Flexible mineral insulation anti-interference fireproof cable and preparation method thereof

Abstract

The invention relates to the technical field of cable preparation, and discloses a flexible mineral insulation anti-interference fireproof cable and a preparation method thereof.The method comprises the steps that the number of stranded wire cores is compared with a historical stranded wire database, a stranded wire strategy or a prediction strategy is determined based on the comparison result, and the stranded wire speed is determined according to the stranded wire strategy or the prediction strategy; the qualification of the stranding machine is judged according to the relation between the historical stranding yield and the real-time stranding yield, a cable preparation influence sequence is constructed, preparation influence samples are constructed based on the cable preparation influence sequence, and an execution model association strategy or a model mapping strategy is determined according to the number of the samples. And determining a preparation environment association result based on the extracted preparation environment data and an association rule algorithm, determining a target wire stranding speed based on the preparation environment association result or a model mapping strategy, and completing wire stranding at the target wire stranding speed. According to the invention, the adaptation stability of the stranding speed during the preparation of the insulated cable is ensured.

Description

Technical Field

[0001] This invention relates to the field of cable manufacturing technology, and more specifically, to a flexible mineral-insulated anti-interference fireproof cable and its manufacturing method. Background Technology

[0002] With the rapid development of fields such as construction engineering, rail transit and new energy storage, cables, as the core carriers of power transmission and signal transmission, are facing increasingly stringent requirements in terms of safety performance, installation adaptability and anti-interference capabilities.

[0003] Chinese Patent Publication No. CN119495477B discloses an insulated cable and its preparation method and system. This invention relates to the field of insulated cable technology. The method includes: determining multiple initial stranding speeds using an initial stranding speed determination model based on video footage of the cable being processed at the current stranding speed; sending the multiple initial stranding speeds to a cable stranding device and acquiring images of the finished cable processed at each initial stranding speed; constructing a graph structure based on the images of the finished cable processed at each initial stranding speed; processing the graph structure using a graph convolutional network to determine a target stranding speed; sending the target stranding speed to the cable stranding device and controlling the cable stranding device to strand the cable. Therefore, during the cable stranding process, different numbers of stranded cores in the conductors to be stranded will result in different stranding speeds. Furthermore, factors such as temperature, humidity, and air dust concentration in the stranding environment can alter processing characteristics, leading to a failure in the adaptability of the stranding speed.

[0004] Therefore, it is necessary to design a flexible mineral-insulated anti-interference fireproof cable and its preparation method to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a flexible mineral-insulated anti-interference fireproof cable and its preparation method, aiming to solve the problem that different stranding core numbers of conductors to be stranded will result in different stranding speeds, and that factors such as temperature, humidity, and air dust concentration in the stranding operation environment will change the processing characteristics, thus leading to the failure of the adaptability of stranding speed.

[0006] In one aspect, the present invention provides a method for preparing a flexible mineral-insulated anti-interference fire-resistant cable, comprising: Obtain the number of stranded cores of the conductor to be stranded, compare the number of stranded cores with the historical stranded wire database, determine the stranding strategy or prediction strategy based on the comparison results, and determine the stranding speed according to the stranding strategy or prediction strategy. Obtain the historical stranded wire yield and the real-time stranded wire yield at the stated stranding speed. Determine the qualification of the stranding machine based on the relationship between the historical stranded wire yield and the real-time stranded wire yield. When the stranding machine is determined to be unqualified, divide the cable preparation area into several preparation data points and construct a cable preparation influence series. Based on the cable preparation influence series, preparation influence samples are constructed. The preparation environment data of the cable preparation influence series is extracted according to the sample size of the preparation influence samples. The execution model association strategy or model mapping strategy is determined according to the sample size. When it is determined to execute the model association strategy, the preparation environment association result is determined based on the extracted preparation environment data and association rule algorithm, the target stranding speed is determined based on the preparation environment association result or the model mapping strategy, and the stranding is completed at the target stranding speed.

[0007] Furthermore, when comparing the number of stranded cores with a historical stranded wire database, determining a stranding strategy or prediction strategy based on the comparison results, and determining the stranding speed according to the stranding strategy or prediction strategy, the process includes: The historical stranded wire database includes several historical stranded wire core counts and several historical stranded wire speeds, and each historical stranded wire core count corresponds to a historical stranded wire core count. When the historical stranded wire database contains a historical stranded wire number that is the same as the number of stranded wire cores, the stranded wire strategy is determined to be executed. The stranded wire strategy determines the historical stranded wire speed corresponding to the historical stranded wire number that is the same as the number of stranded wire cores as the stranded wire speed. If the historical stranded wire database does not contain a historical stranded wire core number that is the same as the stated stranded wire core number, then the prediction strategy is executed. The prediction strategy determines the stranding speed based on the random forest model and the stated stranded wire core number.

[0008] Furthermore, when the prediction strategy determines the stranding speed based on the random forest model and the number of stranded cores, it includes: Obtain the conductor dataset and merge it with the historical stranded wire database to determine the stranded wire dataset; An initial random forest model is pre-selected, with the number of trees in the initial random forest model set to 200, and the maximum depth of the trees adjusted based on the distribution of the number of strands in the input. The minimum number of samples for node splitting is set to 10, the minimum number of samples for leaf nodes is set to 5, and the maximum number of leaf nodes is 100. The twisted wire dataset is divided into a training set and a test set. The initial random forest model is iteratively trained based on the training set, and the F1 score of the iteratively trained initial random forest model is tested based on the test set. If the F1 score of the initial random forest model after the current iteration is less than the F1 score of the initial random forest model after the previous iteration, then adjust the learning rate of the initial random forest model and continue iterative training. If the F1 score of the initial random forest model after the current iteration is greater than or equal to the F1 score of the initial random forest model after the previous iteration, then stop the iteration training and determine the initial random forest model after the current iteration as the random forest model. The stranding speed is determined by substituting the number of stranded cores into the random forest model.

[0009] Furthermore, when obtaining the historical stranding yield and the real-time stranding yield at the stated stranding speed, and determining the qualification of the stranding machine based on the relationship between the historical stranding yield and the real-time stranding yield, the process includes: When the real-time stranding yield is greater than or equal to the historical stranding yield, the stranding machine is deemed qualified. If the real-time stranding yield is less than the historical stranding yield, the stranding machine is deemed unqualified.

