Preparation management method and system for aviation atomic oxygen-resistant superconducting mixed metal nickel fiber

Abstract

The invention discloses a preparation management method and system for aviation atomic oxygen-resistant superconductive mixed metal nickel fibers, and relates to the technical field of nickel fiber preparation management, and the method comprises the steps: obtaining a plurality of groups of process response data under different conditions under each adjustment working condition, determining the optimal dynamic control range of each interaction influence parameter in a core process link, and obtaining the optimal dynamic control range; taking the edge process parameters of the core process link as the associated reference parameters of the next link to be calibrated, sequentially determining the cooperative control value of each interactive influence parameter in each process link, completing the cooperative calibration of the interactive parameters of all process links except the raw material pretreatment link, and taking the final target of the preparation process as the reference. And performing global consistency adjustment on the interactive influence parameters calibrated in each link. The management system realizes high-performance stable preparation of the aviation atomic oxygen-resistant superconducting mixed metal nickel fiber through a method of combining multi-parameter independent calibration, cyclic collaborative optimization and dynamic self-adaptive regulation and control.

Description

Technical Field

[0001] This invention relates to the field of nickel fiber preparation and management technology, and specifically to a method and system for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers. Background Technology

[0002] In the aerospace field, superconducting hybrid nickel fibers, as key functional materials, must possess excellent superconducting properties and resistance to atomic oxygen erosion to meet the stringent requirements of the space environment. However, in existing fabrication technologies, core characteristic parameters (such as melt temperature and tensile tension) and parameters that interact with process steps (such as thermal gradient and cooling rate matching) are often controlled in isolation, leading to inconsistencies between microstructural defects and macroscopic properties, especially rapid performance degradation under atomic oxygen exposure. Given the extremely high requirements for atomic oxygen resistance and superconducting stability in aerospace-grade superconducting fibers, traditional single-parameter optimization methods cannot meet the needs of multi-parameter coupled synchronous control, and conventional fabrication processes cannot achieve precise management of the entire process from basic properties to process coordination.

[0003] Based on this need, this patent application proposes a method that combines multi-parameter independent calibration, cyclic collaborative optimization, and dynamic adaptive control to achieve high-performance and stable preparation of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for the preparation and management of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, in order to address the shortcomings in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, the method comprising the following steps: Step A. Calculate the characteristic mapping relationship corresponding to each type of characteristic parameter based on multiple sets of experimental data, and combine the characteristic mapping relationships of the same characteristic parameter under different experimental conditions into a set of mathematical equations. Solve the set of mathematical equations to determine the optimal control range of the characteristic parameter. Step B. Perform cyclical collaborative calibration of the interaction parameters between various process steps in the preparation process: Step B1. Obtain multiple sets of process response data under different conditions in each adjusted operating condition, and determine the optimal dynamic control range of each interactive parameter in the core process link; Step B2. Using the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, determine the collaborative control values ​​of each interactive parameter in each process step in turn, and complete the collaborative calibration of the interactive parameters of all process steps except the raw material pretreatment step. Step B3. Based on the final goal of the preparation process, adjust the interaction parameters of each step of the calibration to ensure global consistency. Step C. Collect historical data of characteristic parameters and interaction parameters throughout the entire process, and combine them with the final fiber performance characterization results to construct a dynamic adaptive control model. When a set of parameters is detected to fall into the historical risk range, the control command is automatically triggered.

[0006] In one embodiment disclosed in this application, step B1, obtaining multiple sets of process response data under different conditions in each adjusted operating condition, and determining the optimal dynamic control range of each interactive parameter in the core process link, includes the following steps: Select the core process steps in the preparation process, and simultaneously collect multiple sets of transition state data under different process conditions in the core process steps; The stability of the melt solidification interface and the dynamic fluctuation of the drawing tension are monitored in real time by sensors, and the process parameters are actively adjusted. Under each adjusted operating condition, multiple sets of process response data under different conditions are obtained, and the characteristic mapping relationship corresponding to each set of process response data is calculated. By combining the calibrated characteristic parameters, a system of equations is established and solved to determine the optimal dynamic control range of each interactive parameter in the core process.

[0007] In one embodiment disclosed in this application, step B2, using the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, sequentially determines the collaborative control values ​​of each interactive parameter in each process step, and completes the collaborative calibration of interactive parameters of all process steps except the raw material pretreatment step, including the following steps: Following the logical sequence of the preparation process, the parameters of adjacent downstream steps are used as new parameters to be calibrated. By combining the same multi-condition data acquisition and characteristic mapping solutions, the collaborative control values ​​of each interactive parameter in each process step are determined sequentially.

[0008] In one embodiment disclosed in this application, step C. collecting historical data of characteristic parameters and interaction parameters throughout the entire process, and combining this data with the final fiber performance characterization results, constructs a dynamic adaptive control model. When a set of parameter combinations is detected to fall into the historical risk range, a control command is automatically triggered, including the following steps: Multivariate correlation analysis techniques are used to uncover the correlation patterns between characteristic parameters and interaction parameters. Based on the results of this cross-dimensional correlation analysis, a dynamic adaptive control model is constructed to monitor characteristic parameters and interaction parameters in real time. When a combination of parameters is detected to fall into the historical risk range, control commands are automatically triggered, including reducing the pulling rate or increasing the cooling airflow intensity.

[0009] In one embodiment disclosed in this application, step A. calculating the characteristic mapping relationship corresponding to each type of characteristic parameter based on multiple sets of experimental data includes the following steps: For each type of characteristic parameter, multiple preparation experiments were conducted under different process conditions, and the preparation process parameters and initial fiber performance characterization data were recorded simultaneously. Based on the quantitative functional relationship between the changes of multiple characteristic parameters and fiber performance indicators, the characteristic mapping relationship of the same characteristic parameter under different experimental conditions is combined into a system of mathematical equations; The optimal control range of characteristic parameters is determined by solving a system of mathematical equations.

[0010] In one embodiment disclosed in this application, the characteristic parameters include the proportion of raw material components, melting environment temperature, melt flow rate, purity of inert gas protection, and initial drawing tension.

[0011] In one embodiment disclosed in this application, multiple sets of process response data under different conditions are obtained under each adjustment condition. The characteristic mapping relationship corresponding to each set of process response data is calculated. Combined with the calibrated characteristic parameters, the process includes the following steps: Extract real-time monitoring values ​​of interactive influence parameters and fiber transition state performance indicators, and analyze the correlation trend between the two; Combining the calibrated characteristic parameters, the characteristic mapping relationship of the interaction parameters and the characteristic parameter constraints are combined into a system of equations: The characteristic parameters are used as boundary constraint inputs, the interaction influence parameters are used as independent variables, and the fiber transition state performance index is used as the dependent variable. The nonlinear relationship between the parameters is learned through the training set data. Set a target threshold for the transition state performance index, and then deduce the range of values ​​for the interaction parameters that satisfy this threshold.

[0012] In one embodiment disclosed in this application, the coordinated control values ​​of each interactive parameter in each process step are determined sequentially by simultaneously acquiring and solving the same multi-condition data and characteristic mapping, including the following steps: Multiple sets of process response data were obtained under each adjusted operating condition, and the characteristic mapping relationship of the interactive influence parameters was calculated. By combining the upstream reference parameters with the calibrated characteristic parameters, a system of multi-parameter equations is established. Set the fiber performance transfer target for this stage, and deduce the range of values ​​for the parameters to be calibrated to meet the target.

