Wear-resistant cable processing method

Abstract

The invention discloses a wear-resistant cable processing method, and belongs to the field of cable processing, the processing method comprises the following specific steps: S1, wear-resistant cable material selection and pretreatment; s2, balancing and optimizing each performance index of the cable; s3, performing simulation testing on different knitting schemes; s4, selecting and using a wear-resistant coating material; the wear resistance of the cable can be remarkably improved, the replacement frequency and the maintenance cost are reduced, the diversity and comprehensiveness of weaving process optimization are guaranteed, the number of tests and material waste in actual production are reduced, the research and development cost and time are reduced, the research and development efficiency is improved, resource waste can be reduced, and the overall production efficiency is improved. And efficient operation of the production process is guaranteed, the defective rate and the rework rate are reduced, and the quality and consistency of products are improved.

Description

Technical Field

[0001] The invention relates to the field of cable processing, and in particular to a method for processing a wear-resistant cable. Background Art

[0002] Marine cables play a vital role in the marine environment and are mainly used for ship anchoring, mooring, towing and other related operations. Due to the complex and changeable marine environment, marine cables are often exposed to harsh conditions such as salt water, ultraviolet rays, high temperature, low temperature and mechanical wear, which places high demands on their wear resistance, strength and durability. However, traditional cable materials and processing technologies are difficult to meet these stringent requirements, resulting in frequent replacement of cables in actual use, which not only increases maintenance costs, but may also affect the safety and operating efficiency of ships; therefore, it is particularly important to invent a wear-resistant cable processing method.

[0003] After searching, Chinese patent No. CN202011012260.2 discloses a method for processing wear-resistant cables. Although this invention can enhance the anti-aging ability of the cable, making the cable have a longer service life and be stronger and more durable, it cannot guarantee the diversity and comprehensiveness of the weaving process optimization, and the number of tests and material waste in actual production increase; in addition, the existing wear-resistant cable processing method has low overall production efficiency, slow operation of the production process, high defective rate and rework rate, which reduces the quality and consistency of the product; for this reason, we propose a method for processing wear-resistant cables. Summary of the invention

[0004] The purpose of the present invention is to solve the defects in the prior art and to propose a method for processing a wear-resistant cable.

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

[0006] A method for processing a wear-resistant cable, the specific steps of the processing method are as follows:

[0007] S1. Selection and pretreatment of wear-resistant cable materials: Analyze the molecular structure and physical properties of different materials, select the best wear-resistant material combination, analyze and optimize the microstructure of the materials, and clean, dry and adjust the size of the selected materials;

[0008] S2. Balance and optimize the performance indicators of the cable: Based on the strength, flexibility, wear resistance and cost of the cable, a multi-objective optimization model for the wear-resistant cable is established, and the best parameter combination is found to balance the performance indicators;

[0009] S3. Simulate and test different braiding schemes: optimize the braiding process of the cable, including the braiding method, braiding density and braiding angle, and simulate and test different braiding schemes to evaluate their impact on the wear resistance of the cable and select the best scheme;

[0010] S4. Select and use wear-resistant coating materials: Based on the previous optimization results, select wear-resistant coating materials, dynamically control the coating thickness and uniformity, make the coating evenly distributed on the cable surface, and cure the coating by controlling temperature and time;

[0011] S5. Analyze and optimize each link in the production process: Analyze the resource requirements, time consumption and interdependence of each link in the production process, optimize the resource allocation between each production link, dynamically adjust the production schedule according to the optimization results, and coordinate the production progress of each link;

[0012] S6. Detect cable quality and provide feedback and adjustments: Apply the processed wear-resistant cables to actual ships, conduct offshore tests, and analyze the test data to identify potential quality issues and room for improvement, and adjust the parameters and processes of the aforementioned steps.

[0013] As a further solution of the present invention, the specific steps of selecting the optimal wear-resistant material combination in S1 are as follows:

[0014] Q1.1: Collect data on the molecular structure, physical properties and wear resistance of different materials from public literature, databases and experimental measurement data sources, and then clean and standardize the collected data to remove noise and outliers;

[0015] Q1.2: Represent the molecular structure of each material as a graph, with nodes representing atoms and edges representing chemical bonds. Define attribute vectors for the nodes and edges in the graph based on the collected data. Use each set of defined attribute vectors as the initial feature vector of the node and determine the set of neighbor nodes for each group of nodes.

