A method for collaborative manufacturing optimization of an insulated cable accessory

Abstract

The application discloses a kind of synergic production optimization methods of insulated cable accessory, it is related to physical simulation technical field.Method includes: multiple-source monitoring data is collected by sensor array;Coupled model of multiple physical fields is constructed, and real-time generates heat-power-electricity coupling state;Initial lead sealing optimization temperature and straightening optimization parameters are generated according to coupling state;Real-time data of lead sealing and straightening process are interactively shared, and initial parameters are cross-corrected based on shared data, and lead sealing optimization temperature and straightening optimization parameters are obtained;According to lead sealing optimization temperature and straightening optimization parameters, the lead sealing control and straightening control of polypropylene cable accessory are implemented.The technical problem of uneven thermal stress and material damage caused by independent control of lead sealing and straightening process is solved, and the technical effect of improving the quality and performance of cable accessory by real-time data sharing and cross-correction, realizing the synergic optimization of lead sealing and straightening process is achieved.

Description

Technical Field

[0001] This invention relates to the field of physical simulation technology, specifically to a collaborative manufacturing optimization method for insulated cable accessories. Background Technology

[0002] As power systems develop towards higher voltage, larger capacity, and higher reliability, polypropylene insulated cables are increasingly being used in medium and high voltage cables and their accessories due to their excellent dielectric properties, good aging resistance, and environmental friendliness. Polypropylene is a thermoplastic material with high sensitivity to heat deformation; it is prone to deformation or performance degradation under heat and stress, placing higher demands on process control during cable accessory manufacturing.

[0003] In the manufacturing process of polypropylene cable accessories, the lead sealing and straightening processes are critical steps affecting the quality and long-term operational reliability of the accessories. Existing lead sealing processes largely rely on manual experience for heating control, making precise temperature distribution difficult to adjust. This can easily exceed the heat resistance limits of the auxiliary materials in the polypropylene in localized areas, leading to problems such as porosity in the insulation shielding layer and surface defects, thus reducing the insulation performance and sealing reliability of the cable accessories. Meanwhile, the straightening process typically uses fixed heating and pressure parameters, making dynamic adjustments based on the actual heating and stress conditions of the cable difficult. This can easily result in residual stress concentration or uneven deformation, affecting the assembly accuracy and structural stability of the accessories. Furthermore, in existing technologies, the lead sealing and straightening processes are usually implemented independently, lacking data sharing and collaborative control mechanisms between the processes. This prevents comprehensive optimization of heat input, material deformation, and electric field distribution from a holistic perspective, resulting in lag and limitations in process parameter adjustments, making it difficult to effectively suppress defect generation. Summary of the Invention

[0004] This application provides a collaborative manufacturing optimization method for insulated cable accessories, which solves the technical problems of uneven thermal stress and material damage caused by independent control of lead sealing and straightening processes.

[0005] This application provides a collaborative manufacturing optimization method for insulated cable accessories, the method comprising: Multi-source monitoring data is collected in real time using a sensor array. This data includes temperature distribution in the lead-sealing area, temperature and pressure data in the straightening area, and defect characteristics of the insulation shielding layer. Based on this multi-source monitoring data and cable material parameters, a multi-physics coupling model is constructed to generate the thermo-mechanical-electric coupling state in the lead-sealing and straightening processes in real time. According to this thermo-mechanical-electric coupling state, the lead-sealing process is adjusted and the straightening process is optimized, generating initial lead-sealing optimization temperature and initial straightening optimization parameters. Real-time data from the lead-sealing and straightening processes are shared and cross-corrected using the shared data to generate lead-sealing optimization temperature and straightening optimization parameters. Based on these optimized temperatures and parameters, lead-sealing and straightening control of polypropylene cable accessories is performed.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, a sensor array is deployed during the lead sealing and straightening processes to monitor temperature, pressure, and insulation shielding defects in real time. Then, a multiphysics coupling model is established by combining the collected data with cable material parameters to dynamically characterize the combined thermal, mechanical, and electric field states during the lead sealing and straightening processes. Next, the lead sealing and straightening processes are analyzed and adjusted according to the coupling state to obtain the corresponding initial optimized temperatures and straightening parameters. Then, the real-time data from both processes are shared, and the initial parameters are corrected through mutual feedback to form the final optimized lead sealing temperature and straightening parameters. Finally, precise lead sealing and straightening control is implemented on the polypropylene cable accessories based on the optimized parameters to improve process stability and product quality. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic diagram of a collaborative manufacturing optimization method for insulated cable accessories provided in an embodiment of this application.

[0009] Figure 2 This is a schematic diagram illustrating the generation and coupling state process of a collaborative manufacturing optimization method for insulated cable accessories provided in this application embodiment. Detailed Implementation

[0010] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structure, features and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0011] Examples, such as Figure 1 As shown, this application provides a collaborative manufacturing optimization method for insulated cable accessories, the method comprising: Multi-source monitoring data is collected in real time through a sensor array. The multi-source monitoring data includes the temperature distribution of the lead-sealed area, the temperature and pressure data of the straightening area, and the defect characteristics of the insulating shielding layer.

[0012] In this embodiment, during the manufacturing of polypropylene cable accessories, sensor arrays are arranged at the lead-sealing station and the straightening station respectively for continuous and synchronous real-time monitoring and recording of critical areas. In the lead-sealing process, multiple temperature sensors are arranged along the axial and circumferential directions of the lead-sealing area, near the copper shell, aluminum sheath, and insulating shield. These temperature sensors can be thermocouples or infrared temperature probes, used to collect temperature values ​​at different locations during the lead-sealing heating process in real time. By performing time synchronization and spatial interpolation calculations on the multi-point temperature data, the temperature distribution, temperature gradient, and local hotspot information of the lead-sealing area at different times are obtained, used to determine whether there is a risk of local overheating. In the straightening process, temperature sensors are arranged along the cable axis within the heating zone to collect the heating temperature, heating rate, and temperature changes during the holding phase. Simultaneously, pressure sensors or strain sensors are installed on the straightening clamps, angle aluminum fixing structures, or tightening devices to collect the magnitude of pressure, rate of pressure change, and holding time of the cable during the straightening process in real time, thereby obtaining temperature and pressure coupling data of the straightening area, reflecting the real-time state of the cable under heat and stress. In addition, defect detection units are arranged on the surface of the insulating shield layer and in the transition area between it and the insulating layer to perform online or offline inspections of the insulating shield layer. These defect detection units may include industrial cameras, structured light profilometers, or non-destructive testing devices, used to scan the insulating shield layer through visual inspection or non-destructive testing methods to obtain the location information and dimensional characteristics of pores, holes, or surface anomalies, forming defect feature data of the insulating shield layer. Finally, the collected temperature distribution data of the lead sealing area, the temperature and pressure data of the straightening area, and the defect feature data of the insulating shield layer are collected through a unified data acquisition and communication module and processed synchronously according to a unified time base. After filtering, calibration, and outlier removal at the acquisition end, the data is formed into a multi-source monitoring dataset, providing input for subsequent process control and model calculations.

[0013] Based on the multi-source monitoring data and cable material parameters, a multi-physics coupling model is constructed to generate the thermo-mechanical-electric coupling state in the lead sealing and straightening process in real time.

[0014] In one embodiment, after obtaining the temperature distribution in the lead-sealing area, the temperature and pressure data in the straightening area, and the defect characteristics of the insulation shielding layer, these multi-source monitoring data and cable material parameters are used together as model inputs to construct a multi-physics coupled model for lead-sealing-straightening co-analysis in an industrial computer. During this process, the temperature distribution data from the multi-source monitoring data is combined with the thermophysical parameters of the cable material to construct a transient temperature field model for the lead-sealing area; the temperature and pressure data from the straightening area in the multi-source monitoring data are combined with the mechanical performance parameters of the cable material to construct an elasto-plastic stress field model for the straightening area; and the temperature and pressure data from the multi-source monitoring data are combined with the mechanical performance parameters of the cable material to construct an elasto-plastic stress field model for the straightening area. By combining the defect characteristics of the insulation shielding layer with the dielectric performance parameters of the cable material, a three-dimensional electric field distribution model of the insulation shielding layer is constructed. Then, the three models are coupled and iterated in both directions to calculate the thermo-mechanical-electric coupling state in real time during the lead sealing and straightening process. This thermo-mechanical-electric coupling state includes at least the temperature gradient and heat flux density in the lead sealing area, and the deformation displacement and residual stress distribution in the straightening area. This is used to achieve accurate characterization of the heating, stress and electric field response of polypropylene cables during the lead sealing and straightening process, providing a reliable physical basis for the dynamic control and synergistic optimization of subsequent process parameters, and reducing the risk of material thermal damage and defect generation.

