Intelligent corrosion-resistant cable manufacturing method

Abstract

The invention discloses an intelligent corrosion-resistant cable manufacturing method, and belongs to the field of cable manufacturing, and the manufacturing method comprises the following specific steps: (1) collecting and integrating material information of each corrosion-resistant cable; (2) recommending an optimal corrosion-resistant material combination and formulating a cable design scheme; (3) optimizing various process parameters in the cable manufacturing process; according to the method, key process parameters can be more efficiently explored and optimized, the manual intervention and experiment cost in a traditional trial and error method is reduced, the production cost is reduced, and the probability of problems in the production process is reduced, so that the stability of the manufacturing process and the overall quality of the cable are improved; the comprehensive effect of different production plans can be evaluated more comprehensively, it is ensured that the production plans can respond to changes quickly, high efficiency and competitiveness of production are kept, optimization results can be better understood and explained, and transparency of the decision making process is promoted.

Description

Technical Field

[0001] The invention relates to the field of cable manufacturing, and in particular to an intelligent corrosion-resistant cable manufacturing method. Background Art

[0002] In the modern shipping industry, marine cables play a vital role in modern shipping and port operations. Due to their use in extreme marine environments, the quality of their corrosion resistance is directly related to the safety and service life of the cables. As key marine equipment, marine cables are responsible for traction and fixation between ships and port facilities and other ships. Since marine cables are often exposed to salt spray, moisture and other corrosive substances in the marine environment, their corrosion resistance is essential. Traditional methods of manufacturing marine cables rely mainly on experience and manual adjustments, and usually face problems such as low production efficiency, poor quality stability and difficult cost control. With the development of materials science and intelligent manufacturing technology, how to use intelligent means to optimize the manufacturing process of marine cables has become a key issue that needs to be solved urgently; therefore, it is particularly important to invent an intelligent method for making corrosion-resistant cables.

[0003] The existing corrosion-resistant cable manufacturing method cannot efficiently explore and optimize key process parameters, increases the manual intervention and experimental costs in the traditional trial-and-error method, increases production costs, and reduces the stability of the manufacturing process and the overall quality of the cable; in addition, the existing corrosion-resistant cable manufacturing method cannot ensure the stability of the manufacturing process and the overall quality of the cable, and the production plan cannot respond to changes quickly, reducing the efficiency and competitiveness of production; for this reason, we propose an intelligent corrosion-resistant cable manufacturing method. Summary of the invention

[0004] The purpose of the present invention is to solve the defects in the prior art and to propose an intelligent corrosion-resistant cable manufacturing method.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] An intelligent method for manufacturing a corrosion-resistant cable, the specific steps of the manufacturing method are as follows:

[0007] (1) Collect and integrate information on various corrosion-resistant cable materials.

[0008] As a further solution of the present invention, the specific steps of collecting and integrating the corrosion-resistant cable material information in step (1) are as follows:

[0009] S1.1: Collect information on various corrosion-resistant cable materials from various data sources including academic papers, technical documents, patent databases, manufacturer data, and environmental data through web crawlers, API interfaces, and database exports;

[0010] S1.2: Eliminate redundant, irrelevant, and erroneous data, unify the data formats from different sources, use interpolation, mean filling, or deletion to handle missing data, and convert the processed data into text format;

[0011] S1.3: Use NLTK or spaCy natural language processing tools to segment the collected text data into words or phrases, remove stop words from the word segmentation results, restore the words to their basic form or roots, use NLP technology to label each word with its part of speech, identify the part of speech of each word, and then use the pre-trained NER model to identify specific types of entities;

[0012] S1.4: Analyze sentence structure through Stanford NLP to determine the dependency relationship between words. Then, based on the dependency relationship and context information, identify the relationship pattern between entities, unify the same entities in different expressions into a standard form, and unify the identified relationships into predefined relationship types.