[0016] Q1.3: The initial feature vectors of each group of nodes are imported into the graph convolutional network as input data. In each layer of GCN, the feature vector of each node is aggregated with the feature vectors of its neighboring nodes, and the aggregated features are nonlinearly transformed through the ReLU activation function. The loss function is calculated based on the predicted output of the model and the true label.

[0017] Q1.4: Calculate the gradient of the loss function relative to the model parameters through the back-propagation algorithm, then use the gradient descent algorithm to update the model parameters, evaluate the prediction performance of the model on the test set, and adjust the model architecture and hyperparameters based on the evaluation results;

[0018] Q1.5: Using the trained graph neural network, the molecular structure of each material is extracted through multi-layer feature aggregation and transformation to generate the material's feature vector. The wear resistance of different material combinations is evaluated based on the material feature vector through the Bayesian optimization algorithm, and the material combination with the best wear resistance is selected as the final wear-resistant cable material.

[0019] As a further solution of the present invention, the specific calculation formula of the feature aggregation described in Q1.3 is as follows:

[0020]

[0021] In the formula, represents the feature vector of node i in the l+1th layer; σ represents the activation function; N(i) represents the set of neighbor nodes of node i; d i represents the number of node connections of node i; d j represents the number of node connections of node j; W (l) represents the weight matrix of the lth layer; h j (l) Represents the feature vector of node j at layer l.

[0022] As a further solution of the present invention, the specific steps of optimizing the braiding process of the cable in S3 are as follows:

[0023] Q2.1: Encode different weaving process schemes into binary or real vectors, where each vector element represents a parameter of the weaving process. Generate a set of initial populations based on the number of different weaving process schemes, perform simulation tests on the weaving schemes corresponding to each individual, and use the scores of the weaving schemes in the wear resistance test as the fitness values ​​of the individuals;

[0024] Q2.2: Select individuals from the current population through a roulette wheel selection strategy to form an intermediate population, select paired parent individuals from the intermediate population, generate offspring individuals through single-point crossover or multi-point crossover, randomly select one or more mutation points in the gene sequence of the individual, perturb the gene value according to the Gaussian distribution, and add the new individuals generated by the crossover and mutation operations to the original population;

[0025] Q2.3: Place detectors at each individual position in the population, and use the individual parameters of each solution as a set of nodes. Then initialize the pheromone matrix and the initial pheromone value on each path. Each detector selects the next node based on the pheromone concentration and heuristic function value on the path to construct the path.

[0026] Q2.4: Each group of detectors continuously selects the next node until the parameters of the entire weaving scheme are selected to construct a complete path. After each path is selected, the pheromones on the path are locally updated. The influence of the weaving scheme corresponding to the constructed path on the wear resistance of the cable is evaluated through simulation or actual testing, and the fitness value is calculated. Then, the pheromones are globally updated according to the fitness value of the path.

[0027] Q2.5: Compare the fitness values ​​of each group of new weaving schemes generated by the detector with the existing population individuals, and generate a new population through the elite strategy, full replacement strategy or partial replacement strategy. Then repeat the iterative update and exploration of the population until the preset maximum number of iterations is reached. In the final population, select the weaving scheme with the highest fitness as the optimal scheme.

[0028] As a further solution of the present invention, the specific formula for local updating of pheromones in Q2.4 is as follows:

[0029] τ ab =(1-ρ)τ ab +Δτ ab

[0030] In the formula, τ ab represents the updated pheromone concentration; ρ represents the pheromone volatility coefficient; Δτ ab Represents pheromone increment.

[0031] As a further solution of the present invention, the specific steps of dynamic adjustment of production scheduling in S5 are as follows:

[0032] Q3.1: Divide the production process into multiple links. The state of each link represents the specific situation of the current link. The actions that will be taken in each state represent the specific operations that can be taken in this link. Then establish a state transition model.

[0033] Q3.2: Treat each link in the production process as a game participant. The participants play the game through resource allocation and interaction. Calculate the expected total benefits that can be obtained by taking actions according to the random production scheduling strategy starting from any state, and update the strategy through iterative calculation to maximize the benefits of action selection in each state.