[0015] Furthermore, such as Figure 2 As shown, based on the multi-source monitoring data and cable material parameters, a multi-physics coupling model is constructed to generate the thermo-mechanical-electric coupling state in the lead sealing and straightening process in real time, including: Based on the temperature distribution data and the thermophysical parameters of the cable material from the multi-source monitoring data, a transient temperature field model of the lead-sealing area is established; based on the temperature and pressure data of the straightening area and the mechanical performance parameters of the cable material from the multi-source monitoring data, an elastoplastic stress field model of the straightening area is established; based on the defect characteristics of the insulation shielding layer and the dielectric performance parameters of the cable material, a three-dimensional electric field distribution model of the insulation shielding layer is established; through a multi-field coupling solver, the transient temperature field model, the elastoplastic stress field model, and the three-dimensional electric field distribution model are subjected to bidirectional coupling iterative calculation, and the thermo-mechanical-electric coupling state during the lead-sealing and straightening process is output in real time.

[0016] Preferably, for the lead sealing process, firstly, based on the structural form of the polypropylene cable accessories and the characteristics of the lead sealing process, a geometric model consistent with the actual lead sealing area is established in an industrial computer. This geometric model includes at least the aluminum sheath, copper shell, lead sealing material, and adjacent insulation shielding layer areas. The lead sealing area is then divided into spatial grids to meet the accuracy requirements of transient thermal analysis. Next, multi-point temperature distribution data of the lead sealing area is extracted from multi-source monitoring data. Based on the actual installation position of each temperature sensor in the lead sealing area, the temperature data is mapped to the corresponding nodes or boundary surfaces in the geometric model, serving as the initial temperature conditions and real-time updated boundary conditions for the model. Subsequently, the thermophysical parameters of the cable material are read and assigned to the corresponding model area. These thermophysical parameters include at least the material's density, specific heat capacity, and thermal conductivity, and are preferably defined in the form of temperature-related functions to reflect the thermophysical property changes of the polypropylene material and the lead sealing material during the heating process. Simultaneously, external heat input parameters are set according to the lead sealing process conditions, including heating power, heating path or heating surface heat flux density, and environmental convective heat transfer coefficient and radiative heat transfer parameters. Next, based on the fundamental governing equations of heat conduction, a transient heat conduction equation describing the temporal and spatial variations of heat in the lead-sealing region was established. The initial conditions, boundary conditions, and thermophysical parameters were substituted into this transient heat conduction equation, and a numerical solution method was used to discretize and solve it, yielding the temperature distribution of the lead-sealing region at each time step. Finally, through these steps, the transient temperature field of the lead-sealing region was constructed, forming a transient temperature field model that can be updated in real-time with the lead-sealing process. This model characterizes the temperature variation and temperature gradient distribution at various locations during the lead-sealing process, providing a fundamental model support for subsequent lead-sealing process control.

[0017] For the straightening process, based on the actual stress state of the polypropylene cable during the straightening process, a geometric model consistent with the structure of the straightening area is established in an industrial computer. This geometric model includes at least the conductor, main insulation, insulating shielding layer, and the area where the straightening clamp contacts the cable. The straightening area is then divided into spatial meshes to meet the accuracy requirements of elastoplastic mechanical analysis. Temperature and pressure data of the straightening area are extracted from multi-source monitoring data. Based on the actual arrangement of temperature and pressure sensors in the straightening area, the temperature and pressure data are mapped to the corresponding nodes, surfaces, or loading areas in the geometric model. The temperature data serves as a thermal load input to describe the thermal state changes of the material during the straightening process, while the pressure data is converted into force loads or displacement boundary conditions and applied to the corresponding positions where the straightening clamp contacts the cable. Next, the mechanical property parameters of the cable material are read and assigned to the corresponding material region in the geometric model. These mechanical property parameters include at least the elastic modulus, Poisson's ratio, yield strength, plastic hardening parameters, and constitutive parameters characterizing the temperature-dependent elastoplastic behavior of polypropylene material. Preferably, these parameters are defined as temperature-dependent functions to reflect the evolution of the material's mechanical properties during the heating and straightening process. Then, based on continuum mechanics theory, a governing equation describing the stress and deformation relationship of the material in the straightening region is established. Temperature load, pressure load, and mechanical constitutive relations are substituted into the governing equation, and thermal expansion effects and plastic deformation characteristics are considered. An elastoplastic constitutive model is used to numerically solve the straightening process, obtaining the stress, strain, and deformation displacement distribution of the cable structure during the straightening process. Finally, through the above steps, the elastoplastic stress field of the straightening region is constructed, forming a model that can be dynamically updated with straightening process parameters and real-time monitoring data. This model characterizes the stress state and residual stress distribution of the cable during the straightening process, providing a basic model support for subsequent straightening process optimization and parameter control.

[0018] For the insulation shielding layer, based on the structural composition of the polypropylene cable and the actual morphology of the insulation shielding layer, a three-dimensional geometric model is established in an industrial computer, including the conductor, main insulation, insulation shielding layer, and its adjacent metal shielding or sheath. The insulation shielding layer and its transition region are finely meshed to meet the spatial resolution requirements of the electric field numerical calculation. Defect characteristic parameters of the insulation shielding layer are then extracted from multi-source monitoring data and introduced into the three-dimensional geometric model. Based on the defect characteristics, structural or material parameters are corrected at corresponding locations in the model. For example, pores, voids, or debonding areas are modeled as local cavity regions, thus reflecting the dielectric discontinuity of the insulation shielding layer in the model. Then, the dielectric performance parameters of the cable material are read and assigned to the corresponding material region in the model. These dielectric performance parameters include at least the dielectric constant, dielectric loss factor, and conductivity, and are preferably set as functions related to temperature or stress to reflect the dielectric performance changes of the polypropylene insulation and shielding materials under process and service conditions. Next, based on the electrical conditions of the cable accessories under operation or testing, potential or electric field boundary conditions are set in the three-dimensional geometric model. For example, the conductor is set as a high-potential boundary, and the metal shield or sheath is set as a ground boundary. Parameters such as applied voltage amplitude and frequency are introduced when necessary. Based on this, governing equations describing the spatial distribution of the electric field are established and solved in conjunction with changes in dielectric parameters caused by defects. Finally, through numerical calculations of the aforementioned three-dimensional geometric model, dielectric parameters, and boundary conditions, the electric field intensity distribution results of the insulation shield layer and its adjacent area in three-dimensional space are obtained, thus completing the construction of the three-dimensional electric field distribution model of the insulation shield layer. This model can reflect the influence of defects on local electric field distortion and field concentration, providing a basis for electric field analysis for subsequent insulation performance evaluation and process optimization.

[0019] After constructing the transient temperature field model of the lead-sealed region, the elasto-plastic stress field model of the straightened region, and the three-dimensional electric field distribution model of the insulating shielding layer, a multi-field coupled solver is used to perform bidirectional coupled iterative calculations on the transient temperature field model, the elasto-plastic stress field model, and the three-dimensional electric field distribution model. Specifically, in each time step, the transient temperature field model first outputs the current temperature distribution to update the material constitutive parameters and thermal strain terms in the elasto-plastic stress field model in real time. Then, the strain, density, or contact state changes calculated by the elasto-plastic stress field model are fed back to the transient temperature field model to correct local heat conduction paths and heat transfer efficiency. At the same time, the equivalent heat source terms caused by dielectric loss or local field strength anomalies in the three-dimensional electric field distribution model are fed back to the transient temperature field model, and the changes in dielectric parameters caused by stress-deformation are fed back to the three-dimensional electric field distribution model. By setting a unified time step, data exchange interface, and convergence criteria, the physical fields are iteratively calculated until the preset convergence conditions are met. This allows for continuous and real-time output of the thermo-mechanical-electric coupling state of the interaction between reaction temperature, stress, and electric field during the lead sealing and straightening processes, providing a precise physical basis for subsequent lead sealing temperature control and straightening parameter optimization.

[0020] Furthermore, the thermo-mechanical-electric coupling state includes the temperature gradient distribution and heat flux density of the lead-sealed region, the residual stress distribution and deformation displacement of the straightening region, and the changes in the electric field strength and dielectric constant of the insulating shielding layer.

[0021] Optionally, after constructing the transient temperature field model of the lead-sealing region, the elasto-plastic stress field model of the straightening region, and the three-dimensional electric field distribution model of the insulating shielding layer, a multi-field coupling solver is used to uniformly schedule and perform simultaneous calculations on the above physical fields, outputting the coupling results of each physical quantity under the same time step and spatial coordinate system. Specifically, in the thermal field calculation process, based on the transient temperature field model, the temperature change of each spatial node in the lead-sealing region over time is solved, and the temperature gradient distribution of the lead-sealing region is obtained by spatially differentiating the temperature difference between adjacent nodes. At the same time, combined with the thermal conductivity of the material, the heat flux density at each location is calculated to characterize the direction and intensity of heat transfer in the lead-sealing region. In the mechanical field calculation process, the temperature distribution output by the thermal field is introduced as a thermal load into the elasto-plastic stress field model to solve the structural state after the straightening process, obtaining the stress distribution of the straightening region after unloading, and extracting the irrecoverable part as the residual stress distribution. At the same time, based on the displacement results of nodes or elements in the elasto-plastic stress field model, the overall and local deformation displacement of the straightening region is obtained to reflect the straightening effect and deformation uniformity. During the electric field calculation, based on a three-dimensional electric field distribution model, the electric field of the insulating shielding layer under its current geometric structure and material state is solved, outputting the electric field intensity distribution at each location of the insulating shielding layer. Simultaneously, considering the influence of temperature and stress fields on dielectric parameters, the dielectric constant of the insulating shielding layer material is dynamically updated, obtaining the results of how the dielectric constant changes with spatial location and process. Through the synchronous output of these multi-field results within the same computational framework, a thermo-mechanical-electric coupling state is formed, encompassing the temperature gradient distribution and heat flux density in the lead-sealing region, the residual stress distribution and deformation displacement in the straightening region, and the changes in the electric field intensity and dielectric constant of the insulating shielding layer. This comprehensively characterizes the interaction relationships between various physical fields during the lead-sealing and straightening processes, providing an intuitive and quantifiable basis for precise control of process parameters and early prediction of defect risks.