[0013] In one embodiment disclosed in this application, step B3 involves globally consistent adjustment of the interaction parameters calibrated at each stage, based on the final goal of the preparation process, including the following steps: Summarize the actual values ​​of the interaction parameters at each stage and analyze their correlation with the final performance indicators; If a systematic deviation is detected due to the optimization of parameters in a local process, the parameter values ​​are adjusted through a dynamic weight allocation algorithm to eliminate parameter conflicts between processes.

[0014] This application also provides a management system for the preparation of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, including: Characteristic parameter optimization module: Based on multiple sets of experimental data, calculate the characteristic mapping relationship corresponding to each type of characteristic parameter, put the characteristic mapping relationship of the same characteristic parameter under different experimental conditions into a set of mathematical equations, and solve the set of mathematical equations to determine the optimal control range of the characteristic parameter. Global Adjustment Module: Performs cyclical collaborative calibration of the interaction parameters between various process steps in the preparation process: Under each adjustment condition, obtains multiple sets of process response data under different conditions, determines the optimal dynamic control range of each interaction parameter in the core process step, uses the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, sequentially determines the collaborative control value of each interaction parameter in each process step, completes the collaborative calibration of the interaction parameters of all process steps except the raw material pretreatment step, and adjusts the global consistency of the calibrated interaction parameters of each step based on the final goal of the preparation process. Control module: Collects historical data of characteristic parameters and interactive parameters throughout the entire process, and combines them with the final performance characterization results of the fiber to construct a dynamic adaptive control model. When a set of parameters is detected to fall into the historical risk range, the control command is automatically triggered.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention uses historical data mining and a dynamic adaptive control model to identify high-risk parameter combinations (such as the combination of excessively high melting temperature and excessively fast drawing rate) in real time, and automatically triggers control commands (such as dynamically adjusting cooling intensity or drawing speed) when an anomaly is detected. This reduces the atomic oxygen erosion rate, decreases the superconducting performance decay rate, and improves the preparation yield. It breaks through the limitations of isolated and static adjustment of parameters in traditional preparation technologies, and ultimately achieves high-performance stable output and long-life reliable service of aerospace-grade superconducting fibers in extreme environments.

[0016] This invention establishes a quantitative mapping relationship between characteristic parameters and key fiber performance indicators (such as superconducting critical current density and atomic oxygen erosion rate), and solves the optimal control range to achieve precise and independent control of basic characteristic parameters such as melting temperature and tensile tension. This reduces microstructural defects (such as lattice distortion and abnormal porosity) caused by parameter deviations from the source, and provides a reliable guarantee for the intrinsic properties of the fiber.

[0017] This invention employs a cyclical collaborative calibration approach. It first focuses on the dynamic optimization of interactive parameters such as thermal gradient and cooling rate matching in core process steps (e.g., the initial drawing stage). Then, it iteratively calibrates the collaborative values ​​of parameters in each step along the preparation process, ultimately achieving consistent matching of process parameters across the entire chain. This effectively solves the problem of parameter conflicts between steps in traditional step-by-step control (e.g., the exacerbation of lattice defects caused by excessively fast drawing rate after high-temperature melting), and significantly improves the macroscopic consistency of fibers. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart of the management method of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example: This example provides a method for the preparation and management of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers. Please refer to [link to relevant documentation]. Figure 1 As shown, the management method includes the following steps: A. Precise Analysis of Characteristic Parameters of Superconducting Hybrid Nickel Fibers In the preparation of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, key characteristic parameters (such as raw material composition ratio, melting environment temperature, melt flow rate, inert gas protection purity, and initial tensile tension) play a decisive role. For each type of characteristic parameter (such as the matching conditions of the synergistic relationship between melting temperature and melt flow rate), multiple sets of preparation experiments were conducted under different process conditions. The preparation process parameters and the initial performance characterization data of the fiber (such as room temperature resistivity measurements) were recorded simultaneously to ensure that multiple sets of comparable process response information could be obtained under each set of parameter conditions. Based on multiple sets of experimental data, the characteristic mapping relationship corresponding to each type of characteristic parameter was calculated, that is, the quantitative functional relationship between the change of characteristic parameter and the key performance index of the fiber (such as superconducting critical current density). The characteristic mapping relationship of the same characteristic parameter under different experimental conditions was combined into a set of mathematical equations (such as constructing a nonlinear correlation model through machine learning algorithms). By solving the set of mathematical equations, the optimal control range of the characteristic parameter was determined to complete the independent and accurate calibration of all characteristic parameters.

[0022] B. Dynamic calibration of coordinated parameters across multiple stages of the preparation process A cyclical collaborative calibration was performed on the interaction parameters between various process steps in the fabrication process (e.g., the thermal gradient change from the completion of melting to the initial drawing stage). These parameters directly affect the macroscopic consistency and overall resistance to atomic oxygen of the fiber. Specifically, this was divided into three progressive stages: (1) Select the most critical core process step in the preparation process (by manual input through an interactive interface (usually based on expert experience)). In this step, collect multiple sets of transition state data under different process conditions simultaneously. Monitor key process indicators such as the stability of the melt solidification interface and the dynamic fluctuation of the drawing tension in real time using high-precision sensors, and actively adjust process parameters (e.g., change the drawing rate). Obtain multiple sets of process response data under different conditions for each adjustment. Calculate the characteristic mapping relationship corresponding to each set of process response data. Combined with calibrated characteristic parameters (e.g., the optimized threshold of the melting temperature), solve the system of equations to determine the optimal dynamic control range of each interactive parameter in this core process step.

[0023] (2) The edge process parameters of the core process link (e.g., the exit tension of the initial drawing stage) are used as the reference parameters of the next link to be calibrated. Along the logical order of the preparation process, the key parameters of the adjacent downstream links (e.g., the online annealing temperature after fiber forming) are used as new parameters to be calibrated. Through the same multi-condition data acquisition and characteristic mapping solution, the collaborative control values ​​of each interactive parameter in each process link are determined in turn. Finally, the collaborative calibration of the interactive parameters of all key process links except the raw material pretreatment link at the front end is completed.

[0024] (3) Based on the final goal of the preparation process (e.g., the overall performance qualification rate of aerospace-grade atomic oxygen resistant superconducting fiber), the interactive parameters of each link are adjusted globally to eliminate systematic deviations caused by local optimization and ensure the coordinated matching of parameters throughout the entire process from raw materials to finished fibers.