[0010] Furthermore, when the stranding machine is determined to be defective, the cable preparation area is divided into several preparation data points, and a cable preparation influence series is constructed, including: The initial environmental data corresponding to each preparation data point is acquired, and the initial environmental data is preprocessed. The preprocessing includes data noise reduction and removal of erroneous data. The erroneous data is determined based on the parameters of the environmental sensor. The preparation environmental data is determined based on the results of the preprocessing. Extract all preparation environment data of the same type to construct a series of factors affecting cable preparation.

[0011] Furthermore, when constructing the influence sample based on the influence sequence of the cable, the process includes: Determine the standard preparation environment data corresponding to each preparation data point, and construct a cable standard preparation influence series by combining standard preparation environment data of the same type. The preparation environment data in the cable preparation influence series corresponds one-to-one with the standard preparation environment data in the cable standard preparation influence series. The preparation of the affected samples includes samples prepared in a suitable environment and samples prepared in an unsuitable environment. The fabrication environment data in the cable fabrication influence series is compared with the standard fabrication environment data in the cable standard fabrication influence series. If the preparation environment data is not equal to the standard preparation environment data, then construct the preparation environment mismatch sample; If the prepared environmental data are all equal to the standard prepared environmental data, then the environmental conformity sample is constructed.

[0012] Furthermore, when extracting the preparation environment data of the cable preparation influence series based on the sample quantity of the preparation influence samples, and determining whether to execute a model association strategy or a model mapping strategy based on the sample quantity, the process includes: The number of samples that match the environmental conditions affecting the construction of the series and the number of samples that do not match the environmental conditions are statistically analyzed. When the number of discrepancies is greater than 1, the model association strategy is then executed. When the number of discrepancies is equal to 1, the model mapping strategy is then executed.

[0013] Furthermore, when determining to execute the model association strategy, the determination of the preparation environment association result based on the extracted preparation environment data and association rule algorithm includes: Extract the manufacturing environment data that is not equal to the corresponding standard manufacturing environment data from the entire series of cable manufacturing influence data, and record them as the first manufacturing influence data; Different first preparation influence data are combined to form candidate item sets, and the support of each candidate item set in all candidate item sets is determined. Frequent itemsets that meet the minimum support threshold are selected, and the confidence level is determined based on the frequent itemsets to judge the validity of the association relationship. Based on the results of the support and confidence levels, the preparation environment association results between the first preparation influence data are determined.

[0014] Furthermore, when determining the target strand speed based on the aforementioned preparation environment correlation results or model mapping strategy, the following steps are included: Pre-train support vector machine and Bayesian models; The preparation environment correlation results and stranding speed are substituted into the Bayesian model to determine the target stranding speed; When it is determined that the model mapping strategy is to be executed, the manufacturing environment data that is not equal to the corresponding standard manufacturing environment data in the cable manufacturing influence series is obtained and recorded as the second manufacturing influence data; The second set of influence data and strand speed are substituted into the support vector machine model to determine the target strand speed.

[0015] Compared with existing technologies, the advantages of this invention are as follows: By comparing the number of stranded cores of the conductor to be stranded with a historical stranding database, a stranding strategy or prediction strategy is selected to determine the stranding speed, avoiding the blindness of setting the stranding speed based on experience. Furthermore, matching the stranding speed to different numbers of stranded cores reduces problems such as uneven stranding tension and broken wires caused by differences in the number of cores, thereby improving the quality and efficiency of cable manufacturing. By comparing the historical stranding yield with the real-time yield, the qualification of the stranding machine is judged, avoiding the risk of mass production of defective cables. At the same time, the cable manufacturing area is divided into several manufacturing data points and manufacturing data points are constructed. This approach incorporates environmental data into the overall analysis. Based on the number of samples used to prepare the influencing samples, a model association strategy or a model mapping strategy is selected. This avoids the limitations of a single model adapting to different environmental conditions. By using association rule algorithms to mine the correlation relationships in the prepared environmental data, the environmental factors affecting stranding speed and quality can be identified. Whether the target stranding speed is determined based on environmental association results or model mapping strategies, the approach can dynamically respond to changes in environmental factors such as temperature, humidity, and dust concentration. This ensures the stability and reliability of cable preparation in complex environments, ultimately improving the cable yield and enhancing the automation of the cable preparation process.

[0016] On the other hand, this application also provides a flexible mineral-insulated anti-interference fire-resistant cable, applied to a method for preparing the aforementioned flexible mineral-insulated anti-interference fire-resistant cable, comprising: Several conductors; Mineral mica layer: used to wrap the conductor, providing fireproof and insulating protection; Cross-linked insulating layer: used to coat the mineral mica layer, the cross-linked insulating layer is made of cross-linked polyethylene material; Flame-retardant filler layer: fills the gaps around the cross-linked insulation layer; Ceramic fireproof layer: used to wrap the flame-retardant filler layer; Stainless steel armor layer: used to wrap the ceramic fireproof layer; High flame-retardant outer sheath: used to wrap the stainless steel armor layer, the high flame-retardant outer sheath is made of halogen-free low-smoke flame-retardant polyolefin material.

[0017] It is understandable that the above-mentioned flexible mineral-insulated anti-interference fireproof cable and its preparation method have the same beneficial effects, and will not be elaborated further here. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a method for preparing a flexible mineral-insulated, anti-interference, fire-resistant cable according to an embodiment of the present invention; Figure 2 This is a structural schematic diagram of a flexible mineral-insulated anti-interference fireproof cable provided in an embodiment of the present invention.

[0019] The components include: 1. High flame-retardant outer sheath; 2. Stainless steel armor layer; 3. Ceramic fireproof layer; 4. Flame-retardant filler layer; 5. Cross-linked insulation layer; 6. Mineral mica layer; 7. Conductor. Detailed Implementation

[0020] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] See Figure 1 As shown in some embodiments of this application, a method for preparing a flexible mineral-insulated anti-interference fire-resistant cable includes: S100: Obtain the number of stranded cores of the conductor to be stranded, compare the number of stranded cores with the historical stranded wire database, determine the stranding strategy or prediction strategy based on the comparison results, and determine the stranding speed according to the stranding strategy or prediction strategy.

