distribution calculation sub-model is used to calculate the distribution of temperature field, electric field and stress field of the high-voltage cable joint based on its geometric structure, material properties and operating conditions. The parameter transfer function is used to convert the current material state parameters output by the material state evolution sub-model into the material property parameters required by the whole field distribution calculation sub-model.

3. The method according to claim 2, characterized in that, The parameter transfer function specifically includes: converting the current molecular chain orientation degree and residual stress value calculated by the material state evolution sub-model into the thermal conductivity correction coefficient and dielectric constant correction coefficient of the insulating material in the whole field distribution calculation sub-model; and converting the current true contact area ratio calculated by the material state evolution sub-model into the contact resistance coefficient and contact thermal resistance coefficient in the whole field distribution calculation sub-model.

4. The method according to claim 1, characterized in that, In step S5, the step of substituting real-time monitoring data into the coupled degradation model for inverse solution specifically includes the following sub-steps: S51: Construct an objective function with the goal of minimizing the weighted sum of squared residuals between the theoretical temperature distribution and theoretical electric field distribution output by the overall field distribution calculation sub-model and the measured temperature data and partial discharge pulse data in step S4; S52: Take the real-time intrinsic state parameter set as the variable to be optimized and set its value range; S53: Iteratively adjust the variable to be optimized using the adjoint state method or integrated Kalman filter algorithm until the objective function converges, and output the real-time intrinsic state parameter set at the current moment.

5. The method according to claim 4, characterized in that, The real-time monitoring data mentioned in step S4 also includes vibration acceleration signals; the objective function mentioned in step S51 simultaneously considers the spatial distribution residual of the temperature field, the phase distribution characteristics of the partial discharge pulse, and the spectral energy distribution differences of the vibration signal.

6. The method according to claim 1, characterized in that, In step S6, the calculation of remaining service life based on the coupled degradation model specifically includes: Using the current real-time intrinsic state parameter set as the initial value, and the preset future typical load curve and ambient temperature curve as input conditions, the electro-thermal-mechanical coupling degradation model is driven to perform time-domain progressive simulation. During the simulation, the key performance parameters output by the whole field distribution calculation sub-model are monitored in real time. The key performance parameters include at least the maximum field strength, the maximum temperature, and the contact resistance value. When any of the key performance parameters reaches its preset failure threshold, the simulation is stopped, and the time from the start time to the stop time of the simulation is taken as the remaining service life.

7. The method according to claim 1, characterized in that, It also includes step S7: S7: When the insulation performance degradation level determined in step S6 reaches the preset warning level, based on the abnormal parameter components in the real-time intrinsic state parameter set and combined with the inverse function of the mapping model, the original process parameter deviation range that caused the abnormal parameter components is traced back in reverse, and a process optimization report containing specific improvement suggestions is generated.

8. The method according to claim 7, characterized in that, In step S7, the reverse tracing specifically includes: if the current contact resistance increase in the real-time intrinsic state parameter set exceeds a preset threshold, the corresponding assembly torque deviation value and torque application rate deviation value are calculated through the inverse contact mechanics function in the mapping model, and the bolt position where the deviation occurs and the suggested correction value are indicated in the process optimization report.

9. The method according to claim 1, characterized in that, In step S2, the assembly parameters include the torque time-series curve and the rotation time-series curve during the bolt tightening process; the mapping model is established based on contact mechanics and is used to calculate the pressure distribution of the contact interface and the actual contact area of ​​the interface according to the torque time-series curve and the rotation time-series curve.

10. An online monitoring system for the production cycle of high-voltage cable joints, used to implement the method according to any one of claims 1 to 9, characterized in that, include: The identification code management unit is used to generate and assign identification codes to high-voltage cable joints, and to establish the association between the identification codes and the full-cycle data; The process data acquisition unit is used to acquire the timing data of the insulation extrusion process parameters and the timing data of the bolt fastening assembly parameters associated with the identification code from the extrusion production line control system and intelligent assembly tools. The parameter parsing and modeling unit has the mapping model built in, which is used to calculate and output the initial intrinsic state parameter set based on the process parameter time series data and the assembly parameter time series data, and to construct the electro-thermal-mechanical coupling degradation model based on the initial intrinsic state parameter set. The operation monitoring data acquisition unit is deployed at the high-voltage cable joint site to collect temperature, partial discharge pulse, load current and vibration acceleration data associated with the identification code in real time; The state inversion calculation unit is connected to the operation monitoring data acquisition unit and the parameter analysis and modeling unit. It is used to drive the electro-thermal-mechanical coupling degradation model to perform inverse solution with the real-time monitoring data as input, and calculate and output the real-time intrinsic state parameter set. The health assessment and life prediction unit is used to calculate the current offset and rate of change of each parameter based on the initial intrinsic state parameter set and the real-time intrinsic state parameter set, determine the insulation performance degradation level, and simulate and calculate the remaining service life. The source tracing diagnosis and process feedback unit is used to generate and output a process optimization report when the insulation performance degradation level reaches the warning level, based on the abnormal parameter components in the real-time intrinsic state parameter set and combined with the inverse function of the mapping model. The data storage and visualization unit is used to store the full-cycle data associated with the identification code and to provide a visual query interface for digital archives.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the online monitoring method for the production cycle of high-voltage cable joints as described in any one of claims 1 to 9.

Images