[0025] C. Intelligent adaptive mechanism based on multi-parameter coupling Considering the cross-scale nonlinear interaction between characteristic parameters (e.g., melt temperature) and interaction parameters (e.g., drawing rate) (e.g., if the drawing rate is too fast after high-temperature melting, it will aggravate the lattice defects inside the fiber, thereby significantly reducing the resistance to atomic oxygen erosion), this application collects historical data of characteristic parameters and interaction parameters throughout the entire process, and combines them with the final performance characterization results of the fiber (e.g., surface morphology changes after atomic oxygen exposure, and the decay rate of the superconducting critical temperature), and uses multivariate correlation analysis techniques (e.g., principal component analysis PCA) to uncover the hidden correlation between the two types of parameters (e.g., it was found that when the melt temperature exceeds 1600℃ and the drawing rate is greater than 5m / s, the fiber atomic oxygen erosion rate increases exponentially). Based on the results of this cross-dimensional correlation analysis, a dynamic adaptive control model is constructed to monitor key characteristic parameters (such as the current melt temperature) and interactive parameters (such as the real-time drawing rate) in real time. When a combination of parameters is detected to fall into the historical high-risk range, control commands are automatically triggered (such as reducing the drawing rate or increasing the cooling airflow intensity). Through cross-dimensional correlation analysis and dynamic control, the adaptive capability of the preparation system is improved, ultimately ensuring the performance consistency and long-term reliability of aerospace atomic oxygen resistant superconducting hybrid nickel fiber.

[0026] This embodiment also provides a management system for the preparation and management of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, including: Characteristic parameter optimization module: Based on multiple sets of experimental data, calculate the characteristic mapping relationship corresponding to each type of characteristic parameter, put the characteristic mapping relationship of the same characteristic parameter under different experimental conditions into a set of mathematical equations, solve the set of mathematical equations to determine the optimal control range of the characteristic parameter, and send the optimal control range of the characteristic parameter to the global adjustment module and the control module. Global Adjustment Module: Performs cyclical collaborative calibration of the interaction parameters between various process steps in the preparation process: Under each adjustment condition, obtains multiple sets of process response data under different conditions, determines the optimal dynamic control range of each interaction parameter in the core process step, uses the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, sequentially determines the collaborative control value of each interaction parameter in each process step, completes the collaborative calibration of the interaction parameters of all process steps except the raw material pretreatment step, and uses the final goal of the preparation process as a benchmark to perform global consistency adjustment of the calibrated interaction parameters of each step, and sends the interaction parameters to the control module; Control module: Collects historical data of characteristic parameters and interactive parameters throughout the entire process, and combines them with the final performance characterization results of the fiber to construct a dynamic adaptive control model. When a set of parameters is detected to fall into the historical risk range, the control command is automatically triggered.

[0027] The following is a detailed explanation of all the steps involved in this application: Furthermore, the characteristic parameters are divided into single control parameters (such as melting ambient temperature and inert gas protective purity) and composite related parameters (such as the synergistic relationship between melting temperature and melt flow rate, and the matching conditions between protective gas purity and initial tensile tension). For each type of characteristic parameter (or parameter combination), multiple sets of control experiments under different process conditions are designed: by adjusting the values ​​of single parameters or parameter combinations (for example, setting the melting temperature to 1500℃, 1550℃, 1600℃, and 1650℃ respectively, while fixing other parameters within the baseline range), preparation experiments are carried out in a highly controlled laboratory environment.

[0028] During the experiment, preparation process parameters (such as real-time melt temperature fluctuation curve, melt flow rate monitoring value, inert gas flow and purity monitoring data, tensile tension sensor readings, etc.) and fiber initial performance characterization data (such as room temperature resistivity measurement value, microstructure scanning electron microscope image, phase composition ratio in X-ray diffraction pattern, etc.) were collected simultaneously to ensure that multiple sets of comparable process response information could be obtained under each set of parameter conditions (for example, repeating the experiment 3 times at the same melting temperature and recording the mean and standard deviation of resistivity each time).

[0029] Furthermore, the characteristic mapping relationship corresponding to each type of characteristic parameter is calculated based on multiple sets of experimental data. The specific processing logic is as follows: The values ​​of characteristic parameters (independent variables) and the measured values ​​of key fiber performance indicators (dependent variables, such as superconducting critical current density, atomic oxygen erosion rate, fiber tensile strength, etc.) are extracted from each experimental group. After data preprocessing (such as outlier removal and normalization), the correlation trend between the two is analyzed. For example, regarding the relationship between melting temperature and superconducting critical current density, the distribution law of fiber critical current density at different melting temperatures is statistically analyzed; regarding the composite parameter melting temperature and melt flow rate, the combined effect of the two on microstructure uniformity is analyzed. This process requires clarifying the correspondence between independent and dependent variables (e.g., the critical current density corresponding to a melting temperature of 1500℃ is XXX A / cm2), and forming a quantitative description of the correlation between characteristic parameters and performance indicators.

[0030] Furthermore, the characteristic mapping relationship of the same characteristic parameter (or combination of parameters) under different experimental conditions is combined into a system of mathematical equations, specifically: A nonlinear correlation model is constructed using data modeling methods. First, historical experimental data is divided into training and validation sets. The model is trained using the training set data to capture the complex nonlinear relationship between characteristic parameters and performance indicators (e.g., the secondary effect of the interaction term between melting temperature and melt flow rate on critical current density). The model's prediction accuracy is evaluated using the validation set (e.g., the mean square error is less than a set threshold). The final model is equivalent to a set of mathematical equations describing how changes in characteristic parameters quantitatively affect fiber performance (e.g., input melting temperature T and melt flow rate V, and the model outputs the predicted critical current density Jc).

[0031] Furthermore, the optimal control range of this characteristic parameter is determined by solving this system of mathematical equations: Based on the model's predictions, threshold values ​​for the fiber's target performance indicators are set (e.g., superconducting critical current density ≥ XXX A / cm², atomic oxygen erosion rate ≤ XXX nm / h). The range of characteristic parameters that satisfy these threshold values ​​is then derived (e.g., by traversing discrete points within the melting temperature range of 1400℃-1700℃ using a grid search method to select the temperature range that allows the critical current density to meet the target). For composite parameters (e.g., melting temperature-melt velocity), multiple parameter combinations need to be optimized simultaneously (e.g., determining that the fiber's overall performance is optimal when the melting temperature is 1600℃±20℃ and the melt velocity is XXX m / s±0.5 m / s). Finally, the optimal control range (including the center value and allowable fluctuation range) for each type of characteristic parameter (or parameter combination) is output.

[0032] Phase (1): Independent optimization and calibration of interaction parameters of core process links Based on expert experience or historical process data analysis (usually manually input via an interactive interface), the most critical core process steps (such as the initial drawing stage after melting) are identified from the entire preparation process. This stage is a key transition point in the transformation of fiber microstructure into macroscopic properties, and its interactive parameters (such as changes in thermal gradient, stability of melt solidification interface, dynamic fluctuations in drawing tension, etc.) directly affect the initial forming quality of the fiber and its adaptability to subsequent processes.

[0033] In this core process, a high-precision sensor network (e.g., infrared thermal imagers monitoring the temperature gradient at the melt solidification interface, strain gauges acquiring real-time tension fluctuations during drawing, and laser displacement sensors tracking melt flow rate changes) is used to simultaneously collect multiple sets of transition state data (i.e., process response data) under different process conditions. During the experiment, key process parameters are actively adjusted (e.g., changing the drawing rate and adjusting the cooling airflow intensity), and multiple sets of process response data under different conditions are obtained for each adjustment (e.g., repeating the experiment three times at the same drawing rate and recording the interface stability index and tension fluctuation range each time).