[0022] Based on the aforementioned thermo-mechanical-electric coupling state, the lead sealing process is controlled and the straightening process is optimized, generating the initial lead sealing optimization temperature and the initial straightening optimization parameters.

[0023] In one embodiment, after obtaining the thermo-mechanical-electric coupling state, it is used as real-time input for process decisions, and the lead sealing process and straightening process are independently optimized in an industrial computer. Specifically, during lead sealing process control, the temperature gradient distribution and heat flux density of the lead sealing area are first read, and key locations such as the lead sealing overlap area and the area adjacent to the insulation shielding layer are identified to determine whether there are hot spots, areas with excessively rapid temperature rise, or areas with excessively large temperature gradients. Constrained by the material's heat resistance threshold and allowable temperature gradient, a temperature curve satisfying "uniform temperature, limited peak value, and controllable gradient" is obtained through comprehensive calculation, forming the initial optimized lead sealing temperature. During straightening process optimization, the residual stress distribution and deformation displacement of the straightening area are read, and the deformation uniformity index and residual stress index of the current straightening result are calculated. With "meeting the deformation uniformity target and minimizing residual stress" as the optimization objective, the material effects are judged in conjunction with real-time temperature data of the straightening area, and then the pre-compensation correction amount used to compensate for deformation errors is obtained through solution, thereby generating the initial optimized straightening parameters. These two initial results serve as the basic inputs for subsequent linkage correction and actual control of lead sealing and straightening, thereby enabling pre-constraint and controllable adjustment of thermal damage risk, deformation risk and electric field abnormality risk before process implementation.

[0024] Furthermore, based on the aforementioned thermo-mechanical-electric coupling state, the lead sealing process is controlled to generate an initial optimized lead sealing temperature, including: Based on the temperature gradient distribution and heat flux density in the thermo-mechanical-electric coupling state, a response relationship matrix between the temperature field and heating power is constructed. According to this response relationship matrix, the heating power partitions of the lead-sealing area are dynamically divided, generating a partitioned heating control strategy. Based on the changes in the electric field strength and dielectric constant of the insulating shielding layer in the thermo-mechanical-electric coupling state, anti-interference control logic is designed. When an abnormal electric field strength is detected, a power attenuation mechanism for adjacent heating partitions is automatically triggered. Based on the heating trajectory and defect correlation data in the historical process database, a heating path optimization knowledge base is established, and the heating mode with the lowest risk of porosity defects is prioritized for path planning, generating an optimized heating path. Combining the partitioned heating control strategy, the anti-interference control logic, and the optimized heating path, an initial optimized lead-sealing temperature curve is generated.

[0025] Preferably, after obtaining the thermo-mechanical-electric coupling state corresponding to the lead sealing process, the lead sealing area is first divided into multiple spatial units along the axial and circumferential directions, and the power output of the corresponding heating device is divided into multiple independently adjustable power channels. During each sampling period, the input changes of each power channel, as well as the temperature change rate, steady-state temperature value, and temperature gradient changes within the corresponding spatial unit, are recorded. Through linear regression, a response relationship matrix is ​​established with the heating power change as input and the temperature gradient and heat flux density response as outputs, used to quantify the influence of different heating power adjustments on the temperature field distribution of the lead sealing area. Based on this, according to the response relationship matrix, the lead sealing area is dynamically partitioned by heating power based on temperature gradient, initial power, and graded cooling protocol, resulting in a partitioned heating control strategy for dividing the lead sealing area into several heating characteristic partitions. Subsequently, based on the changes in the electric field strength and dielectric constant of the insulating shielding layer in the thermo-mechanical-electric coupling state, an anti-interference control logic is designed to cooperate with the partitioned heating control. Specifically, an industrial computer is configured with a safe threshold range for electric field strength and an allowable range for dielectric constant variation. When the electric field strength at a certain location exceeds the preset threshold or an abnormal change in dielectric constant occurs, the abnormal location is mapped to the corresponding lead-sealing heating zone. Furthermore, adjacent heating zones with thermal coupling relationships are identified, and a power attenuation mechanism is automatically triggered. Based on a preset attenuation ratio, graded power reduction and localized heating pauses are implemented in the relevant zones to suppress the adverse effects of localized overheating on the dielectric properties of the insulating shielding layer. In addition, a heating path optimization knowledge base is established based on heating trajectory and defect correlation data in a historical process database. The heating start position, heating sequence, power change trajectory, and the number, size, and distribution of the final detected porosity defects recorded in the historical lead-sealing process in the knowledge base are statistically analyzed. Heating sequence patterns and power change characteristics with low defect occurrence probabilities are extracted and stored as callable heating path rules. Under current process conditions, low-defect-risk heating modes that match the current material parameters, structural dimensions, and temperature response characteristics are preferentially selected from the knowledge base. Finally, by integrating the zonal heating control strategy, anti-interference control logic, and optimized heating path, the power setpoints of each heating zone in different time periods are uniformly arranged to calculate the target temperature distribution that changes over time during the lead sealing process. This is then converted into a continuous initial lead sealing optimized temperature curve. This initial lead sealing optimized temperature curve can effectively reduce the temperature gradient and local overheating risk in the lead sealing area while ensuring overall heating efficiency, providing a basic control target for real-time control and dynamic correction of the subsequent lead sealing process.

[0026] Furthermore, the dynamic division of the heating power zone for the lead sealing area includes: Based on the temperature gradient distribution of the lead-sealed area in the thermo-mechanical-electric coupling state, the lead-sealed area is divided into a core heating zone, a transition heating zone, and an outer protection zone. Based on the heat flux density distribution of the lead-sealed area in the thermo-mechanical-electric coupling state, the required heat input power ratio for each zone is calculated, and the initial power setting value for each zone is determined. A temperature buffer zone is set in the outer protection zone, and a graded cooling protocol is configured. When the local temperature is detected to be close to the material's heat resistance threshold, a stepped cooling program is automatically started. A sliding window is used to process the real-time temperature data and dynamically update the boundary range of each heating zone.

[0027] Optionally, temperature gradient distribution data for each spatial location within the lead-sealed region is first read from the thermo-mechanical-electric coupling state. The magnitude of the temperature gradient is then used as the partitioning criterion to initially divide the lead-sealed region. When the temperature gradient per unit area within a region exceeds a preset first gradient threshold, that region is designated as the core heating zone; when the temperature gradient is between the first and second gradient thresholds, it is designated as the transition heating zone; and when the temperature gradient is below the second gradient threshold and is close to the insulating shielding layer or other heat-sensitive structures, it is designated as the outer protection zone. This creates a spatially progressively decreasing heating zone structure from the inside out. Subsequently, based on the heat flux density distribution within the lead-sealed region under the thermo-mechanical-electric coupling state, the required heat input power ratio for each heating zone is calculated. Specifically, the average heat flux density per unit area is calculated for the core heating zone, transition heating zone, and outer protection zone, and the required heat input power ratio for each zone under the target temperature distribution condition is calculated using the response relationship matrix. Based on this, the heat input power ratios are allocated to each heating zone, and the initial power setpoints for each zone are determined. This ensures that the core heating zone receives the primary heat input, the transition heating zone receives secondary heat input, and the outer protection zone maintains only low power or a heat preservation state. Subsequently, a temperature buffer zone is set within the outer protection zone. This buffer zone is located at the boundary between the transition heating zone and the outer protection zone, used to reduce the speed and intensity of heat diffusion outwards. A graded cooling protocol is configured for the temperature buffer zone, containing multiple preset temperature approach thresholds. When the temperature at any measuring point within the buffer zone or outer protection zone gradually approaches the heat resistance threshold of the cable material, a stepped cooling program is automatically initiated according to a preset sequence. This includes sequentially reducing the output power of adjacent heating zones, decreasing the power adjustment step size, extending the heating interval time, and pausing local heating if necessary, to ensure that the local temperature never exceeds the material's safe upper limit. Furthermore, throughout the entire lead sealing process, a sliding window method is used to continuously process real-time temperature data. In this process, a sliding time window of preset length is constructed, and statistical analysis is performed on the temperature change rate, temperature gradient, and trends of each measuring point within each window period. When the temperature gradient distribution calculated within the sliding window shifts, causing an increased outward expansion trend of the original core heating zone or transition heating zone boundaries, the zoning criteria are recalculated. The spatial range of the core heating zone and transition heating zone is reduced proportionally, while the coverage of the outer protection zone or temperature buffer zone is expanded, thus dynamically updating the boundary positions of each heating zone. Through the coordinated control of the above-mentioned zoning criteria, power allocation, buffer protection, and dynamic boundary updates, the heat-affected zone during the lead sealing process is always strictly limited to a predetermined safe area, effectively preventing the disorderly diffusion of heat to the insulating shielding layer and other sensitive structures, and reducing the risk of porosity or performance degradation in polypropylene materials due to localized overheating.