[0022] S200: Obtain historical stranded wire yield and real-time stranded wire yield at stranding speed. Determine the qualification of the stranding machine based on the relationship between historical and real-time stranded wire yield. When the stranding machine is deemed unqualified, divide the cable preparation area into several preparation data points and construct a cable preparation influence series.

[0023] S300: Construct a sample of influences based on the influence series of cable manufacturing, extract the environmental data of the influence series of cable manufacturing based on the number of samples, and determine the model association strategy or model mapping strategy to be executed based on the number of samples.

[0024] S400: When it is determined to execute the model association strategy, the preparation environment association result is determined based on the extracted preparation environment data and association rule algorithm. The target stranding speed is determined based on the preparation environment association result or the model mapping strategy, and the stranding is completed at the target stranding speed.

[0025] Specifically, the conductors to be stranded are stranded using a stranding machine. The number of stranded cores refers to the total number of strands involved in the stranding process. A higher number of stranded cores indicates greater structural complexity of the insulated cable and requires higher manufacturing precision. First, the number of stranded cores in the conductor to be stranded is obtained and compared with a historical stranding database. This database is a structured database storing different stranding core numbers and stranding speeds from previous years. If a matching record exists in the database, a stranding strategy is adopted, directly calling a historically verified and feasible stranding speed to avoid the risk of repeated trial and error. If no matching record exists, it indicates a discrepancy between the historical data and the current manufacturing process. The stranding speed is then inferred based on the difference between the historical database and the number of stranded cores. This comparison with historical data avoids the bias caused by setting the stranding speed based on human experience. Multiple stranded conductor segments are collected within 1-2 minutes at a predetermined stranding speed. The specific time can be dynamically adjusted according to the required precision in the insulated cable manufacturing process. Image acquisition and various testing equipment are used to determine the real-time stranding yield of these stranded conductors. The stranding yield is the percentage of qualified conductors out of the total conductors; a higher yield indicates better conductor quality. Historical stranding yields are obtained, and the real-time yield is compared with historical yields to assess the stranding machine's competence. In cable stranding, two variables affect the stranding yield: the number of stranded cores (determining stranding difficulty) and the manufacturing environment (temperature, humidity, dust, etc., affecting processing characteristics). Changes in the manufacturing environment can interfere with the stranding machine's insulated cable manufacturing process. By comparing the real-time and historical yields, if there is a difference, it indicates that the real-time yield may not meet the requirements for good cable manufacturing. If the stranding speed needs to be adjusted to adapt to the change in the preparation environment, the stranding machine is deemed unqualified. The cable preparation area is divided into several preparation data points. The cable preparation area refers to the overall working area during stranding operations. The number of preparation data points is dynamically adjusted according to the size of the cable preparation area. This embodiment does not limit the specific number. Data from each preparation data point is collected in real time by devices such as temperature sensors, humidity sensors, and particle sensors to construct a cable preparation influence series, thereby transforming the fuzzy environmental interference into a quantifiable variable cable preparation influence series.

[0026] Understandably, the influence samples are constructed based on the influence series. These influence samples reflect the impact of various environmental types on cable manufacturing. Under different environmental influences, the manufacturing environment data in the cable manufacturing influence series is extracted based on the sample size of the influence samples. If the sample size is large, it indicates that multiple different environmental types of manufacturing environment data may have a comprehensive impact on manufacturing. If the sample size is small, it indicates that only one type of manufacturing environment data may have a comprehensive impact on manufacturing. By dynamically executing the model association strategy or model mapping strategy through the sample size, the adaptability of the stranding speed to the environment during cable manufacturing is ensured. If the model association strategy is executed, the correlation between data is mined by extracting the preparation environment data and association rule algorithm, such as "when the temperature is 25-28℃, the humidity is usually 40%-50%", thereby determining the environmental correlation result and then obtaining the target stranding speed. If the model mapping strategy is executed, a target stranding speed that adapts to the environment is predicted by using the currently extracted preparation environment data and deep learning model, and stranding is completed at the target stranding speed. Even if the temperature and humidity fluctuate and the dust changes, the reliability of the prepared cable can be ensured by the environmental correlation result or the model mapping strategy, thereby ensuring the performance of the insulated cable. In some embodiments of this application, when comparing the number of stranded wire cores with a historical stranded wire database, determining a stranding strategy or prediction strategy based on the comparison results, and determining the stranding speed according to the stranding strategy or prediction strategy, the process includes: the historical stranded wire database includes several historical stranded wire core numbers and several historical stranded wire speeds, and each historical stranded wire core number corresponds to a historical stranded wire core number; when the historical stranded wire database contains a historical stranded wire core number with the same number of stranded wire cores, the stranding strategy is determined to be executed, and the stranding strategy determines the historical stranded wire speed corresponding to the historical stranded wire core number with the same number of stranded wire cores as the stranding speed; when the historical stranded wire database does not contain a historical stranded wire core number with the same number of stranded wire cores, the prediction strategy is determined to be executed, and the prediction strategy determines the stranding speed based on a random forest model and the number of stranded wire cores.

[0027] Specifically, the historical stranded wire database has a one-to-one correspondence between the historical stranded wire core count and the historical stranded wire speed. This correspondence originates from adaptation data in past manufacturing processes, ensuring the reliability of the matching between each set of stranded wire core counts and corresponding stranded wire speeds. When the same historical stranded wire core count can be found in the historical stranded wire database, the stranding strategy is executed, directly using the historical stranded wire speed corresponding to the same historical stranded wire core count as the current stranded wire speed. This essentially reuses mature process parameters. When no matching historical stranded wire core count exists, a prediction strategy is executed, using a random forest model to determine the stranded wire speed. The random forest model excels at integrating multi-dimensional data. Based on the characteristics of potential correlations, the model learns the inherent patterns of historical stranded wire core counts and historical stranding speeds with similar core counts in the historical stranded wire database. This fills the gaps when there is no direct matching data. The random forest model is derived based on existing data, reducing the blind spots in cable preparation and shortening the process time. At the same time, it ensures the reliability of stranding speed and lays the foundation for subsequent optimization of stranding speed by incorporating environmental factors. This avoids stranding defects caused by improper stranding speed, thus meeting the stranding accuracy requirements of flexible mineral-insulated anti-interference fireproof cables, and improving the overall process efficiency and stability.