[0034] Furthermore, for each set of process response data, the corresponding characteristic mapping relationship is calculated: Real-time monitoring values ​​of interactive parameters (such as thermal gradient ΔT / μm and tension fluctuation amplitude ΔF / N) and fiber transition state performance indicators (such as solidification interface defect density and initial tensile uniformity coefficient) are extracted. After data preprocessing (such as sliding window smoothing and outlier removal), the correlation trend between the two is analyzed (for example, when the thermal gradient exceeds XXX℃ / μm, the interface defect density increases significantly).

[0035] Furthermore, combining the calibrated characteristic parameters (e.g., the optimized threshold for melting temperature 1600℃±10℃), the characteristic mapping relationship of the interactive parameters and the characteristic parameter constraints are combined into a system of equations: The characteristic parameter (melting temperature) is used as the boundary constraint input, the interaction parameters (thermal gradient, tension fluctuation) are used as independent variables, and the fiber transition state performance index (defect density, uniformity coefficient) is used as the dependent variable. The nonlinear relationship between parameters (e.g., the secondary effect of the interaction term of thermal gradient and melting temperature on defect density) is learned through the training set data, and the model prediction accuracy is verified by the validation set (e.g., the mean square error is lower than the set threshold).

[0036] Furthermore, by solving this system of equations, the optimal dynamic control range of each interactive parameter in the core process is determined: Target thresholds for transition state performance indicators are set (e.g., interface defect density ≤ XXX pieces / mm², uniformity coefficient ≥ XXX). The range of values ​​for interactive parameters that satisfy these thresholds is then derived (e.g., by using a genetic algorithm to traverse combinations of thermal gradient ranges of XXX-XXX℃ / μm and tension fluctuation ranges of XXX-XXXN, the parameter ranges that allow the performance indicators to meet the standards are selected). This result provides a benchmark input for the collaborative calibration of subsequent process steps.

[0037] Collect key process data for core process steps, such as using an infrared thermal imager to monitor the thermal gradient change at the melt solidification interface (unit: ℃ / μm), strain gauges to record the dynamic fluctuations of tensile tension (unit: N), and laser displacement sensors to track melt flow rate (unit: m / s); and simultaneously acquire fiber transition state performance indicators (e.g., solidification interface defect density observed by scanning electron microscopy, unit: defects / mm2, or initial tensile uniformity coefficient measured by laser interferometer, dimensionless).

[0038] For example, under a certain adjustment condition (drawing rate set to 4.5 m / s, cooling airflow intensity of XXX L / min), five sets of process response data at different time points were collected: The first set of data, with a thermal gradient ΔT / μm = 120℃ / μm, tension fluctuation ΔF = 15N, and melt flow velocity V = 2.1m / s, corresponds to a solidification interface defect density of 8 defects / mm². The second set of data, with a thermal gradient ΔT / μm = 150℃ / μm, tension fluctuation ΔF = 22N, and melt flow velocity V = 2.3m / s, corresponds to a defect density of 15 defects / mm². This pattern continues, recording the performance index values ​​for each parameter combination. Comparative analysis revealed that when the thermal gradient exceeds 140℃ / μm, the defect density increases significantly (e.g., from 8 defects / mm² to 18 defects / mm²), initially indicating a positive correlation between thermal gradient and defect density. Simultaneously, when the tension fluctuation exceeds 20N, the initial tensile uniformity coefficient decreases (e.g., from 0.92 to 0.85), indicating that tension fluctuation has a significant impact on macroscopic uniformity.

[0039] Based on the correlation trends extracted in the first step, the characteristic mapping relationships of the interactive parameters (thermal gradient ΔT / μm, tension fluctuation ΔF, melt flow velocity V) (i.e., the correlation between each parameter and defect density and uniformity coefficient) are further combined with the calibrated characteristic parameters (e.g., melting temperature T = 1600℃ ± 10℃) into a system of equations: A multi-parameter correlation model is constructed by using characteristic parameters (melting temperature) as boundary constraint inputs (e.g., setting T to be within the range of 1590℃-1610℃), interaction parameters (thermal gradient, tension fluctuation, melt flow rate) as independent variables, and fiber transition state performance indicators (defect density, uniformity coefficient) as dependent variables. For example, the Gradient Boosting Decision Tree (GBDT) algorithm from machine learning is used as the modeling method, as detailed below: Historical experimental data (including parameter combinations and performance indicators under multiple adjusted operating conditions) were divided into a training set (80% of the data) and a validation set (20% of the data). In the training set, the input features were interaction parameters (ΔT / μm, ΔF, V) and characteristic parameters (T=1600℃), with output labels being defect density (numbers / mm²) and uniformity coefficient (0-1). The nonlinear relationships between parameters were learned through the training set data (e.g., the model captured that when T=1600℃ and ΔT / μm>140℃ / μm, the defect density increases at an accelerating rate with increasing ΔF). The model's prediction accuracy was verified using the validation set (e.g., mean squared error MSE<0.05, coefficient of determination R²>0.9). This model is equivalent to a set of equations describing the quantitative relationship between interaction parameters and performance indicators, and implicitly includes constraints on the characteristic parameter (melting temperature) (T must be within the range of 1590℃-1610℃).

[0040] Based on the equation set (i.e., the multi-parameter correlation model) established in the second step, target thresholds for fiber transition state performance indicators are set (e.g., defect density ≤ 10 defects / mm², uniformity coefficient ≥ 0.90). The range of values ​​for interaction parameters that satisfy these thresholds is determined through reverse derivation. The algorithm searches for suitable combinations of interactive parameters within the parameter space. First, it defines the search range (e.g., thermal gradient ΔT / μm ∈ [100℃ / μm, 200℃ / μm], tension fluctuation ΔF ∈ [10N, 30N], melt flow velocity V ∈ [2.0m / s, 2.5m / s]), and uses the objective function (e.g., minimizing defect density and maximizing uniformity coefficient) as the optimization goal. During algorithm iteration, multiple candidate parameter combinations are generated (e.g., candidate group 1: ΔT / μm = 130℃ / μm, ΔF = 18N, V = 2.2m / s; candidate group 2: ΔT / μm = 135℃ / μm, ΔF = 16N, V = 2.1m / s). The system inputs the corresponding performance indicators predicted by the correlation model (e.g., the first group predicts a defect density of 9 defects / mm² and a uniformity coefficient of 0.91; the second group predicts a defect density of 8 defects / mm² and a uniformity coefficient of 0.92); it then selects parameter combinations that meet the target threshold (defect density ≤ 10 defects / mm² and uniformity coefficient ≥ 0.90) (e.g., both the first and second groups meet the requirements), and further converges to the optimal range (e.g., the optimal control range of thermal gradient ΔT / μm is 130℃ / μm-140℃ / μm, the optimal range of tension fluctuation ΔF is 15N-18N, and the optimal range of melt flow velocity V is 2.1m / s-2.2m / s).