[0028] Furthermore, based on the aforementioned thermo-mechanical-electric coupling state, the straightening process is optimized to generate initial straightening optimization parameters, including: Based on the residual stress distribution and deformation displacement in the thermo-mechanical-electrical coupling state, a coupled dynamic model of material deformation and applied pressure during the straightening process is established. According to the coupled dynamic model, the optimal pressure loading sequence that satisfies the objectives of deformation uniformity and residual stress minimization is calculated. Based on the temperature data of the straightening region in the thermo-mechanical-electrical coupling state, an adaptive straightening control strategy is designed to automatically adjust the duration and temperature gradient requirements of the pressure holding stage when changes in material creep characteristics are detected. Based on the mapping relationship between historical deformation data and final dimensional errors, a deformation error prediction model is established, and the pressure loading sequence is pre-compensated and corrected according to the prediction results. Combining the optimal pressure loading sequence, the adaptive straightening control strategy, and the pre-compensation correction, an initial straightening optimization parameter set containing pressure, time, and temperature parameters is generated.

[0029] Preferably, a coupled dynamic model between material deformation and applied pressure during the straightening process is first established in an industrial computer based on the residual stress distribution and deformation displacement in the straightening region obtained under thermo-mechanical-electrical coupling conditions. Specifically, the residual stress distribution and deformation displacement in the straightening region under thermo-mechanical-electrical coupling conditions are used as the initial state variables of the model. The residual stress distribution describes the internal stress state of the cable before straightening or after the previous step, and the deformation displacement characterizes the current degree of straightening and deformation non-uniformity. Simultaneously, parameters such as the applied pressure, loading rate, and holding time during the straightening process are defined as external input loads and applied to the corresponding boundary of the model according to the actual position of the straightening fixture. Subsequently, the mechanical property parameters of the cable material are introduced. These mechanical property parameters include the elastic modulus, Poisson's ratio, yield strength, plastic hardening parameters, and creep or viscoelastic parameters describing the time-dependent behavior of polypropylene materials. Preferably, these parameters are defined in a temperature-dependent form to reflect the evolution of the material's mechanical properties under heating and straightening conditions. Based on this, a dynamic governing equation describing the relationship between stress, strain, deformation displacement, and external pressure is established, with residual stress introduced as an initial stress term into the equation system. Next, the dynamic governing equation is combined with the boundary and load conditions during the straightening process. Numerical solutions are used to calculate the deformation evolution and stress redistribution process of the material under different pressure loading states during the straightening process, obtaining the response relationship of deformation displacement with pressure and time, and the corresponding update law of residual stress. Finally, through the above steps, a coupled dynamic model of the straightening process is constructed that reflects the intrinsic relationship between "applied pressure - deformation response - residual stress evolution" during the straightening process. This model takes applied pressure as input and deformation displacement and residual stress as outputs, and is used to characterize the dynamic coupling behavior between material deformation and force during the straightening process, providing a basic model support for the subsequent optimization calculation of pressure loading sequences and the design of straightening process parameters.

[0030] Subsequently, based on this coupled dynamic model, and with the objectives of maximizing deformation uniformity and minimizing residual stress, an optimization algorithm was used to obtain a set of Pareto optimal solutions that satisfy the constraints. Then, the optimal pressure loading sequence was selected from this set of Pareto optimal solutions as the optimal pressure loading sequence. Based on this, and combining real-time temperature data of the straightening region in the thermo-mechanical-electrical coupled state, an adaptive straightening control strategy was designed. Specifically, during the straightening process, multi-point temperature data of the straightening region were collected in real time, and the average temperature T at the current moment was calculated. avg The rate of temperature change dT / dt and the axial and circumferential temperature gradients T, and simultaneously, a temperature-creep correlation table for polypropylene materials was established. This correlation, determined based on material test data, is used to classify the material into low-creep, stable-creep, and creep-sensitive regions. When T avgWhen the temperature is below the first temperature threshold T1, the material is determined to be in the low creep region, and the material deformation is mainly elastic-plastic response. At this time, the reference holding time t0 is used in the holding stage, and the allowable temperature gradient is set as the reference gradient. T0. When T avg When the temperature is between the first temperature threshold T1 and the second temperature threshold T2, and dT / d is less than the preset rate of change threshold, the material is determined to have entered the stable creep region. At this time, the holding time is adjusted according to a preset ratio, extending the holding time by 10% to 30% based on the baseline value, while reducing the upper limit of the allowable temperature gradient by 10% to 20%, in order to release residual stress through controllable creep. When T avg Exceeding the second temperature threshold T2, or detecting a temperature gradient. If temperature (T) increases rapidly within a short period, the material is considered to have entered the creep-sensitive zone. At this point, a significant change in the material's creep characteristics is considered to have occurred, automatically triggering an enhanced control strategy. This involves extending the holding pressure stage duration by 30%–50% from the current set value, while simultaneously lowering the temperature gradient control target during the straightening process to 50%–70% of the original set value, and limiting the pressure loading rate to prevent over-plasticization or local instability. The adjusted holding pressure stage duration and temperature gradient are stored to form an adaptive straightening control strategy. These adjustment ratios are not fixed values ​​but are gradually corrected through continuous evaluation of temperature and deformation data using a sliding time window. For example, if the material's creep index continues to rise within two consecutive time windows, the holding pressure time is extended in a stepwise manner, with each extension not exceeding 10% of the baseline value, until the creep index stabilizes. If the creep index falls back, the original parameters are gradually restored to prevent excessive degradation of straightening efficiency.

[0031] For example, the baseline holding time is set to 120 seconds, and the baseline allowable temperature gradient is 3°C / cm. When the average temperature of the calibration area is monitored to rise from 80°C to 88°C and enter the stable creep zone, the holding time is adjusted to approximately 150 seconds (extended by approximately 25%), and the allowable temperature gradient is reduced to approximately 2.5°C / cm. When the temperature further rises to 92°C and the temperature gradient fluctuation intensifies, it is determined that the creep-sensitive zone has been entered. The holding time is automatically adjusted to approximately 180 seconds, while the allowable temperature gradient is limited to within 1.8°C / cm, and the pressure loading pace is slowed down, thereby completing the calibration without causing excessive thermal damage.

[0032] Then, based on the mapping relationship between deformation displacement data and final dimensional error recorded in historical straightening processes, a deformation error prediction model is established. Specifically, representative sample data is selected from the historical straightening process database, covering cables of different specifications, different straightening parameters, and different temperature conditions. For each historical sample, representative deformation characteristic variables during the straightening process are extracted as independent variables, including but not limited to maximum deformation displacement, axial and circumferential deformation uniformity indices, deformation change rate during the pressure holding stage, and springback estimate after unloading. Simultaneously, the corresponding final dimensional error is extracted as the dependent variable, which can be straightness error, outer diameter deviation, or end coaxiality deviation. Based on this, according to the correlation and physical meaning between the deformation characteristic variables, an input vector for a multivariate regression model is constructed, and a suitable regression form, such as linear regression, multinomial regression, or a regression model with cross terms, is selected to describe the comprehensive influence of various deformation characteristics on the final dimensional error. Next, the parameters of the multivariate regression model are solved using the least squares method or regularized regression. By minimizing the sum of squared residuals between the model's predicted values ​​and the actual dimensional errors, the regression coefficients corresponding to each deformation characteristic variable are determined. Cross-validation or leave-one-out method is used to evaluate the model's fitting accuracy and generalization ability, and necessary adjustments and constraints are made to the regression coefficients to avoid overfitting. After the model parameters are determined, the multivariate regression model is solidified into a deformation error prediction model, used to calculate the corresponding final dimensional error prediction value based on real-time collected deformation displacement data during the actual straightening process, thus providing a quantitative basis for the pre-compensation correction of the subsequent pressure loading sequence. During the current straightening process, the real-time collected deformation displacement data is input into the deformation error prediction model to calculate the final dimensional error prediction value under the current working conditions. This prediction value is compared with the target dimensional error range to determine whether there is a trend of exceeding the allowable tolerance. When the prediction results indicate that the final dimensional error has a risk of exceeding the limit, the original pressure loading sequence is pre-compensated and corrected based on the direction and magnitude of the prediction error, thereby obtaining the pre-compensation correction amount. Specifically, if the prediction results show insufficient straightening, the pressure level in the subsequent loading stage will be increased or the holding time in the fine pressing stage will be extended accordingly; if the prediction results show excessive straightening or increased risk of rebound, the pressure peak will be reduced or the holding time will be shortened. The entire correction process will be carried out in a small-scale, phased manner to avoid causing drastic disturbances to the straightening process.