[0028] In some embodiments of this application, when the prediction strategy determines the strand speed based on the random forest model and the number of strand cores, it includes: acquiring a conductor dataset and merging it with a historical strand database; determining the strand dataset; pre-selecting an initial random forest model, where the number of trees in the initial random forest model is set to 200, and the maximum tree depth is adjusted based on the input strand core number distribution; the minimum number of samples for node splits is set to 10, the minimum number of samples for leaf nodes is set to 5, and the maximum number of leaf nodes is 100; dividing the strand dataset into a training set and a test set; and iteratively training the initial random forest model based on the training set. The test set is used to test the F1 score of the initial random forest model after iterative training. If the F1 score of the initial random forest model after the current iteration is less than the F1 score of the initial random forest model after the previous iteration, the learning rate of the initial random forest model is adjusted and iterative training continues. If the F1 score of the initial random forest model after the current iteration is greater than or equal to the F1 score of the initial random forest model after the previous iteration, iterative training is stopped, and the initial random forest model after the current iteration is determined as the random forest model. The number of stranded cores is substituted into the random forest model to determine the stranding speed.

[0029] Specifically, the conductor dataset contains various data affecting stranding characteristics, such as the material and diameter of the conductor to be stranded. This dataset is merged with the historical stranding database to determine the stranding dataset, avoiding the risk of insufficient data dimensionality and biased pattern discovery caused by relying solely on the historical stranding database. The number of trees is set to 200 because random forests require multiple decision trees to vote on each other to reduce variance and improve stability. If the number of trees is too small (e.g., <100), the model is easily affected by the bias of a single tree, making it difficult to cover the correlation between different historical strand core numbers and historical stranding speeds. This could lead to situations where the input historical strand core numbers are similar, but the predicted stranding speeds differ significantly. If the number of trees is too large (e.g., >300), although variance can be further reduced, computation time will increase, which does not meet the efficiency requirement of quickly determining stranding speed in cable manufacturing. The 200-tree ensemble effectively cancels out noise from individual trees, ensuring prediction stability for different strand counts. It also controls training and prediction time, adapting to the real-time requirements of the manufacturing process. The maximum tree depth is adjusted based on the strand count distribution. If a large number of historical strand counts in the dataset are less than or equal to 10, it indicates a simple correlation between historical strand counts and historical stranding speeds. Overly deep trees are prone to overfitting local data (e.g., treating historical stranding speeds for individual strand counts as a pattern). Therefore, the maximum tree depth is set to 8. If a large number of historical strand counts in the dataset are greater than 10, it indicates a more dispersed core count distribution (e.g., core counts from 11 to 30 are included). Shallow trees cannot capture this segmented pattern. Therefore, the maximum tree depth is set to 15. This allows the model to avoid overfitting in simple distribution scenarios while fully capturing patterns in complex distribution scenarios, ensuring that different strand counts yield stranding speeds that closely match actual processes. The minimum number of samples for node splitting is set to 10 to avoid unreliable rules caused by small sample splitting. Node splitting is the key to dividing data in decision trees. If the data sample of the twisted wire dataset is small, the split child nodes may only reflect the accidental correlation of individual data, resulting in the problem that the input historical twisted wire core numbers are similar, but the predicted twisted wire speeds are very different. Setting the minimum number of samples for node splitting to 10 can ensure that the split nodes have certain statistical significance and avoid the impact of local data on the overall prediction reliability. The minimum number of samples for leaf nodes is set to 5. Leaf nodes are the terminal of the model output twisted wire speed. If the minimum number of samples is less than 5, the "core number interval" corresponding to the leaf node is too narrow, and it is easy to include noisy data in the prediction basis (such as two core number samples being used as the standard of the interval due to accidental environmental fluctuations). The minimum number of samples of 5 can ensure that the output of the twisted wire speed within the core number interval covered by the leaf node is stable and reliable, reducing twisting defects caused by prediction deviation.The maximum number of leaf nodes is set to 100, which balances the model's complexity and generalization ability. Too many leaf nodes (e.g., >150) will cause the model to overfit (treating random fluctuations in historical data as inevitable patterns), resulting in a large prediction deviation when facing new strand numbers. Too few leaf nodes (e.g., <50) will cause the model to oversimplify the patterns and fail to accurately match the stranding requirements of different strand numbers. 100 leaf nodes can cover the common core number range in cable manufacturing while avoiding model redundancy. This ensures that the model finds a balance between accurate prediction and avoiding overfitting, providing a reliable basis for subsequent optimization of stranding speed in combination with environmental factors.

[0030] Understandably, the random forest dataset is divided into training and testing sets in a 4:1 ratio to ensure the model's generalization ability. The training set is used to train the initial random forest model, while the testing set is used to evaluate the performance of the trained model. The initial random forest model is iteratively trained using data from the training set. In each iteration, the model attempts to learn patterns and relationships in the data to improve its predictive ability. After each iteration, the model is tested using data from the testing set. If the F1 score of the initial random forest model after the current iteration is lower than that after the previous iteration, it indicates a decline in model performance. In this case, the learning rate of the initial random forest model needs to be adjusted, using methods such as cosine annealing. Then, iterative training continues until the F1 score of the initial random forest model after the current iteration is greater than or equal to that after the previous iteration. This helps the model more stably approach the global optimum. If the F1 score of the initial random forest model after the current iteration is greater than or equal to the F1 score of the random forest model after the previous iteration, it indicates that the model performance has improved or remained stable. At this point, the iterative training can be stopped, and the model is considered to have reached a satisfactory performance level. The initial random forest model after the current iteration is then determined as the random forest model. By training the initial random forest model, the stranding speed corresponding to each strand core number can be accurately output, so as to avoid judgment errors and improve the stability of the cable manufacturing process.

[0031] In some embodiments of this application, when obtaining the historical stranding yield and the real-time stranding yield at the stranding speed, and determining the qualification of the stranding machine based on the relationship between the historical stranding yield and the real-time stranding yield, the following steps are taken: when the real-time stranding yield is greater than or equal to the historical stranding yield, the stranding machine is determined to be qualified; when the real-time stranding yield is less than the historical stranding yield, the stranding machine is determined to be unqualified.