[0041] Suppose that under a certain adjustment condition, the calibrated melting temperature characteristic parameter is T = 1600℃ ± 10℃ (boundary constraint), and the initial experimental data of the interaction parameters are as follows: When the thermal gradient ΔT / μm = 120℃ / μm, the tension fluctuation ΔF = 15N, and the melt flow rate V = 2.1m / s, the defect density = 8 defects / mm², and the uniformity coefficient = 0.92. When the thermal gradient ΔT / μm = 150℃ / μm, the tension fluctuation ΔF = 22N, and the melt flow rate V = 2.3m / s, the defect density = 15 defects / mm², and the uniformity coefficient = 0.85. When the thermal gradient ΔT / μm = 130℃ / μm, the tension fluctuation ΔF = 18N, and the melt flow rate V = 2.2m / s, the defect density = 9 defects / mm², and the uniformity coefficient = 0.91. When the thermal gradient ΔT / μm = 160℃ / μm, the tension fluctuation ΔF = 25N, and the melt flow rate V = 2.4m / s, the defect density = 22 defects / mm², and the uniformity coefficient = 0.80.

[0042] Correlation model analysis revealed that when the thermal gradient exceeded 140℃ / μm and the tension fluctuation exceeded 20N, the defect density increased significantly (exceeding 10 defects / mm²), and the uniformity coefficient fell below 0.90. Setting the target threshold as defect density ≤ 10 defects / mm² and uniformity coefficient ≥ 0.90, reverse derivation yielded: The optimal control range for the thermal gradient is 130℃ / μm-140℃ / μm (e.g., a defect density of 9 defects / mm² at 135℃ / μm), the optimal range for tension fluctuation is 15N-18N (e.g., a uniformity coefficient of 0.92 at 16N), and the recommended range for melt flow rate is 2.1m / s-2.2m / s (e.g., both indicators meet the requirements at 2.15m / s). The optimal dynamic control range for the interactive parameters in the core process steps is ultimately determined as follows: thermal gradient ΔT / μm = 130℃ / μm-140℃ / μm, tension fluctuation ΔF = 15N-18N, and melt flow rate V = 2.1m / s-2.2m / s. This ensures that the fiber transition state performance meets the initial quality requirements for aerospace-grade superconducting fibers and provides stable input for the collaborative calibration of subsequent process steps.

[0043] Phase (2): Recursive Collaborative Calibration of Process Links and Global Parameter Transfer After optimizing the interaction parameters of the core process steps, the edge process parameters of that step (e.g., the exit tension value in the initial drawing stage) are used as the associated reference parameters for the next step to be calibrated. Following the logical sequence of the manufacturing process (e.g., from initial drawing → fiber forming → online annealing → final winding), the key parameters of adjacent downstream steps (e.g., online annealing temperature and annealing time after fiber forming) are used as new parameters to be calibrated, thus initiating a recursive collaborative calibration process. For each downstream stage, the data acquisition and analysis process of stage (1) is repeated. The transition state data of this stage (such as temperature uniformity and fiber stress release rate in the online annealing stage) is collected by high-precision sensors, and the parameters to be calibrated are actively adjusted (such as changing the annealing temperature or time). Multiple sets of process response data are obtained under each adjustment condition (such as repeating the test 3 times at the same annealing temperature and recording the fiber stress residual value each time). The characteristic mapping relationship of the interactive influence parameters (such as the matching relationship between annealing temperature and time) is calculated (such as the fiber stress residual amount is significantly reduced when the annealing temperature exceeds XXX℃), and combined with the upstream related reference parameters (such as the initial drawing exit tension value) and the calibrated characteristic parameters (such as melting temperature and melt flow rate), a multi-parameter equation system is established.

[0044] Furthermore, the coordinated control values ​​of the interactive parameters in each process step are determined by solving the system of equations. The processing logic is as follows: The fiber performance transfer target for this stage is set (e.g., fiber diameter tolerance ≤ XXX μm after forming, residual stress ≤ XXX MPa after annealing), and the range of values ​​for the parameters to be calibrated to meet the target is derived in reverse (e.g., by using a particle swarm optimization algorithm to traverse the combination of annealing temperature and time, and selecting the parameter range that simultaneously meets the target for residual stress and diameter tolerance). This process sequentially completes the collaborative calibration of interactive parameters for all key process stages (e.g., drawing, forming, annealing, winding, etc.) except for the initial raw material pretreatment stage, ensuring dynamic matching between parameters in each stage.

[0045] For each process step to be calibrated (e.g., fiber forming), based on expert experience or historical data analysis (usually manually input via an interactive interface), the key interactive parameters of that step (e.g., mold temperature uniformity during the forming stage, the matching relationship between fiber traction speed and cooling rate) and the related benchmark parameters transmitted from upstream (e.g., exit tension value from core process steps, melt flow rate stability index) are determined.

[0046] By actively adjusting the process parameters of this stage (e.g., changing the heating power of the molding die and adjusting the cooling water flow rate), multiple sets of process response data under different conditions are collected simultaneously under each adjustment condition. For example, a high-precision temperature sensor is used to monitor the surface temperature distribution of the molding die (unit: °C), a linear speed encoder records the fiber traction speed (unit: m / s), and an online imaging system collects the fiber cross-sectional morphology (e.g., real-time images obtained through an industrial camera). This data is then combined with preliminary characterization data of downstream performance (e.g., fiber diameter tolerance and surface roughness).

[0047] Multiple sets of process response data were obtained under each set of adjusted operating conditions (e.g., three repeated tests were conducted at the same mold temperature, recording the fiber diameter deviation and surface roughness value each time). Real-time monitoring values ​​of interactive parameters (e.g., mold temperature uniformity ΔTm / ℃, traction speed Vpm / s, cooling rate Rcm / s) and fiber transition state or initial performance indicators (e.g., diameter tolerance ΔDmm, surface roughness Raμm) were extracted from these data. The correlation trends between these data were analyzed (e.g., when mold temperature uniformity exceeds ±15℃, fiber diameter tolerance increases from 0.02mm to 0.05mm; when traction speed and cooling rate are mismatched, surface roughness increases significantly).

[0048] The characteristic mapping relationship of the interaction parameters in this stage is calculated as follows: Using interactive parameters (mold temperature uniformity, traction speed, cooling rate) as independent variables and fiber performance indicators (diameter tolerance, surface roughness) as dependent variables, and combining upstream-transferred reference parameters (e.g., the core component outlet tension value FtN) with calibrated characteristic parameters (e.g., melt temperature T = 1600℃ ± 10℃), a multi-parameter correlation model is constructed: Machine learning algorithms (such as random forest regression or support vector machine) are employed, using upstream correlation benchmark parameters and characteristic parameters as boundary constraint inputs (e.g., setting the outlet tension Ft to be within the range of XXX-XXXN, and the melt temperature T fixed at 1600℃), interaction parameters as independent variables, and fiber performance indicators as dependent variables. The nonlinear correlation between parameters is learned using training set data (80% of the adjustment condition data) (e.g., the model captures an exponential correlation between diameter tolerance ΔD and cooling rate Rc when the die temperature uniformity ΔTm > ±15℃ and the traction speed Vp > XXXm / s). The model's prediction accuracy is verified using validation set data (20% of the adjustment condition data) (e.g., mean square error MSE < 0.03, coefficient of determination R² > 0.85). This model is equivalent to a multi-parameter equation system describing the quantitative relationship between interaction parameters and performance indicators in this stage, and implicitly includes constraints on upstream parameters.