[0034] Q3.3: Through potential game theory, the production links adjust their respective strategies to maximize the overall potential function of the production link, evaluate the performance of the current strategy, calculate the value function of each state, and then adjust the strategy to obtain higher benefits and update the strategy based on the evaluation of the value function;

[0035] Q3.4: Repeatedly evaluate and improve the strategy until the action selected in each state no longer changes, so as to obtain the optimal strategy, and optimize the resource allocation between each production link based on the optimal strategy.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. The wear-resistant cable processing method encodes different weaving process schemes into binary or real number vectors, generates a group of initial populations according to the number of different weaving process schemes, and selects individuals from the current population through a roulette selection strategy based on the fitness values ​​of the scoring individuals of the weaving schemes in the wear resistance test. New individuals are generated and added to the original population through crossover and mutation operations, and detectors are placed at each individual position in the population. The parameters of each individual scheme are used as a group of nodes, and then the pheromone matrix and the initial pheromone value on each path are initialized. Each detector selects the next node according to the pheromone concentration and the heuristic function value on the path to construct the path. Each group of detectors continuously selects the next node until the parameters of the entire weaving scheme are completed. Selection is performed to construct a complete path, and the influence of the braiding scheme corresponding to the constructed path on the wear resistance of the cable is evaluated through simulation or actual test, and the fitness value is calculated, and the fitness value of each group of new braiding schemes generated by the detector is compared with the fitness value of the existing population individuals, and a new population is generated through an elite strategy, a full replacement strategy or a partial replacement strategy, and then the population is iteratively updated and explored repeatedly until the preset maximum number of iterations is reached. In the final population, the braiding scheme with the highest fitness is selected as the optimal scheme, which can significantly improve the wear resistance of the cable, reduce the replacement frequency and maintenance cost, ensure the diversity and comprehensiveness of the braiding process optimization, reduce the number of trials and material waste in actual production, reduce R&D costs and time, and improve R&D efficiency.

[0038] 2. The wear-resistant cable processing method divides the production process into multiple links. The state of each link represents the specific situation of the current link. The action to be taken in each state represents the specific operation that can be taken in the link. Then, a state transfer model is established, and each link in the production process is used as a game participant. The participants play games through resource allocation and interaction, and the expected total benefit that can be obtained by taking actions starting from any state and following the random production scheduling strategy is calculated. The strategy is updated through iterative calculation to maximize the benefit of action selection in each state. Through potential game theory, the production links adjust their respective strategies to maximize the overall potential function of the production link, evaluate the performance of the current strategy, calculate the value function of each state, and then adjust the strategy according to the evaluation of the value function to obtain higher benefits, update the strategy, and repeatedly perform strategy evaluation and strategy improvement until the action selected in each state no longer changes, so as to obtain the optimal strategy, and optimize the resource allocation between the production links according to the optimal strategy, which can reduce resource waste, improve overall production efficiency, ensure the efficient operation of the production process, reduce the defective rate and rework rate, and improve product quality and consistency. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] 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.

[0040] Figure 1 The present invention is a flowchart of a method for processing wear-resistant cables. DETAILED DESCRIPTION

[0041] Example 1

[0042] Reference Figure 1 , a wear-resistant cable processing method, the specific steps of the processing method are as follows:

[0043] Wear-resistant cable material selection and pretreatment: Analyze the molecular structure and physical properties of different materials, select the best wear-resistant material combination, analyze and optimize the material microstructure, and clean, dry and size the selected materials.