[0013] S1.5: Create a node for each standardized entity in the manufacturing graph, create an edge for each standardized relationship in the manufacturing graph, connect each group of nodes, and check the entities and relationships in the manufacturing graph through manual or automated tools to ensure their correctness and consistency. Based on the verification results, correct the erroneous entities and relationships in the identification process;

[0014] S1.6: Import the constructed manufacturing graph data into the Neo4j database, continuously collect new data sets of corrosion-resistant cables, and dynamically update the manufacturing graph based on the new data, adding new entities and relationships.

[0015] (2) Recommend the optimal combination of corrosion-resistant materials and develop a cable design plan.

[0016] As a further solution of the present invention, the specific steps for optimizing various process parameters in the cable manufacturing process in step (2) are as follows:

[0017] S2.1: Collect all possible combinations of process parameters in the cable manufacturing process, as well as the process parameters and their ranges adjusted in each step, and construct the state space S = {(T, P, V)} and action space A = {ΔT, ΔP, ΔV} based on the collected information, where T represents temperature, P represents pressure, V represents stretching speed, ΔT represents temperature adjustment, ΔP represents pressure adjustment, and ΔV represents stretching speed adjustment;

[0018] S2.2: Create a root node N0 according to the cable design scheme to represent the initial process parameter combination, and initialize the number of visits and reward value of the root node. Then, generate multiple groups of child nodes according to the parameter combination in the state space, and connect the root node with the child nodes using the corresponding adjustment information in the selected action space as edges to complete the construction of the simulation tree;

[0019] S2.3: Starting from the root node N0, select the node with the highest UCB value through the upper confidence bound selection strategy, repeatedly perform calculations and selections until a leaf node is reached for expansion, perform any action on the selected child node, generate a new state, and add the new state as a new node to the tree;

[0020] S2.4: Starting from the new node, randomly select actions to simulate the process of adjusting process parameters until the preset simulation time is reached, and then calculate the reward value of the simulation result, propagate the reward value of the simulation result back to all nodes on the path, and update the number of visits and reward value of each node;

[0021] S2.5: Repeat the iterative search, step-by-step expansion, simulation, and backtracking process until the set number of iterations or computing resource limit is reached. After the simulation is completed, the process parameter combination with the most visits or the highest reward value is selected as the optimal combination;

[0022] S2.6: Apply the selected optimal process parameter combination to the actual cable manufacturing process, evaluate its quality indicators and manufacturing costs, verify its optimization effect, and continuously adjust and optimize the process parameters based on the verification results and real-time production data.

[0023] As a further solution of the present invention, the specific calculation formula of the upper confidence bound selection strategy described in S2.3 is as follows:

[0024]

[0025] Where N represents the node selected in the next step; argmax i () represents selecting the child node with the highest UCB value among all child nodes; N i Represents a child node; W(N i ) represents the sum of the reward values ​​of the child nodes; n(N i ) represents the number of visits to the child node; c represents the exploration parameter; n(N p ) represents the parent node N p The number of visits.

[0026] (3) Optimize various process parameters in the cable manufacturing process;

[0027] (4) Simulate the optimized process and output a simulation report;

[0028] (5) Analyze the cable quality and cost of different production plans and select the optimal plan.

[0029] As a further solution of the present invention, the specific steps of selecting the optimal plan in step (4) are as follows:

[0030] S3.1: Collect various groups of data in the cable production process, and use the optional production plan and process parameters as decision nodes, the various groups of random factors in the production process as random nodes, and the final quality indicators and manufacturing costs as utility nodes. Determine the dependencies between nodes based on the constructed manufacturing graph, and use the dependencies as edges to connect the groups of nodes.