[0049] Based on the multi-parameter equations (i.e., the correlation model) established in the first step, the transfer target of fiber properties in this process step is set (e.g., fiber diameter tolerance ≤ 0.03 mm, surface roughness ≤ XXX μm, to ensure process compatibility with the downstream online annealing step). The range of values ​​for the parameters to be calibrated (i.e., the interaction parameters of this step) that meet the target is determined through reverse derivation. The algorithm searches for suitable combinations of interactive parameters within the parameter space. First, it defines the search range (e.g., mold temperature uniformity ΔTm ∈ [±10℃, ±20℃], traction speed Vp ∈ [XXXm / s, XXXm / s], cooling rate Rc ∈ [XXXm / s, XXXm / s]), and uses the objective function (e.g., minimizing diameter tolerance and surface roughness) as the optimization objective. During algorithm iteration, multiple sets of candidate parameter combinations are generated (e.g., candidate set 1: ΔTm = ±12℃, Vp = XXXm / s, Rc = XXXm / s; candidate set 2: ΔTm = ±14℃, Vp = XXXm / s, Rc = Input the corresponding performance indicators predicted by the correlation model (e.g., the first group predicts a diameter tolerance of 0.025 mm and a surface roughness of XXX μm; the second group predicts a diameter tolerance of 0.035 mm and a surface roughness of XXX μm); filter out the parameter combinations that meet the transfer target (diameter tolerance ≤ 0.03 mm and surface roughness ≤ XXX μm) (e.g., the first group meets the requirements), and further converge to the optimal range (e.g., the optimal control range for mold temperature uniformity is ±12℃-±15℃, the optimal range for traction speed is XXX-XXX m / s, and the optimal range for cooling rate is XXX-XXX m / s).

[0050] Let the upstream reference parameter of a certain fiber forming process be the outlet tension Ft = XXXN (a value transmitted from the core process), and the calibrated characteristic parameter be the melt temperature T = 1600℃ ± 10℃. The interactive parameters of this process include mold temperature uniformity ΔTm (unit: ℃), traction speed Vp (unit: m / s), and cooling rate Rc (unit: m / s). The following typical data were collected through multi-condition adjustment experiments: When ΔTm=±10℃, Vp=XXXm / s, and Rc=XXXm / s, the diameter tolerance ΔD=0.02mm and the surface roughness Ra=XXXμm; When ΔTm=±18℃, Vp=XXXm / s, and Rc=XXXm / s, the diameter tolerance ΔD=0.04mm and the surface roughness Ra=XXXμm; When ΔTm=±14℃, Vp=XXXm / s, and Rc=XXXm / s, the diameter tolerance ΔD=0.03mm and the surface roughness Ra=XXXμm; When ΔTm=±12℃, Vp=XXXm / s, and Rc=XXXm / s, the diameter tolerance ΔD=0.025mm and the surface roughness Ra=XXXμm.

[0051] The transfer target for this step is set as diameter tolerance ≤ 0.03 mm and surface roughness ≤ XXX μm (to ensure the process stability of the subsequent online annealing step). This is derived through reverse derivation: The optimal control range for mold temperature uniformity is ±12℃-±15℃ (e.g., diameter tolerance is 0.022mm at ±13℃), the optimal range for traction speed is XXX-XXXm / s (e.g., surface roughness is XXXμm at XXXm / s), and the optimal range for cooling rate is XXX-XXXm / s (e.g., both indicators meet the standard at XXXm / s).

[0052] The final coordinated control values ​​for the interactive parameters in the fiber forming process were determined as follows: mold temperature uniformity ΔTm = ±12℃-±15℃, traction speed Vp = XXX-XXXm / s, and cooling rate Rc = XXX-XXXm / s, to ensure that the fiber performance meets the transfer requirements and achieves seamless connection with downstream processes.

[0053] Phase (3): Global consistency adjustment of parameters affecting the entire process interaction After completing the collaborative calibration of the interaction parameters of all process steps, the interaction parameters calibrated in each step of the process are adjusted globally to ensure consistency, based on the final goal of the preparation process (e.g., the overall performance pass rate of aerospace-grade atomic oxygen resistant superconducting fiber ≥ XXX%). The processing logic is as follows: summarize the actual values ​​of the interaction parameters of each step (e.g., the thermal gradient of the core step XXX℃ / μm, the annealing temperature XXX℃), and analyze their correlation with the final performance indicators (e.g., the overall performance pass rate, the atomic oxygen erosion resistance rate) (e.g., extract the combination of dominant parameters affecting the pass rate through principal component analysis).

[0054] Furthermore, if a systematic deviation is detected due to the optimization of local parameters (e.g., excessive optimization of the thermal gradient in the core stage leads to an increase in residual stress in the annealing stage), the parameter values ​​are fine-tuned through a dynamic weight allocation algorithm (e.g., adjusting the priority based on the contribution of each stage to the final performance). This can be done by appropriately lowering the upper limit of the thermal gradient in the core stage and simultaneously increasing the temperature compensation in the annealing stage. Ultimately, this eliminates parameter conflicts between stages and ensures coordinated matching of parameters throughout the entire process from raw material melting to finished product winding. This achieves a comprehensive improvement in fiber macroscopic consistency (e.g., diameter tolerance ≤ XXX μm, surface defect rate ≤ XXX%) and overall resistance to atomic oxygen (e.g., erosion rate ≤ XXX nm / h).

[0055] The actual values ​​of the interaction parameters of each key process step in the entire process (such as the core initial drawing step, the intermediate fiber forming step, the subsequent online annealing step, and the final winding step) after calibration in steps B1 and B2 are integrated. For example, the actual thermal gradient of the initial drawing step is set to 135℃ / μm and the drawing tension fluctuation is 16N; the die temperature uniformity of the fiber forming step is ±13℃, the traction speed is 2.2m / s, and the cooling rate is 1.5m / s; the temperature of the online annealing step is XXX℃ and the annealing time is XXXmin; the tension control of the winding step is XXXN, etc. At the same time, the final performance characterization data of the corresponding batch of fibers are extracted (erosion rate measured by atomic oxygen exposure test chamber, attenuation rate obtained by superconducting critical temperature tester, diameter tolerance measured by laser diameter gauge, surface defect density observed by scanning electron microscope, etc.) to form a parameter-performance correlation database.

[0056] Based on this database, the correlation between the interaction parameters of each stage and the final performance indicators is explored. The processing logic is as follows: After standardizing the interaction parameters of each stage (such as thermal gradient, traction speed, annealing temperature, etc.) and the final performance indicators (such as erosion rate, critical temperature decay rate, diameter tolerance, etc.), the data are input into the analysis model. Principal components are extracted or feature importance scores are calculated to identify the parameter combinations that have the greatest impact on the final performance. For example, the analysis found that the thermal gradient in the initial drawing stage (contribution 35%) and the temperature in the online annealing stage (contribution 28%) are the main parameters affecting the fiber's resistance to atomic oxygen erosion, while the traction speed in the fiber forming stage (contribution 20%) and the tension control in the winding stage (contribution 17%) jointly determine the diameter tolerance. If a parameter in a certain stage (such as the initial drawing thermal gradient) is over-optimized (set too low, resulting in insufficient melt fluidity), it may cause fiber cross-section distortion in subsequent forming stages, ultimately leading to the diameter tolerance exceeding the standard (e.g., increasing from ±0.02mm to ±0.05mm), i.e., a systematic deviation occurs.