[0044] Specifically, data on the molecular structure, physical properties and wear resistance of different materials are collected from public literature, databases and experimental measurement data sources. The collected data are then cleaned and standardized to remove noise and outliers. The molecular structure of each material is represented as a graph, with nodes representing atoms and edges representing chemical bonds. Attribute vectors are defined for the nodes and edges in the graph based on the collected data. At the same time, each group of defined attribute vectors is used as the initial feature vector of the node, and the set of neighbor nodes for each group of nodes is determined. The initial feature vectors of each group of nodes are imported as input data into the graph convolutional network. In each layer of GCN, the feature vector of each node is feature aggregated with the feature vectors of its neighbor nodes. , and perform nonlinear transformation on the aggregated features through the ReLU activation function. According to the predicted output of the model and the true label, calculate the loss function, calculate the gradient of the loss function relative to the model parameters through the back propagation algorithm, and then use the gradient descent algorithm to update the model parameters. Evaluate the prediction performance of the model on the test set, and adjust the model architecture and hyperparameters according to the evaluation results. Use the trained graph neural network to extract the features of the molecular structure of each material through multi-layer feature aggregation and transformation, and generate the feature vector of the material. Through the Bayesian optimization algorithm, based on the material feature vector, evaluate the wear resistance of different material combinations, and select the material combination with the best wear resistance as the final wear-resistant cable material.

[0045] It should be further explained that the specific calculation formula for feature aggregation is as follows:

[0046]

[0047] In the formula, represents the feature vector of node i in the l+1th layer; σ represents the activation function; N(i) represents the set of neighbor nodes of node i; d i represents the number of node connections of node i; d j represents the number of node connections of node j; W (l) Represents the weight matrix of the lth layer; Represents the feature vector of node j at layer l.

[0048] Balance and optimize the performance indicators of the cable: Based on various factors such as the strength, flexibility, wear resistance and cost of the cable, a multi-objective optimization model for wear-resistant cables is established, and the best parameter combination is found to balance various performance indicators.

[0049] In addition, it should be noted that the specific process of balancing and optimizing the performance indicators of the cable is as follows: randomly select several initial sample points from the material feature space, evaluate their wear resistance, obtain the initial data set, construct a Gaussian process model, use the initial data set to train the Gaussian process model, use the expected improvement as the acquisition function, and then maximize the acquisition function in the feature space, select new sampling points, evaluate the wear resistance of the new sampling points to obtain new data points, and then use the updated data set to retrain the Gaussian process model, update the distribution of the proxy objective function, repeat the iterative update until the maximum number of iterations is reached, and select the material combination with the highest wear resistance among all sampling points to balance the performance indicators.

[0050] Simulate and test different braiding schemes: Optimize the braiding process of the cable, including the braiding method, braiding density and braiding angle, and simulate and test different braiding schemes to evaluate their impact on the wear resistance of the cable and select the best scheme.

[0051] Specifically, different weaving process schemes are encoded into binary or real number vectors, where each vector element represents a parameter of the weaving process. According to the number of different weaving process schemes, a set of initial populations is generated, and the weaving schemes corresponding to each individual are simulated and tested. The fitness value of the individual scoring the weaving scheme in the wear resistance test is selected from the current population through a roulette selection strategy to form an intermediate population, and paired parent individuals are selected from the intermediate population. Daughter individuals are generated through single-point crossover or multi-point crossover, and one or more mutation points are randomly selected in the gene sequence of the individual, and the gene value is perturbed according to the Gaussian distribution. The new individuals generated by the crossover operation and the mutation operation are added to the original population, and the detection body is placed at the position of each individual in the population. The individual parameters of each scheme are used as a set of nodes, and then the pheromone matrix and each path are initialized. The initial pheromone value on the path is used. Each detector selects the next node according to the pheromone concentration and the heuristic function value on the path to construct the path. Each group of detectors continuously selects the next node until the parameter selection of the entire weaving scheme is completed to construct a complete path. After each path is selected, the pheromones on the path are locally updated. The influence of the weaving scheme corresponding to the constructed path on the wear resistance of the cable is evaluated through simulation or actual test, and the fitness value is calculated. Then, the pheromone is globally updated according to the fitness value of the path. The fitness values ​​of each group of new weaving schemes generated by the detector are compared with those of the individuals in the existing population, and a new population is generated through the elite strategy, full replacement strategy or partial replacement strategy. Then, the iterative update and exploration of the population are repeated until the preset maximum number of iterations is reached and the weaving scheme with the highest fitness is selected as the optimal scheme in the final population.

[0052] In this embodiment, the specific formula for local updating of pheromones is as follows:

[0053] τ ab =(1-ρ)τ ab +Δτ ab

[0054] In the formula, τ ab represents the updated pheromone concentration; ρ represents the pheromone volatility coefficient; Δτ ab Represents pheromone increment.