[0033] Finally, the optimal pressure loading sequence, adaptive straightening control strategy, and pre-compensation correction amount obtained from deformation error prediction are integrated to form an initial straightening optimization parameter set for the straightening equipment. This initial straightening optimization parameter set includes at least the pressure setpoint, loading and holding time, pressure change rate, and temperature and temperature gradient control requirements corresponding to each straightening stage. It is used to guide the precise control of the straightening process in stages, thereby effectively reducing the residual stress level while ensuring the uniformity of straightening deformation, and improving the stability and consistency of the straightening process of polypropylene cable accessories.

[0034] Furthermore, based on the coupled dynamics model, the optimal pressure loading sequence that satisfies both the deformation uniformity objective and the residual stress minimization objective is calculated, including: Based on the coupled dynamic model, with the objectives of maximizing deformation uniformity and minimizing residual stress, an optimization algorithm is used to solve the problem and obtain a Pareto optimal solution set. From the Pareto optimal solution set, a comprehensive optimal solution is selected to generate a segmented pressure loading curve containing the pre-compression stage, the main pressure stage, and the fine pressure stage, which serves as the optimal pressure loading sequence.

[0035] Optionally, when calculating the coupled dynamics model, deformation uniformity and residual stress are first defined as quantifiable evaluation indices. Deformation uniformity is characterized by the dispersion of deformation displacement at each measuring point within the straightening area, using the maximum difference, standard deviation, or root mean square deviation of deformation displacement as the quantifiable index. Residual stress is characterized by the peak value, average value, or energy density of the residual stress distribution after straightening. By weighting these two indices, a multi-objective optimization function with maximizing deformation uniformity and minimizing residual stress as parallel objectives can be constructed. Subsequently, the pressure loading process is modeled in a segmented parameterized manner, dividing the entire straightening pressure loading process into a pre-compression stage, a main compression stage, and a fine-compression stage. Key parameters of each stage are introduced into the coupled dynamics model as optimization variables. These optimization variables include at least the initial pressure value and loading rate of the pre-compression stage, the target pressure peak value and duration of the main compression stage, and the pressure fine-tuning amplitude and holding time of the fine-compression stage. Simultaneously, engineering constraints are set, including the maximum output pressure of the straightening equipment, the maximum allowable loading rate, the upper limit of the material's safe pressure, and temperature-pressure coupling constraints, to ensure the optimization results are practically feasible. Subsequently, a multi-objective optimization algorithm was used to solve the aforementioned optimization problem. Specifically, within a pre-defined parameter search space, multiple sets of candidate pressure loading parameter combinations were generated according to iterative rules. For each set of candidate parameters, they were substituted into a coupled dynamic model for simulation calculation to obtain the corresponding deformation uniformity index and residual stress index values. By judging the dominance relationship of all candidate solutions, solutions that are not simultaneously superior to other solutions in both deformation uniformity and residual stress objectives were selected, thus constructing a Pareto optimal solution set. This solution set reflects the optimal trade-off between different pressure loading strategies and the two objectives. After obtaining the Pareto optimal solution set, a comprehensive optimal solution was selected based on a multi-objective optimization function. That is, the deformation uniformity and residual stress of each candidate parameter in the Pareto optimal solution set were weighted according to the multi-objective optimization function, the comprehensive evaluation value corresponding to each candidate parameter was calculated, and the solution with the best comprehensive evaluation value was selected as the final optimized solution. Finally, based on the pressure parameters corresponding to the selected optimal solution, pressure-time curves are generated in chronological order, including pre-compression, main compression, and fine compression stages. In the pre-compression stage, lower pressure and a slow loading rate are used to achieve fixture fit and initial stress release. In the main compression stage, higher pressure is used to complete the main deformation correction. In the fine compression stage, small pressure adjustments and stable pressure holding are used to homogenize deformation and reduce residual stress. By summarizing the pressure-time curves of each stage, the resulting segmented pressure loading curves are used as the optimal pressure loading sequence for subsequent automated staged control of the straightening equipment.

[0036] The real-time data of the lead sealing and straightening processes are shared and interactively, and the shared data is used to cross-correct the initial lead sealing optimization temperature and the initial straightening optimization parameters to generate the lead sealing optimization temperature and straightening optimization parameters.

[0037] In one embodiment, after obtaining the initial lead sealing optimization temperature and initial straightening optimization parameters, real-time data from both processes are shared through the same data channel, and cross-correction is performed using correlation mapping, bidirectional feedback, and model correction. Specifically, a process parameter correlation map is first established, mapping the residual stress distribution measured after the straightening process to the partition coordinates of the lead sealing process according to its axial position and circumferential angle. A preset mapping function is then used to convert the residual stress level into a temperature compensation coefficient for each partition. When the residual stress in a partition is too high, a cooling compensation or a reduction in the heating gradient is applied to that partition, thereby correcting the initial lead sealing temperature setting. Simultaneously, a bidirectional process feedback mechanism is constructed based on the defect characteristics of the insulating shielding layer. The location and size parameters of pore defects detected online during the lead sealing stage are automatically projected to the corresponding positions in the straightening area, forming a pressure correction increment. This increment is used to correct the initial pressure loading sequence by reducing pressure or extending the fine pressure holding period in the corresponding segments, thus preventing defects from expanding under stress. Subsequently, the model trained from historical shared data is invoked to quickly predict the correlation between preceding parameters, real-time monitoring, and subsequent optimization parameters. The prediction results are used as the basis for verification and secondary correction. The initial lead sealing optimization temperature and initial straightening optimization parameters are cross-updated until the temperature safety constraints, residual stress threshold, and defect risk threshold are met. Finally, the lead sealing optimization temperature and straightening optimization parameters are output for the coordinated control of lead sealing and straightening of polypropylene cable accessories, thereby achieving global optimization of thermal management and reducing the risk of porosity defects and deformation damage.

[0038] Furthermore, real-time data exchange and sharing between the lead sealing and straightening processes are implemented, and the shared data is used to cross-correct the initial lead sealing optimization temperature and initial straightening optimization parameters to generate the lead sealing optimization temperature and straightening optimization parameters, including: A process parameter correlation map is established, and the residual stress distribution data measured after the straightening process is converted into temperature compensation coefficients for each zone of the lead sealing process through a mapping function. Based on the defect characteristics of the insulation shielding layer in the multi-source monitoring data, a two-way process feedback mechanism is designed to automatically convert the location and size data of pore defects detected in the lead sealing process into pressure correction increments in the corresponding areas of the straightening process. A neural network model is trained using historical shared data. This model takes the preceding process parameters and real-time monitoring data as input and the optimized parameters of the subsequent process as output to perform collaborative optimization of process parameters. The trained neural network model is applied to cross-correct and verify the initial lead sealing optimization temperature and the initial straightening optimization parameters to generate the lead sealing optimization temperature curve and the straightening optimization parameter set for actual control.

[0039] Preferably, before cross-correction, the axial position and circumferential angle of the polypropylene cable accessories are used as a unified spatial coordinate reference. The key areas involved in the lead sealing and straightening processes are spatially discretized to form several structural units with clear physical location meanings. Simultaneously, key process parameters in both processes are used as graph node types, including parameters such as zone heating temperature, heating power, and heating rate in the lead sealing process, and pressure loading parameters, deformation displacement, and residual stress level in the straightening process. Each parameter node is labeled with its associated process, physical meaning, and dimensional attributes. Subsequently, based on historical process data and real-time monitoring data, statistical and physical constraint analyses are performed on the correlation between lead sealing process parameters and straightening process result parameters. For example, the correlation between the lead sealing zone temperature level and the subsequent straightening residual stress distribution and deformation non-uniformity is analyzed. For parameter pairs with significant correlation, correlation edges are established in the graph, and weights or functions are assigned to these edges to represent the influence strength and direction of the parameters. The correlation can be quantified based on the Pearson correlation coefficient. Next, based on the residual stress distribution data obtained from the actual measurements after the straightening process, a "residual stress-temperature compensation" mapping node is constructed in the graph. The residual stress value is converted into a temperature compensation coefficient for the corresponding zone in the lead sealing process using a mapping function. This mapping function can be set based on the material's heat resistance characteristics and historical process experience. For example, a piecewise linear function can be used. When the residual stress is higher than a preset threshold, a temperature compensation coefficient less than 1 is output to reduce the target temperature or heating rate of the corresponding lead sealing zone. When the residual stress is within a safe range, a compensation coefficient close to 1 is output to maintain the original temperature setting, thus achieving fine-grained temperature compensation for lead sealing based on the residual stress distribution. By storing the above node information, relationships, and mapping functions in a graph data structure, a process parameter relationship graph is constructed. During process execution, the latest monitoring data is used to dynamically correct the node attribute values ​​and associated edge weights, enabling the graph to reflect changes in parameter relationships under different material batches, environmental conditions, and process states.