[0057] When the analysis results show that the optimization of parameters in a certain stage has a negative chain reaction on other stages (for example, local thermal gradient optimization leads to an increase in residual stress in the annealing stage, or excessive traction speed in the forming stage leads to an imbalance in winding tension), it is necessary to adjust the parameter values ​​globally using a dynamic weight allocation algorithm. Initial weights are assigned to the interaction parameters of each stage (e.g., the initial drawing thermal gradient weight is set to 0.35, the online annealing temperature weight is set to 0.28, and the fiber forming traction speed weight is set to 0.20, etc., based on their contribution to the final performance), and the total weights are 1. Based on the compliance of the final performance indicators (e.g., whether the erosion rate exceeds 5 nm / h, whether the diameter tolerance exceeds ±0.02 mm), the deviation contribution value of each stage parameter is calculated (e.g., if the initial drawing thermal gradient is too high and the erosion rate exceeds the standard by 10%, then its deviation contribution value is +0.10).

[0058] The core of the dynamic weight allocation algorithm is to dynamically adjust the parameter priorities according to the deviation contribution values. If the parameter optimization of a certain link (such as initial drawing) causes the performance of other links to deteriorate (for example, stress residues are caused by insufficient melt fluidity in the annealing link), then reduce the weight of the parameters related to this link (for example, lower the weight of the initial drawing heat gradient from 0.35 to 0.30), and at the same time increase the weight of the parameters of the affected link (such as the annealing link) (for example, raise the weight of the annealing temperature from 0.28 to 0.33), and recalculate the parameter adjustment direction. The specific adjustment strategy is generated through preset parameter-response mapping rules: for example, if it is detected that the combination of the initial drawing heat gradient > 140 °C / μm and the annealing temperature < XXX °C results in an excessive erosion rate, then reduce the upper limit of the initial drawing heat gradient (such as callback from 135 °C / μm to 130 °C / μm), and synchronously increase the temperature compensation of the annealing link (such as increase from XXX °C to XXX °C); if the fiber forming drawing speed is too fast (such as > 2.3 m / s) resulting in winding tension fluctuations (such as > ±5 N), then reduce the drawing speed to the range of 2.1 - 2.2 m / s, and at the same time fine-tune the winding tension control compensation (such as increase by XXX N).

[0059] Furthermore, collect the historical data of characteristic parameters and interaction influence parameters in the whole process, and integrate the fiber final performance characterization results as the analysis benchmark: Extract historical data from the sensor network of the preparation system (such as molten temperature sensors, drawing rate encoders, cooling air flow pressure gauges) and performance detection equipment (such as atomic oxygen exposure test chambers, superconducting critical temperature testers, surface topography scanning electron microscopes), including characteristic parameters (such as molten temperature, melt flow rate), interaction influence parameters (such as drawing rate, cooling rate, annealing temperature), and fiber final performance indicators (such as atomic oxygen erosion rate, superconducting critical temperature decay rate, surface defect density). The data collection covers multiple batches of preparation tests to ensure that it includes parameter combinations under different process conditions and corresponding performance responses (such as low erosion rate samples and high erosion rate samples under the combination of high-temperature melting + high-speed drawing).

[0060] Furthermore, use multivariate correlation analysis techniques to挖掘 the hidden association rules between characteristic parameters and interaction influence parameters: High-dimensional parameter data are preprocessed using dimensionality reduction and feature extraction algorithms (such as principal component analysis, PCA). First, characteristic parameters, interaction parameters, and performance indicators are standardized (e.g., normalized to the 0-1 range). Principal components are extracted using PCA (e.g., the first principal component represents the combined effect of temperature and rate, and the second principal component represents the synergistic effect of cooling and tension). Principal components with a cumulative contribution rate exceeding a set threshold (e.g., 95%) are retained to reduce data dimensionality. Based on the extracted principal components or original parameters (selected according to data characteristics), parameter combinations strongly correlated with performance indicators are identified (e.g., when the melting temperature is >1600℃ and the drawing rate is >5m / s, the atomic oxygen erosion rate is significantly positively correlated with the interaction term between the two).

[0061] Furthermore, based on the above analysis results, the specific correlation patterns between the two types of parameters are summarized as follows: High-risk parameter ranges are defined (for example, when the melt temperature ∈ [1600℃, 1700℃] and the drawing rate ∈ [5m / s, 6m / s], the probability that the fiber atomic oxygen erosion rate exceeds the threshold XXXnm / h is >90%); at the same time, low-risk parameter combinations are recorded (for example, when the melt temperature ≤1550℃ or the drawing rate ≤4m / s, the erosion rate is stable within the safe range).

[0062] For example, historical data analysis revealed that when the melting temperature exceeds 1600℃ and the drawing rate is greater than 5m / s, the density of lattice defects inside the fiber increases sharply, leading to an exponential increase in the atomic oxygen erosion rate (e.g., from XXXnm / h to XXXnm / h). This pattern serves as the basis for subsequent regulation.

[0063] Furthermore, based on the results of cross-dimensional correlation analysis, a dynamic adaptive regulation model is constructed: Using real-time monitoring data (current characteristic parameters and interactive influence parameters) as the core input, a rule engine is used to establish a mapping relationship between parameter combinations and risk levels. For example, parameter combinations (melting temperature, drawing rate) are mapped to risk labels (low risk / medium risk / high risk), where high risk is defined as performance indicators exceeding safety thresholds (such as erosion rate > XXX nm / h or critical temperature decay rate > XXX% / month).

[0064] Furthermore, during the fabrication process, key characteristic parameters (such as the current melt temperature sensor reading) and interaction parameters (such as the real-time drawing rate encoder feedback value) are monitored in real time, and the monitoring data is input into a dynamic adaptive control model for online inference. The processing logic is as follows: The model collects real-time values ​​of characteristic parameters and interaction parameters at set time windows (e.g., every 10 seconds). The model predicts the risk level based on the current parameter combination (e.g., calculates the predicted erosion rate corresponding to the current melting temperature of 1620℃ and drawing rate of 5.2m / s, and determines whether it exceeds the threshold).

[0065] Furthermore, when a set of parameters is detected to fall into a historical high-risk range (e.g., the model outputs a high-risk level), an adjustment command is automatically triggered to correct the parameter deviation: The control commands are generated based on preset parameter-response mapping rules (e.g., for a high-risk combination with a melt temperature >1600℃ and a drawing rate >5m / s, the corresponding control action is to reduce the drawing rate to 4.5m / s ±0.2m / s or increase the cooling airflow intensity to XXXL / min). The process parameters are adjusted in real time via actuators (such as a drawing rate servo motor or cooling system valves). For example, if the melt temperature is detected to be 1620℃ and the drawing rate to be 5.3m / s (falling into the high-risk range), the system automatically reduces the drawing rate to 4.8m / s and increases the cooling airflow intensity, thereby suppressing lattice defect formation and reducing the atomic oxygen erosion rate.