[0032] Specifically, the historical yield rate is the average pass rate of conductors proven in past cable manufacturing. When the real-time yield rate is greater than or equal to the historical yield rate, it indicates that the stranding accuracy of the stranding machine can stably produce qualified conductors under the current manufacturing environment. The stranding machine is then deemed qualified, and cable manufacturing can continue at the current stranding speed. When the real-time yield rate is less than the historical yield rate, it indicates that a certain number of unqualified conductors exist after stranding. The stranding machine is then deemed unqualified. Since stranding machines are usually regularly maintained and serviced, this unqualification is not caused by the stranding machine itself (such as component wear or control system failure), but rather by the surrounding environment disrupting the equipment's parameters. The need for a proper balance between the quantity and manufacturing process characteristics makes it impossible to output qualified conductors according to the stranding speed benchmark. For example, excessive humidity can cause moisture to adhere to the surface of the conductor to be stranded, increasing the frictional resistance with the stranding machine's guide wheel. Alternatively, excessive dust concentration can accumulate on the surface of components such as the stranding cage and guide wheel of the stranding machine. Fine dust can alter the fit between components, leading to a risk of reduced stranding speed. By comparing the real-time stranding yield with the historical stranding yield, the continuity of cable manufacturing is ensured, thereby promptly identifying the qualification of the stranding machine and avoiding the waste of raw materials caused by the continuous production of defective conductors, further improving the stability of the cable manufacturing process.

[0033] In some embodiments of this application, when the stranding machine is determined to be unqualified, the cable preparation area is divided into several preparation data points, and a cable preparation influence series is constructed. This includes: obtaining the initial environmental data corresponding to each preparation data point, and preprocessing the initial environmental data. The preprocessing includes data noise reduction and removal of erroneous data. The erroneous data is determined based on the parameters of the environmental sensor. Based on the results of the preprocessing, the preparation environmental data is determined, and all preparation environmental data of the same type are extracted to construct a cable preparation influence series.

[0034] Specifically, environmental sensors include devices such as temperature sensors, humidity sensors, and particle sensors. These sensors collect initial environmental data from each preparation data point in real time. During data acquisition, the initial environmental data may contain noise, errors, and invalid information due to objective interference in the acquisition process. For example, when environmental sensors operate in industrial environments, they are susceptible to voltage fluctuations and mechanical vibrations, resulting in instantaneous value jumps. This type of "noise data" is not a reflection of actual environmental changes but rather an error in the sensor's acquisition. Preprocessing eliminates these interferences, and data denoising filters out these occasional sensor fluctuations, such as value jumps caused by instantaneous voltage instability. Based on the environmental sensor parameters, such as sensor range and accuracy thresholds, erroneous data, such as outliers exceeding the range, is removed to ensure that the retained preparation environmental data accurately reflects the actual environmental state of each preparation data point. Finally, similar types of preparation environmental data, such as temperature or humidity data from all preparation data points, are extracted to construct a cable preparation influence series. This transforms the dispersed environmental data into an ordered and analyzable sequence. The ordered preparation influence series allows for the quantification and traceability of environmental factors, providing reliable environmental data support for determining the target stranding speed.

[0035] In some embodiments of this application, when constructing a preparation influence sample based on the cable preparation influence series, the process includes: determining the standard preparation environment data corresponding to each preparation data point, and constructing a cable standard preparation influence series with the same type of standard preparation environment data. The preparation environment data in the cable preparation influence series corresponds one-to-one with the standard preparation environment data in the cable standard preparation influence series. The preparation influence sample includes preparation environment conforming samples and preparation environment non-conforming samples. The preparation environment data in the cable preparation influence series is compared with the standard preparation environment data in the cable standard preparation influence series. If there are preparation environment data that are not equal to the standard preparation environment data, then a preparation environment non-conforming sample is constructed. If all preparation environment data are equal to the standard preparation environment data, then an environment conforming sample is constructed.

[0036] Specifically, the standard preparation environment data is determined based on the equipment parameter range of the stranding machine. For example, the rated operating temperature range of the stranding machine's motor is -5℃ to 40℃. When the temperature exceeds 40℃, the motor's heat dissipation efficiency decreases, causing fluctuations in the stranding speed and thus affecting the stranding pitch accuracy. Therefore, the standard preparation environment data for this temperature range is set to -5℃ to 40℃. Alternatively, if the stranding machine's guide wheel bearing uses a precision ball bearing structure, its moisture-proof rating requires the ambient humidity to not exceed 60%. Excessive humidity accelerates bearing corrosion and increases the coefficient of friction, leading to unstable wire traction and affecting the fluctuation of the stranding speed. Therefore, the standard preparation environment data for humidity is limited to within 60%, and the same standard preparation environment data is used for the same type of stranding machine. A standard cable manufacturing influence series was constructed, ensuring a one-to-one correspondence between the standard and actual cable manufacturing influence series in terms of data type and collection location, guaranteeing comparability. Subsequently, a point-by-point comparison of the two series was used to construct manufacturing influence samples. If the manufacturing environment data differed from the standard manufacturing environment data, a discrepancy sample was constructed. For example, if the temperature was below -5℃ or above 40℃, a discrepancy sample was constructed for the stranding machine, reflecting the manufacturing impact caused by temperature deviations. If all manufacturing environment data equaled the standard manufacturing environment data, it indicated that the current ambient temperature did not affect the stranding machine's manufacturing, and a conforming sample was constructed. By comparing the cable manufacturing influence series with the standard cable manufacturing influence series, the impact of data such as temperature, humidity, and dust on cable manufacturing could be quantified. This allowed for the determination of which environmental types met the manufacturing requirements and which did not, avoiding misclassification of normal environmental data as interfering factors in subsequent analysis. Simultaneously, focusing on existing abnormal data ensured the accuracy of the environmental optimization direction derived from the manufacturing influence samples, guaranteeing the reliability of cable manufacturing.