[0031] S3.2: Use statistical methods or machine learning models to quantify the causal relationship between nodes based on historical data, and define the probability distribution of random nodes through normal distribution, initialize the number of exploration bodies, set the pheromone concentration on each path to the same value, and initialize heuristic information based on the initial attractiveness of each path;

[0032] S3.3: The selection probability of each path in the influence diagram is calculated by pheromone concentration and heuristic information. Then the explorer selects the next node according to the selection probability, and builds a specific production plan based on the selected path, including the selected process parameters and decision steps, and evaluates the quality indicators and manufacturing costs of the selected path;

[0033] S3.4: The overall quality of the path is evaluated through a comprehensive utility function, and the pheromone concentration on the selected path is updated after each path selection. After all explorers complete the path selection, the pheromone concentration is updated centrally;

[0034] S3.5: Iteratively execute path construction, utility calculation and pheromone update, and gradually optimize the production plan until the change value of the utility value of the optimal path for multiple consecutive times reaches the preset convergence threshold. Select the path with the highest utility value from all iterative results as the optimal production plan, and apply the selected optimal production plan to the actual cable manufacturing process.

[0035] As a further solution of the present invention, the specific calculation formula of the selection probability in S3.3 is as follows:

[0036]

[0037] In the formula, p hj represents the selection probability of path h to j; τ hj represents the pheromone concentration of path h to j; η hj represents the heuristic information of the path from h to j; α represents the importance factor of pheromone; β represents the importance factor of heuristic information; k represents a node in the set of all optional neighbor nodes of node h; Mh Represents the set of all optional neighbor nodes of node h; τ hk represents the path from h to j; η hk represents the path from h to j;

[0038] The specific calculation formula of the comprehensive utility function described in S3.4 is as follows:

[0039] U=w1Q-w2C

[0040] Where U represents the utility value; Q represents the cable quality index; C represents the manufacturing cost; w1 and w2 represent the weights of quality and cost respectively.

[0041] (6) Conduct quality inspection on the produced cables and store the inspection results;

[0042] (7) Analyze market demand and competitor situation to adjust production strategy and product positioning.

[0043] As a further solution of the present invention, the main participants in the market, including the enterprise itself and the main competitors, and the strategy set that each participant can choose from are determined, and the profit function of each participant under different strategy combinations is calculated based on the market demand function and the cost structure. Then, the payment matrix of the enterprise and the competitors is constructed based on the obtained groups of data to represent the profits under different strategy combinations. By analyzing the payment matrix, the strategy combination that makes it impossible for each participant to increase the profit by changing his own strategy when the opponent's strategy remains unchanged, that is, the Nash equilibrium point, is found. A sensitivity analysis of the Nash equilibrium is performed, the parameters of the market demand function and the cost function are changed, the changes in the Nash equilibrium point are observed, and the impact of different market scenarios on the equilibrium strategy is analyzed. Based on the found Nash equilibrium point, the production plan and market strategy of the enterprise are formulated, and the reactions of competitors are considered to formulate countermeasures in advance.

[0044] (8) Collect customer feedback on corrosion-resistant cables and improve the design and manufacturing process.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The intelligent corrosion-resistant cable manufacturing method collects all possible combinations of process parameters in the cable manufacturing process, as well as the process parameters and their ranges adjusted in each step, constructs the state space and action space based on the collected information, creates a root node based on the formulated cable design plan to represent the initial process parameter combination, and initializes the number of visits and reward values ​​of the root node. Then, based on the parameter combination in the state space, multiple groups of child nodes are generated. Starting from the root node, the node with the highest UCB value is selected through the upper confidence bound selection strategy, and the calculation and selection are repeated until the leaf node is reached to expand and generate a new state, and the new state is added to the tree as a new node. Starting from the new node, actions are randomly selected to simulate the process of process parameter adjustment, and the reward value of the simulation result is back-propagated to the path All nodes on the cable are searched, the number of visits and reward values ​​of each node are updated, and the iterative search, step-by-step expansion, simulation and backtracking process are repeated until the set number of iterations or computing resource limits are reached. After the simulation, the process parameter combination with the largest number of visits or the highest reward value is selected as the optimal combination. The selected optimal process parameter combination is applied to the actual cable manufacturing process, and its quality indicators and manufacturing costs are evaluated to verify its optimization effect. According to the verification results and real-time production data, the process parameters are continuously optimized and adjusted, which can more efficiently explore and optimize key process parameters, reduce manual intervention and experimental costs in traditional trial and error methods, reduce production costs, and reduce the probability of problems in the production process, thereby improving the stability of the manufacturing process and the overall quality of the cable.