[0066] This dynamic control process is continuously iterated. After each parameter adjustment, the performance response is monitored (such as whether the subsequent fiber erosion rate falls back to a safe range), and the control strategy is optimized (such as adjusting the correction range of the drawing rate according to the actual effect).

[0067] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0068] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, characterized in that: The management method includes the following steps: Step A. Calculate the characteristic mapping relationship corresponding to each type of characteristic parameter based on multiple sets of experimental data, and combine the characteristic mapping relationships of the same characteristic parameter under different experimental conditions into a set of mathematical equations. Solve the set of mathematical equations to determine the optimal control range of the characteristic parameter. Step B. Perform cyclical collaborative calibration of the interaction parameters between various process steps in the preparation process: Step B1. Obtain multiple sets of process response data under different conditions in each adjusted operating condition, and determine the optimal dynamic control range of each interactive parameter in the core process link; Step B2. Using the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, determine the collaborative control values ​​of each interactive parameter in each process step in turn, and complete the collaborative calibration of the interactive parameters of all process steps except the raw material pretreatment step. Step B3. Based on the final goal of the preparation process, adjust the interaction parameters of each step of the calibration to ensure global consistency. Step C. Collect historical data of characteristic parameters and interaction parameters throughout the entire process, and combine them with the final fiber performance characterization results to construct a dynamic adaptive control model. When a set of parameters is detected to fall into the historical risk range, the control command is automatically triggered.

2. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 1, characterized in that: Step B1. Obtain multiple sets of process response data under different conditions in each adjusted operating condition, and determine the optimal dynamic control range of each interactive parameter in the core process link, including the following steps: Select the core process steps in the preparation process, and simultaneously collect multiple sets of transition state data under different process conditions in the core process steps; The stability of the melt solidification interface and the dynamic fluctuation of the drawing tension are monitored in real time by sensors, and the process parameters are actively adjusted. Under each adjusted operating condition, multiple sets of process response data under different conditions are obtained, and the characteristic mapping relationship corresponding to each set of process response data is calculated. By combining the calibrated characteristic parameters, a system of equations is established and solved to determine the optimal dynamic control range of each interactive parameter in the core process.

3. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 2, characterized in that: Step B2. Using the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, determine the coordinated control values ​​of each interactive parameter in each process step in sequence, and complete the coordinated calibration of interactive parameters of all process steps except the raw material pretreatment step, including the following steps: Following the logical sequence of the preparation process, the parameters of adjacent downstream steps are used as new parameters to be calibrated. By combining the same multi-condition data acquisition and characteristic mapping solutions, the collaborative control values ​​of each interactive parameter in each process step are determined sequentially.

4. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 1, characterized in that: Step C. Collect historical data of characteristic parameters and interaction parameters throughout the entire process, and combine them with the final fiber performance characterization results to construct a dynamic adaptive control model. When a set of parameters is detected to fall into the historical risk range, a control command is automatically triggered, including the following steps: Multivariate correlation analysis techniques are used to uncover the correlation patterns between characteristic parameters and interaction parameters. Based on the results of this cross-dimensional correlation analysis, a dynamic adaptive control model is constructed to monitor characteristic parameters and interaction parameters in real time. When a combination of parameters is detected to fall into the historical risk range, control commands are automatically triggered, including reducing the pulling rate or increasing the cooling airflow intensity.

5. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 1, characterized in that: Step A. Calculate the characteristic mapping relationship corresponding to each type of characteristic parameter based on multiple sets of experimental data, including the following steps: For each type of characteristic parameter, multiple preparation experiments were conducted under different process conditions, and the preparation process parameters and initial fiber performance characterization data were recorded simultaneously. Based on the quantitative functional relationship between the changes of multiple characteristic parameters and fiber performance indicators, the characteristic mapping relationship of the same characteristic parameter under different experimental conditions is combined into a system of mathematical equations; The optimal control range of characteristic parameters is determined by solving a system of mathematical equations.

6. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 5, characterized in that: The characteristic parameters include the proportion of raw material components, melting environment temperature, melt flow rate, purity of inert gas protection, and initial tensile tension.

7. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 2, characterized in that: For each adjusted operating condition, multiple sets of process response data under different conditions are obtained. The characteristic mapping relationship corresponding to each set of process response data is calculated. Combined with the calibrated characteristic parameters, the following steps are included: Extract real-time monitoring values ​​of interactive influence parameters and fiber transition state performance indicators, and analyze the correlation trend between the two; Combining the calibrated characteristic parameters, the characteristic mapping relationship of the interaction parameters and the characteristic parameter constraints are combined into a system of equations: The characteristic parameters are used as boundary constraint inputs, the interaction influence parameters are used as independent variables, and the fiber transition state performance index is used as the dependent variable. The nonlinear relationship between the parameters is learned through the training set data. Set a target threshold for the transition state performance index, and then deduce the range of values ​​for the interaction parameters that satisfy this threshold.

8. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 3, characterized in that: By using the same multi-condition data acquisition and characteristic mapping to solve the system simultaneously, the coordinated control values ​​of each interactive parameter in each process step are determined sequentially, including the following steps: Multiple sets of process response data were obtained under each adjusted operating condition, and the characteristic mapping relationship of the interactive influence parameters was calculated. By combining the upstream reference parameters with the calibrated characteristic parameters, a system of multi-parameter equations is established. Set the fiber performance transfer target for this stage, and deduce the range of values ​​for the parameters to be calibrated to meet the target.

9. The method for preparing and managing aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers according to claim 3, characterized in that: In step B3, based on the final goal of the preparation process, the interaction parameters of each step are adjusted globally for consistency, including the following steps: Summarize the actual values ​​of the interaction parameters at each stage and analyze their correlation with the final performance indicators; If a systematic deviation is detected due to the optimization of parameters in a local process, the parameter values ​​are adjusted through a dynamic weight allocation algorithm to eliminate parameter conflicts between processes.

10. A management system for the preparation of aerospace-grade atomic oxygen-resistant superconducting hybrid nickel fibers, used to implement the management method described in any one of claims 1-9, characterized in that: include: Characteristic parameter optimization module: Based on multiple sets of experimental data, calculate the characteristic mapping relationship corresponding to each type of characteristic parameter, put the characteristic mapping relationship of the same characteristic parameter under different experimental conditions into a set of mathematical equations, and solve the set of mathematical equations to determine the optimal control range of the characteristic parameter. Global Adjustment Module: Performs cyclical collaborative calibration of the interaction parameters between various process steps in the preparation process: Under each adjustment condition, obtains multiple sets of process response data under different conditions, determines the optimal dynamic control range of each interaction parameter in the core process step, uses the edge process parameters of the core process step as the associated reference parameters of the next process step to be calibrated, sequentially determines the collaborative control value of each interaction parameter in each process step, completes the collaborative calibration of the interaction parameters of all process steps except the raw material pretreatment step, and adjusts the global consistency of the calibrated interaction parameters of each step based on the final goal of the preparation process. Control module: Collects historical data of characteristic parameters and interactive parameters throughout the entire process, and combines them with the final performance characterization results of the fiber to construct a dynamic adaptive control model. When a set of parameters is detected to fall into the historical risk range, the control command is automatically triggered.

Images