[0040] After obtaining the process parameter correlation map, the residual stress distribution data measured after the straightening process is spatially calibrated and discretized, mapping the residual stress data to the corresponding cable structure zones. Simultaneously, heating zone information for the lead sealing process is introduced into the process parameter correlation map, establishing a one-to-one spatial correspondence between the straightening process zones and the lead sealing process zones. Based on this, the received residual stress values ​​are calculated according to the mapping functions stored in the mapping nodes of the process parameter correlation map, generating temperature compensation coefficients for each zone of the lead sealing process, used to achieve precise compensation of the lead sealing temperature. Finally, based on the insulation shielding layer defect characteristics in the multi-source monitoring data, a bidirectional process feedback mechanism is designed. During the lead sealing process, the location, size, and number of pore defects in the insulation shielding layer are extracted in real time from multi-source monitoring data to form defect features. These defect features are then matched with spatial partitions in the process parameter correlation map. Based on the defect size and distribution density, a preset defect-pressure segmentation mapping rule is invoked to automatically convert the pore defect features into pressure correction increments or holding time corrections for the corresponding regions in the straightening process. For example, when a large-sized or high-density pore defect is detected, the pressure peak value of the main pressure stage in the corresponding straightening region is automatically reduced, and the holding time of the fine pressure stage is extended. At the same time, abnormal deformation or residual stress concentration areas detected in the straightening process can also be fed back to the lead sealing process to dynamically correct the power settings of the subsequent lead sealing heating zones, forming a two-way constraint and compensation mechanism between lead sealing and straightening.

[0041] Then, a neural network model was trained using historical shared data from the lead sealing and straightening processes. This neural network model employed a multilayer feedforward neural network (MLP) architecture for training. The training process used historical shared data as the sample set to complete a nonlinear regression mapping from temperature compensation coefficients and pressure correction increments to the optimization parameters of the lead sealing and straightening processes. Specifically, the temperature compensation coefficients of each heating zone in the lead sealing process were first arranged in a fixed order to form an input vector xT. The pressure correction increments of the straightening process in the pre-pressing, main press, and fine press stages and their corresponding spatial regions were arranged in a fixed order to form an input vector xP. These two vectors were then concatenated with the initial parameters to form a unified input x. The output was also arranged in a fixed order to form a target vector y, containing the key control parameters of the lead sealing optimized temperature curve and the straightening optimized parameter set, ensuring that each historical record corresponds to a set of (x, y) training samples. To ensure training stability, both input and output were standardized, and outlier samples were subjected to amplitude limiting to prevent the model from being influenced by a small number of abnormal data points.

[0042] This neural network model employs a fully connected configuration consisting of an input layer, three hidden layers, and an output layer to balance expressive power and deployment stability. The input layer has a dimension of x (the number of parameters), with 128 neurons in the first hidden layer, 64 neurons in the second hidden layer, and 32 neurons in the third hidden layer. The output layer has m neurons. The hidden layer activation function is ReLU, although LeakyReLU can be used to improve sparse gradients. The output layer uses linear activation to satisfy continuous parameter regression output. To suppress overfitting, Dropout is added after the first and second hidden layers with a dropout rate of 0.2, and L2 regularization is introduced into the weights of each fully connected layer with a weight decay coefficient of 1×10⁻⁶. -4 This ensures the model maintains its generalization ability under perturbations such as different material batches and varying ambient temperatures. Weights are initialized using He initialization, and biases are initialized to 0 to ensure stable variance during the initial forward propagation. To enhance training convergence speed, batch normalization can be added after the first or second hidden layer to reduce internal covariate bias. The momentum parameter for batch normalization is set to 0.9, and the numerical stability term... Take 1×10 -5 During training, historical shared data is grouped and segmented in batches to avoid falsely high validation values ​​due to leaks under the same operating conditions. The loss function uses weighted mean square error to reflect the importance of key process parameters. Weighting coefficients are set for outputs related to sealing temperature and straightening pressure. For example, the key parameter of the sealing temperature curve has a weight of 1.0, the straightening pressure and holding time parameters have a weight of 1.2, and safety-related parameters can have a higher weight of 1.5, thus ensuring that the model prioritizes thermal and mechanical safety. The optimizer is Adam, with an initial learning rate of 1×10⁻⁶. -3 The coefficients of the first and second moments are set to 0.9 and 0.999, respectively. The maximum number of training epochs is set to 300. A learning rate decay strategy is configured, where the learning rate is multiplied by 0.5 until it is no less than 1 × 10⁻⁶ when the validation set loss shows no improvement within 10 consecutive epochs. -5 To ensure fine-grained convergence in later stages, early stopping is introduced during training to further guarantee deployability. If the validation set loss shows no improvement within 20 consecutive epochs, training is stopped, and the model is rolled back to the weight parameters with the minimum validation set loss. Gradient clipping can be set to a global norm of 5.0 to avoid gradient explosion caused by individual batches. After training, the prediction error is evaluated using an independent test set, such as the MAE of each output dimension, and models that meet the preset error threshold are solidified.

[0043] In the actual process execution, the initial lead sealing optimization temperature, initial straightening optimization parameters, and corresponding temperature compensation coefficients and pressure correction increments under the current operating conditions are input into a trained neural network model to obtain a collaborative optimization output. This collaborative optimization output is then cross-corrected with the initial parameters. Specifically, the target temperature, heating rate, and holding time for the lead sealing side are compensated and corrected according to the zone, generating a corrected lead sealing temperature curve. Similarly, the pressure peak, loading rate, holding time, and temperature gradient requirements for the straightening side are corrected according to the stage, generating corrected straightening parameters. Immediately after correction, the process enters the online verification stage. In this stage, the latest multi-source monitoring data is used to recalculate the thermo-mechanical-electric coupling state, checking whether constraints such as temperature peak not exceeding the heat resistance threshold, temperature gradient being controlled, residual stress not exceeding limits, and defect risk indicators decreasing are met. If these constraints are not met, the compensation coefficients and correction increments are adjusted and re-inferred according to a preset number of iterations until the constraints are satisfied. The final output includes optimized temperature curves for lead sealing and optimized parameters for actual control, thereby achieving coordinated control of lead sealing and straightening processes at the system level, reducing the risk of porosity defects and stress concentration, and improving the stability of the polypropylene cable accessory manufacturing process and the consistency of finished products.

[0044] Furthermore, the coordinated optimization of the process parameters also includes: A multi-objective process evaluation system is constructed, including process stability indicators, comprehensive energy efficiency indicators, and product defect rate indicators. An evolutionary algorithm is used to globally optimize the prediction parameters of the neural network model, solve for the Pareto optimal frontier solution set under the multi-objective process evaluation system, and dynamically update the optimal process parameter database. An abnormal operating condition response strategy library is established based on the multi-source monitoring data. When real-time monitoring data indicates batch fluctuations or abnormalities in material performance, the system automatically matches and switches from the optimal process parameter database to the backup set of optimized parameters.

[0045] Optionally, after the neural network model predicts the parameters for the coordinated optimization of lead sealing and straightening, a multi-objective process evaluation system will be constructed. This involves defining core evaluation indicators based on historical process data and real-time monitoring data. These core indicators include process stability indicators, comprehensive energy efficiency indicators, and product defect rate indicators. The process stability indicator reflects the fluctuation and repeatability of the lead sealing and straightening process, specifically including the standard deviation of temperature in the lead sealing area, the amplitude of temperature gradient fluctuations, the repeatability deviation of straightening pressure loading, and the dispersion of residual stress. A comprehensive stability score is formed through normalized weighted summation. The comprehensive energy efficiency indicator reflects the energy consumption level corresponding to a unit of qualified product. Specifically, it calculates the unit output energy consumption value using parameters such as total energy consumption for lead sealing heating, energy consumption for straightening loading, and process cycle time, and normalizes it based on the target energy consumption benchmark. The product defect rate indicator reflects the final product quality risk, specifically obtained through statistics on the occurrence rate of porosity defects detected during the lead sealing stage, the number of defects in the insulation shielding layer, and the dimensional deviation rate after straightening. The three types of indicators mentioned above serve as independent objective functions in multi-objective optimization. The process stability and energy efficiency indicators aim to be maximized, while the product defect rate indicator aims to be minimized, thus forming a multi-objective process evaluation system. Subsequently, an evolutionary algorithm is used to globally optimize the process parameters output by the neural network model. Specifically, the sealing optimization temperature curve parameters and straightening optimization parameters predicted by the neural network are used as initial population individuals. For each individual's corresponding parameter combination, the multi-objective process evaluation system is used to calculate its process stability score, energy efficiency score, and defect rate value. Then, new candidate parameter combinations are generated in the parameter space through selection, crossover, and mutation operations. The selection operation prioritizes individuals that perform better in the multi-objective sense; the crossover operation combines the sealing and straightening parameter characteristics of different individuals; and the mutation operation applies small-scale random perturbations to parameters such as temperature, pressure, and time within safe constraints to enhance the global search capability. In each generation of evolution, non-dominated solutions are screened according to the Pareto dominance relationship, gradually approaching the Pareto optimal front solution set that achieves the optimal balance among process stability, energy efficiency, and defect rate. As the number of generations increases, the Pareto optimal solution set is continuously updated, and the parameter combinations that satisfy the engineering constraints and quality constraints are stored in the optimal process parameter database. At the same time, the corresponding material batch characteristics, environmental conditions and process applicability range are labeled to realize the dynamic accumulation and updating of parameters.