[0037] In some embodiments of this application, when extracting the preparation environment data of the cable preparation influence series based on the number of samples of the preparation influence samples, and determining whether to execute a model association strategy or a model mapping strategy based on the number of samples, the process includes: counting the number of samples that conform to the environment of the entire cable preparation influence series and the number of samples that do not conform to the preparation environment; when the number of discrepancies is greater than 1, the model association strategy is determined to be executed; when the number of discrepancies is equal to 1, the model mapping strategy is determined to be executed.

[0038] Specifically, since each cable preparation influence series is compared with the cable standard preparation influence series, an environmental discrepancy sample or an environmental conformity sample will be generated. When the stranding machine is deemed unqualified, there will definitely be an environmental type that deviates from the stranding machine standard. When the number of discrepancies is greater than 1, it means that at least two or more environmental types affect the stranding machine, such as temperature and humidity. When the number of discrepancies is equal to 1, it means that one environmental type affects the stranding machine. When at least two or more environmental types affect the stranding machine, the relationship between environmental factors and stranding speed is more complex. Therefore, a model association strategy is used to uncover hidden patterns among multiple variables. When one environmental type affects the stranding machine, it means that the variables of environmental anomalies are more concentrated. Without complex association analysis, the corresponding improvement measures can be quickly matched through a model mapping strategy. Environmental deviations caused by various environmental types can be comprehensively captured using a model association strategy, ensuring that subsequent adjustments to the stranding speed take into account the influence of multiple environmental factors. Conversely, environmental deviations caused by a single environmental type can be addressed using a model mapping strategy, which shortens the analysis time and quickly outputs an optimization direction suitable for the stranding speed, avoiding unnecessary resource consumption. The flexible switching between these two strategies ensures both the accuracy of environmental analysis and the efficiency of the cable manufacturing process.

[0039] In some embodiments of this application, when determining the execution of the model association strategy, the determination of the preparation environment association result based on the extracted preparation environment data and the association rule algorithm includes: extracting preparation environment data that are not equal to the corresponding standard preparation environment data from all cable preparation influence data series, and recording them as first preparation influence data; combining different first preparation influence data to form candidate item sets; determining the support of each candidate item set in all candidate item sets; screening out frequent itemsets that meet the minimum support threshold; determining the confidence level based on the frequent itemsets to judge the validity of the association relationship; and determining the preparation environment association result between the first preparation influence data based on the support and confidence results.

[0040] Specifically, firstly, the first set of preparation impact data (i.e., all preparation environment data that do not conform to the standard preparation environment data) is extracted. Different first set of preparation impact data are combined into candidate item sets, such as combinations of excessive temperature and excessive humidity, or excessive dust and excessive temperature, etc. Then, the support of each candidate item set is calculated, i.e., the frequency of the combination appearing in all candidate item sets. Frequent itemsets that meet the minimum support threshold are selected, i.e., anomalous combinations with high frequency and statistical significance. The confidence level is determined based on these frequent itemsets, such as "the probability that humidity also exceeds the standard when temperature exceeds the standard," thereby verifying the correlation between the first set of preparation impact data. A higher confidence level indicates a stronger correlation. The more likely the two sets of first preparation influence data are to appear together, the less likely they are to be accidental. Finally, the preparation environment association results between the first preparation influence data are determined by combining support (reflecting the generality of the combination) and confidence (reflecting the reliability of the association), such as "when the temperature is 25-28℃, the humidity is usually 40%-50%". The minimum support threshold is determined according to the actual process scenario of cable preparation and the number of first preparation influence data. This implementation does not limit the minimum support threshold to ensure that abnormal combinations with actual process significance are retained, while meaningless combinations that only appear by chance are eliminated. The confidence determination combines the dual standards of "statistical significance" and "process rationality". Statistically, the confidence level must be significantly higher than the random probability. For example, if the probability of humidity exceeding the standard when temperature exceeds the standard is much higher than the base probability of humidity exceeding the standard alone, it indicates a non-random correlation between the two. From a technological perspective, it must conform to the interaction patterns of environmental parameters, such as the physical logic that air humidity tends to decrease under high temperature conditions. If the confidence level of a combination is high (e.g., 90%) but violates common sense in the process, such as the uncommon occurrence of high temperature and high humidity simultaneously in a cable manufacturing workshop, it is not considered a valid high confidence level. In practice, high confidence levels are usually manifested as values ​​significantly higher than most other combinations (e.g., exceeding 60%-70%), ensuring that the correlation truly reflects the synergistic effect of the environment on the stranding speed. The association rule algorithm can transform scattered and different types of primary manufacturing influence data into logically related anomalous combinations, while excluding accidental anomalous combinations, ensuring the reliability of the manufacturing environment correlation results and thus improving the efficiency of the cable manufacturing process.

[0041] In some embodiments of this application, when determining the target stranded wire speed based on the preparation environment correlation results or model mapping strategy, the process includes: pre-training a support vector machine model and a Bayesian model; substituting the preparation environment correlation results and stranded wire speed into the Bayesian model to determine the target stranded wire speed; when it is determined to execute the model mapping strategy, acquiring the preparation environment data in the cable preparation influence series that is not equal to the corresponding standard preparation environment data, and recording it as the second preparation influence data; and substituting the second preparation influence data and stranded wire speed into the support vector machine model to determine the target stranded wire speed.

[0042] Specifically, the Support Vector Machine (SVM) and Bayesian models are pre-trained, and the training process is consistent with the initial Random Forest model, which will not be repeated here. The Bayesian model excels at probabilistic association analysis, capable of understanding the adjustment patterns of strand speed under the combined influence of multiple anomalies. The SVM model, on the other hand, excels at rapid matching with small samples, capable of understanding the mapping relationship between a single anomaly environment type and strand speed. When executing the model association strategy, the preparation environment association results clearly define the synergistic relationships of multiple sets of first preparation influence data. Substituting this data, along with the current strand speed, into the Bayesian model, the output is a target strand speed adapted to the interactive influence of multiple anomalies to complete cable preparation. When executing the model mapping strategy, the second preparation influence data consists only of preparation environment data of a single environment type, such as temperature or humidity. Substituting this data, along with the current strand speed, into the SVM model, the SVM model can output a target strand speed adapted to the anomaly conditions of that environment to complete cable preparation. Under the model association strategy, when faced with the combined interference of multiple environmental anomalies, the probabilistic association capability of the Bayesian model can integrate multivariate information and avoid ignoring the interaction between anomalies. Under the model mapping strategy, when faced with the interference of one type of environmental anomaly, the mapping capability of the support vector machine model can eliminate complex association analysis and ensure the timeliness of cable manufacturing process. Whether it is the influence of one type of environment or the mutual influence of multiple types of environment, the model association strategy and the model mapping strategy cover different environmental anomaly scenarios at the same time, which can effectively improve the stability and efficiency of cable manufacturing.