[0047] 2. The intelligent corrosion-resistant cable production method collects each group of data in the cable production process, and determines the decision nodes, random nodes and utility nodes, and determines the dependencies between the nodes based on the constructed manufacturing map to connect each group of nodes, initialize the number of explorers, and set the pheromone concentration on each path to the same value. The heuristic information is initialized according to the initial attractiveness of each path, and the selection probability of each path in the influence graph is calculated by the pheromone concentration and the heuristic information. After that, the explorer selects the next node according to the selection probability, and constructs a specific production plan based on the selected path, including the selected process parameters and decision steps, evaluates the quality indicators and manufacturing costs of the selected path, and evaluates it through a comprehensive utility function. The overall quality of the path is evaluated, and the pheromone concentration on the selected path is updated after each path selection. After all explorers complete path selection, the pheromone concentration is updated centrally, and path construction, utility calculation and pheromone update are iteratively performed to gradually optimize the production plan until the utility value change of the optimal path for multiple consecutive times reaches the preset convergence threshold. The path with the highest utility value is selected from all iterative results as the optimal production plan, and the selected optimal production plan is applied to the actual cable manufacturing process. This can more comprehensively evaluate the comprehensive effects of different production plans, ensure that the production plan can respond to changes quickly, maintain the efficiency and competitiveness of production, help to better understand and explain the optimization results, and promote the transparency of the decision-making process. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0049] Figure 1 This is a flowchart of an intelligent corrosion-resistant cable manufacturing method proposed in the present invention. DETAILED DESCRIPTION

[0050] Example 1

[0051] Reference Figure 1 , an intelligent corrosion-resistant cable manufacturing method, the specific steps of the manufacturing method are as follows:

[0052] Collect and integrate information on various corrosion-resistant cable materials.

[0053] Specifically, through web crawlers, API interfaces and database export methods, the information of various corrosion-resistant cable materials is collected from academic papers, technical documents, patent databases, manufacturer information and environmental data sources, redundant, irrelevant and erroneous data are eliminated, the data formats of different sources are unified, and missing data are processed by interpolation, mean filling or deletion. The processed data is converted into text format, and the collected text data is segmented into words or phrases using NLTK or spaCy natural language processing tools. Stop words are removed from the word segmentation results, and the words are restored to their basic form or roots. The part of speech of each word is marked by NLP technology, and the part of speech of each word is identified. Then, the pre-trained NER model is used to identify specific types of entities. NLP analyzes the sentence structure and determines the dependency between words. Then, based on the dependency and context information, it identifies the relationship patterns between entities, unifies the same entities in different expressions into a standard form, and unifies the identified relationships into predefined relationship types. In the manufacturing graph, a node is created for each standardized entity. In the manufacturing graph, an edge is created for each standardized relationship. Each group of nodes is connected. The entities and relationships in the manufacturing graph are checked by manual or automated tools to ensure their correctness and consistency. Based on the verification results, the erroneous entities and relationships in the identification process are corrected. The constructed manufacturing graph data is imported into the Neo4j database. New groups of data on corrosion-resistant cables are continuously collected, and the manufacturing graph is dynamically updated based on the new data to add new entities and relationships.

[0054] Recommend the best combination of corrosion-resistant materials and develop cable designs.