[0046] Subsequently, an abnormal operating condition response strategy library was established based on multi-source monitoring data. Online analysis was performed on raw material batch parameters, real-time temperature response characteristics, pressure-deformation response characteristics, and defect evolution trends. When batch-specific fluctuations or anomalies in material properties were detected—for example, a significant deviation of the temperature rise rate from the historical range under the same heating power, delayed deformation response or premature creep during straightening, or a significant increase in the probability of defect occurrence—the current operating condition was determined to be abnormal. Then, based on the anomaly type, degree, and corresponding characteristic parameters, the historical operating condition tags that best matched the current anomaly characteristics were retrieved from the optimal process parameter database using cosine similarity. A pre-validated set of backup optimized parameters was selected from this set. Next, without interrupting production, the current lead sealing and straightening control parameters were smoothly switched to the backup optimized parameter set, while simultaneously increasing the monitoring frequency. The switched process state was then rapidly evaluated and corrected, thereby achieving rapid response and risk mitigation for abnormal operating conditions. Through the collaborative mechanism of multi-objective process evaluation, global optimization by evolutionary algorithm, and adaptive switching under abnormal operating conditions, the lead sealing and straightening processes can maintain a stable, low-energy-consumption, and low-defect-rate operating state under different material batches and environmental conditions. At the same time, it can achieve continuous accumulation and dynamic evolution of optimal process parameters, providing a long-term optimizable intelligent process control foundation for the production of polypropylene cable accessories.

[0047] The lead sealing and straightening of polypropylene cable accessories are controlled based on the optimized lead sealing temperature and straightening parameters.

[0048] In one embodiment, after obtaining the optimized temperature curve for lead sealing and the optimized parameter set for straightening for actual control, these are sent to the lead sealing control unit and the straightening control unit to implement closed-loop control of the polypropylene cable accessory manufacturing process. Specifically, in the lead sealing control, the target temperature, heating rate, holding time, and cooling rhythm of each heating zone are set according to the optimized temperature curve for lead sealing, and the temperature curve is converted into a corresponding heating power command. This can be achieved by partition power ratio, pulse duty cycle, or continuous power output. During the lead sealing process, the temperature distribution, temperature gradient, and heat flux density of the lead sealing area are collected in real time and compared with the target temperature curve. When a local temperature is too high, the temperature gradient exceeds the limit, or approaches the material's heat resistance threshold, control actions such as power attenuation, partition cooling, or heating suspension are automatically triggered to keep the heat-affected zone of the lead sealing within a safe range and suppress the generation of porosity defects. At the same time, the detection results of insulation shielding layer defects are continuously received. When abnormal defect characteristics are detected, secondary temperature compensation is performed on the corresponding heating zone according to preset mapping rules to form an online adaptive correction of the lead sealing process. In the straightening control, the straightening equipment is controlled in stages according to the optimized straightening parameter set. The pressure loading process is decomposed into control commands for the pre-pressing stage, the main pressure stage, and the fine pressing stage, and the target pressure value, loading rate, holding time, and temperature and temperature gradient requirements are set for each stage. During the straightening process, pressure feedback, deformation displacement, and temperature data are collected in real time, and the actual deformation is compared with the target deformation. When the deviation exceeds the preset threshold, the holding time, pressure peak, or loading rate is automatically adjusted. At the same time, the heating intensity or holding time is corrected according to the temperature gradient requirements to reduce the risk of springback and residual stress concentration. After straightening, the residual stress distribution and final deformation results are output as the basis for process quality judgment and written back to the process database to provide data support for the continuous optimization of subsequent process parameters. Through the above-mentioned collaborative closed-loop control of lead sealing and straightening, the heat input and stress process of polypropylene cable accessories are controllable and traceable, thereby improving molding consistency, reducing the risk of porosity defects and deformation damage, and improving the overall manufacturing quality and reliability of accessories.

[0049] Furthermore, straightening control based on straightening optimization parameters includes: The optimal pressure loading sequence is decomposed into pre-pressure stage control commands, main pressure stage control commands, and fine pressure stage control commands to perform staged control of the straightening equipment; real-time feedback data from the straightening equipment is received, and when the deviation between the actual deformation and the expected deformation exceeds the deviation threshold, the online recalculation of the coupled dynamics model is triggered to dynamically adjust the pressure loading rate and holding time of each stage.

[0050] Optionally, the optimal pressure loading sequence in the straightening optimization parameters is first analyzed and discretized. The continuous pressure-time curve is divided into a pre-pressing stage, a main pressure stage, and a fine-pressing stage according to the pressure change characteristics and process objectives. In the pre-pressing stage, control commands are generated, including the pre-pressing starting pressure, the target pre-pressing pressure, the pressure loading rate, and the pre-pressing duration. These commands drive the straightening equipment to fully contact the straightening clamp with the cable and eliminate clamping gaps using a low-pressure, slow-loading method. In the main pressure stage, control commands are generated, including the main pressure target value, the main pressure loading rate, and the main pressure holding time. These commands are used to complete the main deformation correction of the cable under the temperature and stress conditions allowed by the material. In the fine-pressing stage, control commands are generated, including the fine-pressing pressure fine-adjustment range, the fine-pressing loading rate, and the fine-pressing holding time. These commands are used to refine the local deformation and gradually release residual stress. The above-mentioned staged control commands are issued to the straightening equipment in a predetermined order, and the actuators of the straightening equipment execute them segment by segment. During the phased control process of the straightening equipment, real-time feedback data is continuously received from the straightening equipment at a fixed sampling period. This feedback data includes at least the current applied pressure value, the rate of pressure change, the real-time deformation displacement of the straightening area, and the corresponding temperature data. Based on this feedback data, the actual deformation curve is calculated and compared with the expected deformation curve predicted according to the optimal pressure loading sequence and the coupled dynamics model. The deformation deviation between the two at the same time point is calculated. When the deformation deviation is less than a preset deviation threshold, the control command for the current stage remains unchanged, and the predetermined pressure loading process continues. When the deformation deviation is detected to exceed the preset deviation threshold, or when the deviation shows an increasing trend over multiple consecutive sampling periods, it is determined that the current straightening process has a deviation risk, and the online recalculation of the coupled dynamics model is automatically triggered. During the online recalculation, the real-time state of the current straightening stage is used as the new initial conditions, including the applied actual pressure, the current temperature level, the generated deformation displacement, and the existing residual stress state. The remaining straightening process is then resubmitted into the coupled dynamics model for solution. The updated deformation evolution prediction results and residual stress change trends are obtained through calculation, and the pressure loading strategy for subsequent straightening stages is re-evaluated accordingly. Subsequently, based on the online recalculation results, the control parameters for each stage that has not yet been completed are dynamically corrected. This correction includes, but is not limited to, reducing or increasing the pressure loading rate of subsequent stages, extending or shortening the pressure holding time of corresponding stages, adjusting the pressure fine-tuning range of the fine-tuning stage, and, if necessary, fine-tuning the target pressure value of the current stage, so that the subsequent actual deformation trajectory gradually returns to the expected range. Finally, the corrected staged control commands are sent to the straightening equipment in real time to continue the straightening process. The above-mentioned closed-loop control process of "real-time feedback acquisition - deformation deviation judgment - online model recalculation - dynamic correction of control parameters" is executed cyclically throughout the entire straightening process until the straightening is completed.This phased control and online dynamic adjustment mechanism enables the straightening process to remain stable and controllable under different material conditions and temperatures, thereby effectively suppressing springback and residual stress concentration while ensuring straightening accuracy, and improving the reliability and consistency of the straightening process for polypropylene cable accessories.