[0055] Example 2

[0056] Reference Figure 1 , a wear-resistant cable processing method, the specific steps of the processing method are as follows:

[0057] Select and use wear-resistant coating materials: Based on the previous optimization results, select wear-resistant coating materials, dynamically control the coating thickness and uniformity so that the coating is evenly distributed on the cable surface, and cure the coating by controlling temperature and time.

[0058] Analyze and optimize each link in the production process: analyze the resource requirements, time consumption and interdependence of each link in the production process, optimize the resource allocation between each production link, dynamically adjust the production schedule according to the optimization results, and coordinate the production progress of each link.

[0059] Specifically, the production process is divided into multiple links, the state of each link represents the specific situation of the current link, and the actions to be taken in each state represent the specific operations that can be taken in the link. Then a state transfer model is established, and each link in the production process is regarded as a game participant. The participants play games through resource allocation and interaction, and the expected total revenue that can be obtained by taking actions starting from any state and following the random production scheduling strategy is calculated. The strategy is updated through iterative calculation to maximize the revenue of action selection in each state. Through potential game theory, the production links adjust their respective strategies to maximize the overall potential function of the production link, evaluate the performance of the current strategy, calculate the value function of each state, and then adjust the strategy according to the evaluation of the value function to obtain higher revenue, update the strategy, and repeatedly perform strategy evaluation and strategy improvement until the action selected in each state no longer changes, so as to obtain the optimal strategy, and optimize the resource allocation between the production links based on the optimal strategy.

[0060] Detect cable quality and provide feedback for adjustment: Apply the processed wear-resistant cables to actual ships for offshore testing. Analyze the test data to identify potential quality issues and room for improvement, and adjust the parameters and processes of the aforementioned steps.

Claims

1. A method for processing a wear-resistant cable, characterized in that: The specific steps of the processing method are as follows: S1. Selection and pretreatment of wear-resistant cable materials: Analyze the molecular structure and physical properties of different materials, select the best wear-resistant material combination, analyze and optimize the microstructure of the materials, and clean, dry and adjust the size of the selected materials; S2. Balance and optimize the performance indicators of the cable: Based on the strength, flexibility, wear resistance and cost of the cable, a multi-objective optimization model for the wear-resistant cable is established, and the best parameter combination is found to balance the performance indicators; S3. Simulate and test different braiding schemes: optimize the braiding process of the cable, including the braiding method, braiding density and braiding angle, and simulate and test different braiding schemes to evaluate their impact on the wear resistance of the cable and select the best scheme; S4. Select and use wear-resistant coating materials: Based on the previous optimization results, select wear-resistant coating materials, dynamically control the coating thickness and uniformity, make the coating evenly distributed on the cable surface, and cure the coating by controlling temperature and time; S5. Analyze and optimize each link in the production process: Analyze the resource requirements, time consumption and interdependence of each link in the production process, optimize the resource allocation between each production link, dynamically adjust the production schedule according to the optimization results, and coordinate the production progress of each link; S6. Detect cable quality and provide feedback and adjustments: Apply the processed wear-resistant cables to actual ships, conduct offshore tests, and analyze the test data to identify potential quality issues and room for improvement, and adjust the parameters and processes of the aforementioned steps.

2. A method for processing wear-resistant cables according to claim 1, characterized in that: The specific steps for selecting the best wear-resistant material combination described in S1 are as follows: Q1.1: Collect data on the molecular structure, physical properties and wear resistance of different materials from public literature, databases and experimental measurement data sources, and then clean and standardize the collected data to remove noise and outliers; Q1.2: Represent the molecular structure of each material as a graph, with nodes representing atoms and edges representing chemical bonds. Define attribute vectors for the nodes and edges in the graph based on the collected data. Use each set of defined attribute vectors as the initial feature vector of the node and determine the set of neighbor nodes for each group of nodes. Q1.3: The initial feature vectors of each group of nodes are imported into the graph convolutional network as input data. In each layer of GCN, the feature vector of each node is aggregated with the feature vectors of its neighboring nodes, and the aggregated features are nonlinearly transformed through the ReLU activation function. The loss function is calculated based on the predicted output of the model and the true label. Q1.4: Calculate the gradient of the loss function relative to the model parameters through the back-propagation algorithm, then use the gradient descent algorithm to update the model parameters, evaluate the prediction performance of the model on the test set, and adjust the model architecture and hyperparameters based on the evaluation results; Q1.5: Using the trained graph neural network, the molecular structure of each material is extracted through multi-layer feature aggregation and transformation to generate the material's feature vector. The wear resistance of different material combinations is evaluated based on the material feature vector through the Bayesian optimization algorithm, and the material combination with the best wear resistance is selected as the final wear-resistant cable material.