[0055] Specifically, all possible combinations of process parameters in the cable manufacturing process, as well as the process parameters and their ranges adjusted in each step, are collected. Based on the collected groups of information, the state space S = {(T, P, V)} and the action space A = {ΔT, ΔP, ΔV} are constructed, where T represents temperature, P represents pressure, V represents stretching speed, ΔT represents temperature adjustment, ΔP represents pressure adjustment, and ΔV represents stretching speed adjustment. A root node N0 is created based on the formulated cable design scheme to represent the initial process parameter combination, and the number of visits and reward value of the root node are initialized. Then, based on the parameter combination in the state space, multiple groups of child nodes are generated, and the corresponding adjustment information in the selected action space is used as an edge to connect the root node with the child node to complete the construction of the simulation tree. Starting from the root node N0, the node with the highest UCB value is selected through the upper confidence bound selection strategy, and the calculation selection is repeated. Select until it reaches a leaf node for expansion, perform any action on the selected child node, generate a new state, and add the new state as a new node to the tree, starting from the new node, randomly select actions, simulate the process of process parameter adjustment, stop after reaching the preset simulation time, and calculate the reward value of the simulation result, backpropagate the reward value of the simulation result to all nodes on the path, and update the number of visits and reward value of each node, repeat the iterative search, step-by-step expansion, simulation and backtracking process until the set number of iterations or computing resource limits are reached. After the simulation, select the process parameter combination with the most visits or the highest reward value as the optimal combination, apply the selected optimal process parameter combination to the actual cable manufacturing process, evaluate its quality indicators and manufacturing costs, verify its optimization effect, and continuously optimize and adjust the process parameters according to the verification results and real-time production data.

[0056] In this embodiment, the specific calculation formula of the upper confidence bound selection strategy is as follows:

[0057]

[0058] Where N represents the node selected in the next step; argmax i () represents selecting the child node with the highest UCB value among all child nodes; N i Represents a child node; W(N i ) represents the sum of the reward values ​​of the child nodes; n(N i ) represents the number of visits to the child node; c represents the exploration parameter; n(N p ) represents the parent node N p The number of visits.

[0059] Optimize various process parameters in the cable manufacturing process.

[0060] Simulate the optimized process and output a simulation report.

[0061] Example 2

[0062] Reference Figure 1 , an intelligent corrosion-resistant cable manufacturing method, the specific steps of the manufacturing method are as follows:

[0063] Analyze the cable quality and cost of different production plans and select the best plan.

[0064] Specifically, each group of data in the cable production process is collected, and the optional production plan and process parameters are used as decision nodes, each group of random factors in the production process is used as a random node, and the final quality index and manufacturing cost are used as a utility node. The dependency relationship between nodes is determined based on the constructed manufacturing graph, and the dependency relationship is used as an edge to connect each group of nodes. Statistical methods or machine learning models are used to quantify the causal relationship between nodes based on historical data, and the probability distribution of random nodes is defined through normal distribution. The number of exploration bodies is initialized, and the pheromone concentration on each path is set to the same value. The heuristic information is initialized according to the initial attractiveness of each path, and the selection probability of each path in the influence graph is calculated through the pheromone concentration and heuristic information. After that, the explorer selects the next node according to the selection probability, and builds a specific production plan according to the selected path, including the selected process parameters and decision steps, evaluates the quality indicators and manufacturing costs of the selected path, and evaluates the overall quality of the path through a comprehensive utility function. After each path selection, the pheromone concentration on the selected path is updated. After all explorers complete the path selection, the pheromone concentration is updated centrally, and the path construction, utility calculation and pheromone update are iteratively executed to gradually optimize the production plan until the utility value change value of the optimal path for multiple consecutive times reaches the preset convergence threshold. The path with the highest utility value is selected from all iterative results as the optimal production plan, and the selected optimal production plan is applied to the actual cable manufacturing process.

[0065] As a further solution of the present invention, the specific calculation formula of the selection probability is as follows:

[0066]

[0067] In the formula, p hj represents the selection probability of path h to j; τ hj represents the pheromone concentration of path h to j; η hj represents the heuristic information of the path from h to j; α represents the importance factor of pheromone; β represents the importance factor of heuristic information; k represents a node in the set of all optional neighbor nodes of node h; M h Represents the set of all optional neighbor nodes of node h; τ hk represents the path from h to j; η hk represents the path from h to j;

[0068] The specific calculation formula of the comprehensive utility function is as follows:

[0069] U=w1Q-w2C

[0070] Where U represents the utility value; Q represents the cable quality index; C represents the manufacturing cost; w1 and w2 represent the weights of quality and cost respectively.

[0071] Perform quality inspection on produced ropes and store the inspection results.

[0072] Analyze market demand and competitor situation to adjust production strategy and product positioning.