[0051] In summary, the embodiments of this application have at least the following technical effects: First, multi-source monitoring data is collected in real time using a sensor array. This data includes the temperature distribution in the lead-sealing area, temperature and pressure data in the straightening area, and defect characteristics of the insulation shielding layer. Then, based on the multi-source monitoring data and cable material parameters, a multi-physics coupling model is constructed to generate the thermo-mechanical-electrical coupling state in the lead-sealing and straightening processes in real time. Next, based on this thermo-mechanical-electrical coupling state, the lead-sealing process is adjusted and the straightening process is optimized, generating initial optimized lead-sealing temperature and initial optimized straightening parameters. Then, the real-time data of the lead-sealing and straightening processes are shared and cross-corrected using the shared data to generate optimized lead-sealing temperature and optimized straightening parameters. Finally, lead-sealing and straightening control of polypropylene cable accessories is performed based on these optimized temperatures and straightening parameters. This solves the technical problems of uneven thermal stress and material damage caused by independent control of the lead-sealing and straightening processes, achieving the technical effect of synergistic optimization of the lead-sealing and straightening processes through real-time data sharing and cross-correction, thereby improving the quality and performance of cable accessories.

[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A collaborative manufacturing optimization method for insulated cable accessories, characterized in that, The method includes: Multi-source monitoring data is collected in real time through a sensor array. The multi-source monitoring data includes the temperature distribution of the lead-sealed area, the temperature and pressure data of the straightening area, and the defect characteristics of the insulation shielding layer. Based on the multi-source monitoring data and cable material parameters, a multi-physics coupling model is constructed to generate the thermo-mechanical-electric coupling state in the lead sealing and straightening process in real time. Based on the aforementioned thermo-mechanical-electric coupling state, the lead sealing process is controlled and the straightening process is optimized, generating the initial lead sealing optimization temperature and the initial straightening optimization parameters. The real-time data of the sealing and straightening processes are shared and interactively, and the shared data is used to cross-correct the initial sealing optimization temperature and the initial straightening optimization parameters to generate the sealing optimization temperature and straightening optimization parameters. The lead sealing and straightening of polypropylene cable accessories are controlled based on the optimized lead sealing temperature and straightening parameters.

2. The method for optimizing the collaborative manufacturing of insulated cable accessories as described in claim 1, characterized in that, Based on the multi-source monitoring data and cable material parameters, a multi-physics coupling model is constructed to generate the thermo-mechanical-electric coupling state in the lead sealing and straightening process in real time, including: Based on the temperature distribution data and the thermal properties of the cable material from the multi-source monitoring data, a transient temperature field model of the lead-sealed area is established. Based on the temperature and pressure data of the straightening region and the mechanical property parameters of the cable material in the multi-source monitoring data, an elastic-plastic stress field model of the straightening region is established. Based on the defect characteristics of the insulation shielding layer and the dielectric properties of the cable material, a three-dimensional electric field distribution model of the insulation shielding layer is established. The transient temperature field model, elastoplastic stress field model, and three-dimensional electric field distribution model are bidirectionally coupled and iteratively calculated using a multi-field coupling solver, and the thermo-mechanical-electric coupling state during the lead sealing and straightening process is output in real time.

3. The method for collaborative manufacturing optimization of insulated cable accessories as described in claim 2, characterized in that, The thermo-mechanical-electric coupling state includes the temperature gradient distribution and heat flux density of the lead-sealed region, the residual stress distribution and deformation displacement of the straightening region, and the changes in the electric field strength and dielectric constant of the insulating shielding layer.

4. The method for optimizing the collaborative manufacturing of insulated cable accessories as described in claim 3, characterized in that, Based on the aforementioned thermo-mechanical-electric coupling state, the lead sealing process is controlled to generate an initial optimized lead sealing temperature, including: Based on the temperature gradient distribution and heat flux density in the aforementioned thermo-mechanical-electric coupling state, a response matrix between the temperature field and the heating power is constructed. Based on the response relationship matrix, the heating power zones of the lead sealing area are dynamically divided, and a zone heating control strategy is generated. Based on the changes in electric field strength and dielectric constant of the insulating shielding layer in the thermo-mechanical-electric coupling state, an anti-interference control logic is designed to automatically trigger the power attenuation mechanism of adjacent heating zones when an abnormal electric field strength is detected. Based on the correlation data between heating trajectory and defects in the historical process database, a heating path optimization knowledge base is established, and the heating mode with the lowest risk of porosity defects is selected first for path planning to generate an optimized heating path. By combining the aforementioned zoned heating control strategy, anti-interference control logic, and optimized heating path, an initial lead sealing optimized temperature curve is generated.

5. The collaborative manufacturing optimization method for insulated cable accessories as described in claim 4, characterized in that, The dynamic division of the heating power zone for the lead sealing area includes: Based on the temperature gradient distribution of the lead-sealed area in the thermo-mechanical-electric coupling state, the lead-sealed area is divided into a core heating zone, a transition heating zone, and an outer protection zone. Based on the heat flux density distribution of the lead-sealed region in the thermo-mechanical-electric coupling state, the required heat input power ratio of each region is calculated, and the initial power setting value of each region is determined. A temperature buffer zone is set up within the outer protection zone, and a graded cooling protocol is configured. When the local temperature is detected to be close to the material's heat resistance threshold, the stepped cooling program is automatically activated. A sliding window is used to process real-time temperature data and dynamically update the boundary range of each heating zone.

6. The method for optimizing the collaborative manufacturing of insulated cable accessories as described in claim 1, characterized in that, Based on the aforementioned thermo-mechanical-electric coupling state, the straightening process is optimized to generate initial straightening optimization parameters, including: Based on the residual stress distribution and deformation displacement in the aforementioned thermo-mechanical-electric coupling state, a coupled dynamic model of material deformation and applied pressure during the straightening process is established. Based on the coupled dynamics model, the optimal pressure loading sequence that satisfies the objectives of deformation uniformity and residual stress minimization is calculated; Based on the temperature data of the calibration region in the thermo-mechanical-electric coupling state, an adaptive calibration control strategy is designed to automatically adjust the duration and temperature gradient requirements of the pressure holding stage when the material creep characteristics are detected to change. Based on the mapping relationship between historical deformation data and final dimensional error, a deformation error prediction model is established, and the pressure loading sequence is pre-compensated and corrected according to the prediction results. By combining the optimal pressure loading sequence, adaptive straightening control strategy, and pre-compensation correction, an initial straightening optimization parameter set containing pressure, time, and temperature parameters is generated.

7. The collaborative manufacturing optimization method for insulated cable accessories as described in claim 6, characterized in that, Based on the coupled dynamics model, the optimal pressure loading sequence that satisfies both the deformation uniformity objective and the residual stress minimization objective is calculated, including: Based on the aforementioned coupled dynamics model, with the objectives of maximizing deformation uniformity and minimizing residual stress, an optimization algorithm is used to solve the problem and obtain the Pareto optimal solution set. The optimal solution is selected from the Pareto optimal solution set to generate a segmented pressure loading curve that includes the pre-compression stage, the main pressure stage, and the fine pressure stage, which serves as the optimal pressure loading sequence.

8. The collaborative manufacturing optimization method for insulated cable accessories as described in claim 1, characterized in that, The real-time data exchange and sharing between the lead sealing and straightening processes is utilized, and the shared data is used to cross-correct the initial lead sealing optimization temperature and initial straightening optimization parameters to generate the lead sealing optimization temperature and straightening optimization parameters, including: Establish a correlation graph of process parameters, and convert the measured residual stress distribution data after the straightening process into temperature compensation coefficients for each zone of the lead sealing process through a mapping function; Based on the defect characteristics of the insulation shielding layer in the multi-source monitoring data, a two-way process feedback mechanism is designed to automatically convert the location and size data of the pore defects detected in the lead sealing process into the pressure correction increment in the corresponding area of ​​the straightening process. A neural network model is trained using historical shared data. This model takes preceding process parameters and real-time monitoring data as input and subsequent process optimization parameters as output to perform collaborative optimization of process parameters. The trained neural network model is used to cross-correct and verify the initial lead sealing optimization temperature and the initial straightening optimization parameters, generating the lead sealing optimization temperature curve and the straightening optimization parameter set for actual control.

9. The collaborative manufacturing optimization method for insulated cable accessories as described in claim 8, characterized in that, The collaborative optimization of the process parameters also includes: Construct a multi-objective process evaluation system that includes process stability indicators, comprehensive energy efficiency indicators, and product defect rate indicators; An evolutionary algorithm is used to globally optimize the prediction parameters of the neural network model, solve the Pareto optimal frontier solution set under the multi-objective process evaluation system, and dynamically update the optimal process parameter database. An abnormal operating condition response strategy library is established based on the multi-source monitoring data. When real-time monitoring data indicates that the material properties exhibit batch-to-batch fluctuations or abnormalities, the system automatically matches and switches from the optimal process parameter database to the backup set of optimized parameters.

10. The method for optimizing the collaborative manufacturing of insulated cable accessories as described in claim 7, characterized in that, Straightening control based on straightening optimization parameters includes: The optimal pressure loading sequence is decomposed into pre-pressure stage control commands, main pressure stage control commands, and fine pressure stage control commands to perform staged control of the straightening equipment; The system receives real-time feedback data from the alignment equipment. When the deviation between the actual deformation and the expected deformation exceeds the deviation threshold, it triggers the online recalculation of the coupled dynamics model and dynamically adjusts the pressure loading rate and holding time at each stage.

Images