3. A method for processing a wear-resistant cable according to claim 1, characterized in that: The specific calculation formula for feature aggregation described in Q1.3 is as follows: In the formula, represents the feature vector of node i in the l+1th layer; σ represents the activation function; N(i) represents the set of neighbor nodes of node i; d i represents the number of node connections of node i; d j represents the number of node connections of node j; W (l) represents the weight matrix of the lth layer; h j (l) Represents the feature vector of node j at layer l.

4. A method for processing a wear-resistant cable according to claim 2, characterized in that: The specific steps for optimizing the braiding process of the cable described in S3 are as follows: Q2.1: Encode different weaving process schemes into binary or real vectors, where each vector element represents a parameter of the weaving process. Generate a set of initial populations based on the number of different weaving process schemes, perform simulation tests on the weaving schemes corresponding to each individual, and use the scores of the weaving schemes in the wear resistance test as the fitness values ​​of the individuals; Q2.2: Select individuals from the current population through a roulette wheel selection strategy to form an intermediate population, select paired parent individuals from the intermediate population, generate offspring individuals through single-point crossover or multi-point crossover, randomly select one or more mutation points in the gene sequence of the individual, perturb the gene value according to the Gaussian distribution, and add the new individuals generated by the crossover and mutation operations to the original population; Q2.3: Place detectors at each individual position in the population, and use the individual parameters of each solution as a set of nodes. Then initialize the pheromone matrix and the initial pheromone value on each path. Each detector selects the next node based on the pheromone concentration and heuristic function value on the path to construct the path. Q2.4: Each group of detectors continuously selects the next node until the parameters of the entire weaving scheme are selected to construct a complete path. After each path is selected, the pheromones on the path are locally updated. The influence of the weaving scheme corresponding to the constructed path on the wear resistance of the cable is evaluated through simulation or actual testing, and the fitness value is calculated. Then, the pheromones are globally updated according to the fitness value of the path. Q2.5: Compare the fitness values ​​of each group of new weaving schemes generated by the detector with the existing population individuals, and generate a new population through the elite strategy, full replacement strategy or partial replacement strategy. Then repeat the iterative update and exploration of the population until the preset maximum number of iterations is reached. In the final population, select the weaving scheme with the highest fitness as the optimal scheme.

5. A method for processing a wear-resistant cable according to claim 4, characterized in that: The specific formula for local pheromone update described in Q2.4 is as follows: t ab =(1-ρ)τ ab +Δt ab In the formula, τ ab represents the updated pheromone concentration; ρ represents the pheromone volatility coefficient; Δτ ab Represents pheromone increment.

6. A method for processing a wear-resistant cable according to claim 4, characterized in that: The specific steps of dynamic adjustment of production scheduling described in S5 are as follows: Q3.1: Divide the production process into multiple links. The state of each link represents the specific situation of the current link. The actions that will be taken in each state represent the specific operations that can be taken in this link. Then establish a state transition model. Q3.2: Treat each link in the production process as a game participant. The participants play the game through resource allocation and interaction. Calculate the expected total benefits that can be obtained by taking actions according to the random production scheduling strategy starting from any state, and update the strategy through iterative calculation to maximize the benefits of action selection in each state. Q3.3: Through potential game theory, the production links adjust their respective strategies to maximize the overall potential function of the production link, evaluate the performance of the current strategy, calculate the value function of each state, and then adjust the strategy to obtain higher benefits and update the strategy based on the evaluation of the value function; Q3.4: Repeatedly evaluate and improve the strategy until the action selected in each state no longer changes, so as to obtain the optimal strategy, and optimize the resource allocation between each production link based on the optimal strategy.

Images