[0073] As a further solution of the present invention, the main participants in the market, including the enterprise itself and the main competitors, and the strategy set that each participant can choose from are determined, and the profit function of each participant under different strategy combinations is calculated based on the market demand function and the cost structure. Then, the payment matrix of the enterprise and the competitors is constructed based on the obtained groups of data to represent the profits under different strategy combinations. By analyzing the payment matrix, the strategy combination that makes it impossible for each participant to increase the profit by changing his own strategy when the opponent's strategy remains unchanged, that is, the Nash equilibrium point, is found. A sensitivity analysis of the Nash equilibrium is performed, the parameters of the market demand function and the cost function are changed, the changes in the Nash equilibrium point are observed, and the impact of different market scenarios on the equilibrium strategy is analyzed. Based on the found Nash equilibrium point, the production plan and market strategy of the enterprise are formulated, and the reactions of competitors are considered to formulate countermeasures in advance.

[0074] (8) Collect customer feedback on corrosion-resistant cables and improve the design and manufacturing process.

Claims

1. An intelligent method for manufacturing corrosion-resistant cables, characterized in that: The specific steps of the production method are as follows: (1) Collect and integrate information on various corrosion-resistant cable materials; (2) Recommend the best combination of corrosion-resistant materials and develop a cable design plan; (3) Optimize various process parameters in the cable manufacturing process; (4) Simulate the optimized process and output a simulation report; (5) Analyze the cable quality and cost of different production plans and select the optimal plan; (6) Conduct quality inspection on the produced cables and store the inspection results; (7) Analyze market demand and competitor situation to adjust production strategy and product positioning; (8) Collect customer feedback on corrosion-resistant cables and improve the design and manufacturing process.

2. The method for making an intelligent corrosion-resistant cable according to claim 1, characterized in that: The specific steps for collecting and integrating the corrosion-resistant cable material information in step (1) are as follows: S1.1: Collect information on various corrosion-resistant cable materials from various data sources including academic papers, technical documents, patent databases, manufacturer data, and environmental data through web crawlers, API interfaces, and database exports; S1.2: Eliminate redundant, irrelevant, and erroneous data, unify the data formats from different sources, use interpolation, mean filling, or deletion to handle missing data, and convert the processed data into text format; S1.3: Use NLTK or spaCy natural language processing tools to segment the collected text data into words or phrases, remove stop words from the word segmentation results, restore the words to their basic form or roots, use NLP technology to label each word with its part of speech, identify the part of speech of each word, and then use the pre-trained NER model to identify specific types of entities; S1.4: Analyze sentence structure through Stanford NLP to determine the dependency relationship between words. Then, based on the dependency relationship and context information, identify the relationship pattern between entities, unify the same entities in different expressions into a standard form, and unify the identified relationships into predefined relationship types. S1.5: Create a node for each standardized entity in the manufacturing graph, create an edge for each standardized relationship in the manufacturing graph, connect each group of nodes, and check the entities and relationships in the manufacturing graph through manual or automated tools to ensure their correctness and consistency. Based on the verification results, correct the erroneous entities and relationships in the identification process; S1.6: Import the constructed manufacturing graph data into the Neo4j database, continuously collect new data sets of corrosion-resistant cables, and dynamically update the manufacturing graph based on the new data, adding new entities and relationships.