[0043] In summary, the beneficial effects of this invention are as follows: By comparing the number of stranded cores of the conductor to be stranded with a historical stranding database, a stranding strategy or prediction strategy is selected to determine the stranding speed, avoiding the blindness of setting the stranding speed based on experience. Furthermore, matching the stranding speed to different numbers of stranded cores reduces problems such as uneven stranding tension and broken wires caused by differences in core count, thereby improving the quality and efficiency of cable manufacturing. By comparing the historical stranding yield with the real-time yield, the passability of the stranding machine is judged, avoiding the risk of mass production of defective cables. Simultaneously, the cable manufacturing area is divided into several manufacturing data points and manufacturing data points are constructed, thereby… By incorporating environmental data into the calculation, a model association strategy or a model mapping strategy is selected based on the number of samples used to prepare the influencing samples. This avoids the limitations of a single model in adapting to different environmental conditions. The association rule algorithm is used to mine the correlation relationships in the prepared environmental data, thereby identifying the environmental factors affecting the stranding speed and quality. Whether the target stranding speed is determined based on the environmental association results or the model mapping strategy, it can dynamically respond to changes in environmental factors such as temperature, humidity, and dust concentration. This ensures the stability and reliability of cable preparation in complex environments, ultimately improving the cable yield and enhancing the automation of the cable preparation process.

[0044] In another preferred embodiment based on the above embodiments, see [reference] Figure 2 As shown, this embodiment provides a flexible mineral-insulated anti-interference fire-resistant cable, applied to the preparation method of the aforementioned flexible mineral-insulated anti-interference fire-resistant cable, including: 7. Several conductors.

[0045] Mineral mica layer 6: Used to wrap conductor 7, providing fireproof and insulating protection for conductor 7.

[0046] Cross-linked insulation layer 5: used to coat the mineral mica layer 6. The cross-linked insulation layer 5 is made of cross-linked polyethylene material.

[0047] Flame-retardant filler layer 4: fills the gaps around the cross-linked insulation layer 5.

[0048] Ceramic fireproof layer 3: used to wrap the flame-retardant filler layer 4.

[0049] Stainless steel armor layer 2: used to wrap the ceramic fireproof layer 3.

[0050] High flame retardant outer sheath 1: used to wrap the stainless steel armor layer 2. The high flame retardant outer sheath 1 is made of low smoke halogen-free low smoke flame retardant polyolefin material.

[0051] Specifically, Figure 2This is a schematic diagram of the cross-sectional structure of a flexible mineral-insulated anti-interference fire-resistant cable. The cable is prepared by stranding wires into conductor 7, thus forming a flexible mineral-insulated anti-interference fire-resistant cable to achieve fireproof, anti-interference, and insulation functions. Conductor 7 is the core of power transmission. Preferably, there are five conductors 7. A mineral mica layer 6, utilizing its fire-resistant and insulating properties, wraps around conductor 7, blocking flame attack and maintaining the cable's insulation performance in case of fire. The cross-linked insulation layer 5 uses cross-linked polyethylene material, which further enhances its heat resistance and insulation performance after cross-linking, strengthening the insulation protection of conductor 7 and forming a double insulation guarantee with the mineral mica layer 6. A flame-retardant filler layer 4 fills the gaps around the cross-linked insulation layer 5, making the cable structure more compact and preventing the spread of flames within the gaps using flame-retardant materials. A ceramic fireproof layer 3 wraps around the flame-retardant filler layer 4. When exposed to fire, the ceramic fireproof layer 3 can ceramicize to form a hard fireproof layer. The cable features a barrier that effectively blocks high temperatures and flames. The stainless steel armor layer 2 enhances the cable's mechanical strength to withstand external impacts and reduces electromagnetic interference through its metallic shielding effect. The high flame-retardant outer sheath 1 uses halogen-free, low-smoke, flame-retardant polyolefin material, providing outer protection while producing less smoke and releasing no toxic halides. A multi-layered fireproof system is constructed from the inside out, consisting of a mineral mica layer 6, a cross-linked insulation layer 5, a flame-retardant filler layer 4, a ceramic fireproof layer 3, the stainless steel armor layer 2, and the high flame-retardant outer sheath 1. Electromagnetic shielding is achieved through the stainless steel armor layer 2 to adapt to strong electromagnetic environments, enhancing the safety of the flexible mineral-insulated anti-interference fire-resistant cable. This system meets multiple requirements, including flexibility, mineral insulation, anti-interference, and fire resistance.

[0052] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0053] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0054] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0055] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for preparing a flexible mineral-insulated, interference-resistant, fire-resistant cable, characterized in that, include: Obtain the number of stranded cores of the conductor to be stranded, compare the number of stranded cores with the historical stranded wire database, determine the stranding strategy or prediction strategy based on the comparison results, and determine the stranding speed according to the stranding strategy or prediction strategy. Obtain the historical stranded wire yield and the real-time stranded wire yield at the stated stranding speed. Determine the qualification of the stranding machine based on the relationship between the historical stranded wire yield and the real-time stranded wire yield. When the stranding machine is determined to be unqualified, divide the cable preparation area into several preparation data points and construct a cable preparation influence series. Based on the cable preparation influence series, preparation influence samples are constructed. The preparation environment data of the cable preparation influence series is extracted according to the sample size of the preparation influence samples. The execution model association strategy or model mapping strategy is determined according to the sample size. When it is determined to execute the model association strategy, the preparation environment association result is determined based on the extracted preparation environment data and association rule algorithm, the target stranding speed is determined based on the preparation environment association result or the model mapping strategy, and the stranding is completed at the target stranding speed.

2. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 1, characterized in that, When comparing the number of stranded cores with a historical stranded wire database, determining a stranding strategy or prediction strategy based on the comparison results, and determining the stranding speed according to the stranding strategy or prediction strategy, the process includes: The historical stranded wire database includes several historical stranded wire core counts and several historical stranded wire speeds, and each historical stranded wire core count corresponds to a historical stranded wire core count. When the historical stranded wire database contains a historical stranded wire number that is the same as the number of stranded wire cores, the stranded wire strategy is determined to be executed. The stranded wire strategy determines the historical stranded wire speed corresponding to the historical stranded wire number that is the same as the number of stranded wire cores as the stranded wire speed. If the historical stranded wire database does not contain a historical stranded wire core number that is the same as the stated stranded wire core number, then the prediction strategy is executed. The prediction strategy determines the stranding speed based on the random forest model and the stated stranded wire core number.

3. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 2, characterized in that, When the prediction strategy determines the stranding speed based on the random forest model and the number of stranded cores, it includes: Obtain the conductor dataset and merge it with the historical stranded wire database to determine the stranded wire dataset; An initial random forest model is pre-selected, with the number of trees in the initial random forest model set to 200, and the maximum depth of the trees adjusted based on the distribution of the number of strands in the input. The minimum number of samples for node splitting is set to 10, the minimum number of samples for leaf nodes is set to 5, and the maximum number of leaf nodes is 100. The twisted wire dataset is divided into a training set and a test set. The initial random forest model is iteratively trained based on the training set, and the F1 score of the iteratively trained initial random forest model is tested based on the test set. If the F1 score of the initial random forest model after the current iteration is less than the F1 score of the initial random forest model after the previous iteration, then adjust the learning rate of the initial random forest model and continue iterative training. If the F1 score of the initial random forest model after the current iteration is greater than or equal to the F1 score of the initial random forest model after the previous iteration, then stop the iteration training and determine the initial random forest model after the current iteration as the random forest model. The stranding speed is determined by substituting the number of stranded cores into the random forest model.

4. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 3, characterized in that, When obtaining historical stranding yield and real-time stranding yield at the stated stranding speed, and determining the qualification of the stranding machine based on the relationship between the historical and real-time stranding yields, the process includes: When the real-time stranding yield is greater than or equal to the historical stranding yield, the stranding machine is deemed qualified. If the real-time stranding yield is less than the historical stranding yield, the stranding machine is deemed unqualified.

5. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 4, characterized in that, When the stranding machine is determined to be defective, the cable preparation area is divided into several preparation data points, and a cable preparation influence series is constructed, including: The initial environmental data corresponding to each preparation data point is acquired, and the initial environmental data is preprocessed. The preprocessing includes data noise reduction and removal of erroneous data. The erroneous data is determined based on the parameters of the environmental sensor. The preparation environmental data is determined based on the results of the preprocessing. Extract all preparation environment data of the same type to construct a series of factors affecting cable preparation.

6. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 5, characterized in that, When constructing the influence sample based on the influence sequence of the cable, the following steps are included: Determine the standard preparation environment data corresponding to each preparation data point, and construct a cable standard preparation influence series by combining standard preparation environment data of the same type. The preparation environment data in the cable preparation influence series corresponds one-to-one with the standard preparation environment data in the cable standard preparation influence series. The preparation of the affected samples includes samples prepared in a suitable environment and samples prepared in an unsuitable environment. The fabrication environment data in the cable fabrication influence series is compared with the standard fabrication environment data in the cable standard fabrication influence series. If the preparation environment data is not equal to the standard preparation environment data, then construct the preparation environment mismatch sample; If the prepared environmental data are all equal to the standard prepared environmental data, then the environmental conformity sample is constructed.

7. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 6, characterized in that, When extracting the preparation environment data of the cable preparation influence series based on the sample quantity of the preparation influence samples, and determining whether to execute a model association strategy or a model mapping strategy based on the sample quantity, the process includes: The number of samples that match the environmental conditions affecting the construction of the series and the number of samples that do not match the environmental conditions are statistically analyzed. When the number of discrepancies is greater than 1, the model association strategy is then executed. When the number of discrepancies is equal to 1, the model mapping strategy is then executed.

8. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 7, characterized in that, When determining to execute the model association strategy, the process of determining the association result of the preparation environment based on the extracted preparation environment data and the association rule algorithm includes: Extract the manufacturing environment data that is not equal to the corresponding standard manufacturing environment data from the entire series of cable manufacturing influence data, and record them as the first manufacturing influence data; Different first preparation influence data are combined to form candidate item sets, and the support of each candidate item set in all candidate item sets is determined. Frequent itemsets that meet the minimum support threshold are selected, and the confidence level is determined based on the frequent itemsets to judge the validity of the association relationship. Based on the results of the support and confidence levels, the preparation environment association results between the first preparation influence data are determined.

9. The method for preparing the flexible mineral-insulated anti-interference fire-resistant cable according to claim 8, characterized in that, When determining the target strand speed based on the aforementioned environmental correlation results or model mapping strategy, the following is included: Pre-train support vector machine and Bayesian models; The preparation environment correlation results and stranding speed are substituted into the Bayesian model to determine the target stranding speed; When it is determined that the model mapping strategy is to be executed, the manufacturing environment data that is not equal to the corresponding standard manufacturing environment data in the cable manufacturing influence series is obtained and recorded as the second manufacturing influence data; The second set of influence data and strand speed are substituted into the support vector machine model to determine the target strand speed.

10. A flexible mineral-insulated anti-interference fire-resistant cable, applied to the preparation method of the flexible mineral-insulated anti-interference fire-resistant cable as described in any one of claims 1-9, characterized in that, include: Several conductors; Mineral mica layer: used to wrap the conductor, providing fireproof and insulating protection; Cross-linked insulating layer: used to coat the mineral mica layer, the cross-linked insulating layer is made of cross-linked polyethylene material; Flame-retardant filler layer: fills the gaps around the cross-linked insulation layer; Ceramic fireproof layer: used to wrap the flame-retardant filler layer; Stainless steel armor layer: used to wrap the ceramic fireproof layer; High flame-retardant outer sheath: used to wrap the stainless steel armor layer, the high flame-retardant outer sheath is made of halogen-free low-smoke flame-retardant polyolefin material.

Images