3. The method for making an intelligent corrosion-resistant cable according to claim 1, characterized in that: The specific steps for optimizing various process parameters in the cable manufacturing process described in step (2) are as follows: S2.1: Collect all possible combinations of process parameters in the cable manufacturing process, as well as the process parameters and their ranges adjusted in each step, and construct the state space S = {(T, P, V)} and action space A = {ΔT, ΔP, ΔV} based on the collected information, where T represents temperature, P represents pressure, V represents stretching speed, ΔT represents temperature adjustment, ΔP represents pressure adjustment, and ΔV represents stretching speed adjustment; S2.2: Create a root node N0 according to the cable design scheme to represent the initial process parameter combination, and initialize the number of visits and reward value of the root node. Then, generate multiple groups of child nodes according to the parameter combination in the state space, and connect the root node with the child nodes using the corresponding adjustment information in the selected action space as edges to complete the construction of the simulation tree; S2.3: Starting from the root node N0, select the node with the highest UCB value through the upper confidence bound selection strategy, repeatedly perform calculations and selections until a leaf node is reached for expansion, perform any action on the selected child node, generate a new state, and add the new state as a new node to the tree; S2.4: Starting from the new node, randomly select actions to simulate the process of adjusting process parameters until the preset simulation time is reached, and then calculate the reward value of the simulation result, propagate the reward value of the simulation result back to all nodes on the path, and update the number of visits and reward value of each node; S2.5: Repeat the iterative search, step-by-step expansion, simulation, and backtracking process until the set number of iterations or computing resource limit is reached. After the simulation is completed, the process parameter combination with the most visits or the highest reward value is selected as the optimal combination; S2.6: Apply the selected optimal process parameter combination to the actual cable manufacturing process, evaluate its quality indicators and manufacturing costs, verify its optimization effect, and continuously adjust and optimize the process parameters based on the verification results and real-time production data.

4. The method for making an intelligent corrosion-resistant cable according to claim 3, characterized in that: The specific calculation formula for the upper confidence bound selection strategy described in S2.3 is as follows: Where N represents the node selected in the next step; argmax i () represents selecting the child node with the highest UCB value among all child nodes; N i Represents a child node; W(N i ) represents the sum of the reward values ​​of the child nodes; n(N i ) represents the number of visits to the child node; c represents the exploration parameter; n(N p ) represents the parent node N p The number of visits.

5. The method for making an intelligent corrosion-resistant cable according to claim 3, characterized in that: The specific steps of selecting the optimal plan in step (4) are as follows: S3.1: Collect various groups of data in the cable production process, and use the optional production plan and process parameters as decision nodes, the various groups of random factors in the production process as random nodes, and the final quality indicators and manufacturing costs as utility nodes. Determine the dependencies between nodes based on the constructed manufacturing graph, and use the dependencies as edges to connect the groups of nodes. S3.2: Use statistical methods or machine learning models to quantify the causal relationship between nodes based on historical data, and define the probability distribution of random nodes through normal distribution, initialize the number of exploration bodies, set the pheromone concentration on each path to the same value, and initialize heuristic information based on the initial attractiveness of each path; S3.3: The selection probability of each path in the influence diagram is calculated by pheromone concentration and heuristic information. Then the explorer selects the next node according to the selection probability, and builds a specific production plan based on the selected path, including the selected process parameters and decision steps, and evaluates the quality indicators and manufacturing costs of the selected path; S3.4: The overall quality of the path is evaluated through a comprehensive utility function, and the pheromone concentration on the selected path is updated after each path selection. After all explorers complete the path selection, the pheromone concentration is updated centrally; S3.5: Iteratively execute path construction, utility calculation and pheromone update, and gradually optimize the production plan until the change value of the utility value of the optimal path for multiple consecutive times reaches the preset convergence threshold. Select the path with the highest utility value from all iterative results as the optimal production plan, and apply the selected optimal production plan to the actual cable manufacturing process.

6. The method for making an intelligent corrosion-resistant cable according to claim 5, characterized in that: The specific calculation formula for the selection probability described in S3.3 is as follows: In the formula, p hj represents the selection probability of path h to j; τ hj represents the pheromone concentration of path h to j; η hj represents the heuristic information of path h to j; α represents the importance factor of pheromone; β represents the importance factor of heuristic information; k represents a node in the set of all optional neighbor nodes of node h; M h Represents the set of all optional neighbor nodes of node h; τ hk represents the path from h to j; η hk represents the path from h to j; The specific calculation formula of the comprehensive utility function described in S3.4 is as follows: U=w1Q-w2C Where U represents the utility value; Q represents the cable quality index; C represents the manufacturing cost; w1 and w2 represent the weights of quality and cost respectively.

Images