Method for evaluating environmental damage of group components in south china sea marine atmospheric environment-working condition coupling environment

Abstract

The application relates to a kind of group component south sea marine atmospheric environment-working condition coupling environment damage evaluation methods. Including obtaining multiple groups of corrosion sample data, corresponding actual corrosion result;According to the correlation, the corrosion influencing factor with the largest correlation degree is taken as the current corrosion influencing factor, and the current corrosion identification condition is determined;If the branch attribute corresponding to the current corrosion identification condition is the first branch attribute, the corrosion influencing factor with the largest correlation degree is taken from the corrosion influencing factor except the current corrosion identification condition as the new current corrosion influencing factor, return to the step of determining the current corrosion identification condition according to the current corrosion influencing factor until the branch attribute is the second branch attribute;Based on the corrosion identification result of the current corrosion identification condition and the second branch attribute, the target ship component corrosion identification model is obtained, and the corrosion evaluation data is evaluated by using the target ship component corrosion identification model. The corrosion identification result can be accurately obtained by using the method.

Description

Technical Field

[0001] This application relates to the field of machine learning technology, and in particular to a method for evaluating the environmental damage of components in the South China Sea marine atmospheric environment-operating conditions coupled environment. Background Technology

[0002] Conducting correlation studies on damage in natural-laboratory environments is a significant step in advancing the technology of coupled training between the marine atmospheric environment and operating conditions in the South China Sea for components, and a crucial step in evaluating the adaptability of components to the South China Sea marine environment. Coupled training of the South China Sea atmospheric laboratory environment and operating conditions can acquire corrosion information in a short time. However, the service environment of equipment products is highly variable, and there are differences between the laboratory simulation environment and the actual service environment. Due to significant differences in atmospheric composition, content, humidity, and pollution levels across regions, the mechanisms of single-factor and multi-factor combined effects are not yet fully understood. The challenge of correlation research lies in establishing a correlation analysis method between the South China Sea atmospheric laboratory environment-operating condition coupled training data and the training data under the natural marine atmospheric environment, and accurately converting degradation and corrosion information.

[0003] In the existing technology, there is a lack of evaluation methods for the environmental damage of components in the South China Sea marine atmosphere-operating conditions coupled environment. Summary of the Invention

[0004] Therefore, it is necessary to provide a method for evaluating component damage in the South China Sea marine atmospheric environment-operating condition coupled environment, which can effectively assess corrosion identification of components in the coupled marine atmospheric environment-operating condition environment.

[0005] Firstly, this application provides a method for evaluating the damage of components in the South China Sea marine atmospheric environment coupled with operating conditions, the method comprising:

[0006] Multiple sets of corrosion sample data for sample ship components were obtained, and the actual corrosion results corresponding to each corrosion sample data were obtained; among them, each corrosion sample data corresponds to different corrosion influencing factors;

[0007] Based on the correlation between each corrosion influencing factor and the actual corrosion results, the corrosion influencing factor with the highest correlation is identified as the current corrosion influencing factor, and the current corrosion identification conditions are determined based on the current corrosion influencing factor.

[0008] If the branch attribute of the condition branch corresponding to the current corrosion identification condition is the first branch attribute, which cannot directly identify the corrosion result of the ship component, then based on the degree of relevance, from the corrosion influencing factors other than the current corrosion identification condition, the corrosion influencing factor with the highest degree of relevance is taken as the new current corrosion influencing factor, and the step of determining the current corrosion identification condition based on the current corrosion influencing factor is returned to be executed until the branch attribute of the condition branch corresponding to the current corrosion identification condition is the second branch attribute, which can directly identify the corrosion result of the ship component.

[0009] Obtain the corrosion identification results corresponding to each second branch attribute, and based on each current corrosion identification condition and the corrosion identification results corresponding to each second branch attribute, obtain the corrosion identification model of the target ship component. Use the corrosion identification model of the target ship component to evaluate the corrosion evaluation data of the pre-obtained ship component.

[0010] In one embodiment, the actual corrosion result includes a first corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as uncorroded.

[0011] Before determining the corrosion influencing factor with the highest correlation as the current corrosion influencing factor, based on the correlation between each corrosion influencing factor and the actual corrosion results, the following should also be considered:

[0012] Obtain the target corrosion influencing factors, as well as multiple influencing factor classification intervals pre-defined for the target corrosion influencing factors; the target corrosion influencing factor is any one of the various corrosion influencing factors;

[0013] Based on the number of first corrosion sample data contained in each influencing factor classification interval and the number of second corrosion sample data contained in each influencing factor classification interval, the sub-correlation degree between each influencing factor classification interval and the actual corrosion result is obtained; the first corrosion sample data is the corrosion sample data in each influencing factor classification interval that corresponds to the first corrosion result; the second corrosion sample data is the corrosion sample data in each influencing factor classification interval that corresponds to the second corrosion result.

[0014] Based on the degree of correlation of each sub-correlation, the correlation between the target corrosion influencing factors and the actual corrosion results is obtained.

[0015] In one embodiment, determining the current corrosion identification conditions based on current corrosion influencing factors includes:

[0016] Obtain multiple classification intervals of influencing factors corresponding to the current corrosion influencing factors;

[0017] Based on the classification intervals of multiple influencing factors, multiple corrosion identification conditions corresponding to the current corrosion influencing factor are obtained; the current corrosion identification condition is any one of the multiple corrosion identification conditions corresponding to the current corrosion influencing factor.

[0018] In one embodiment, each corrosion sample data package contains the influence factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the influencing factor classification interval is characterized by the range of influencing factor values.

[0019] Before determining the sub-correlation degree between each influencing factor classification interval and the actual corrosion results, based on the number of first corrosion sample data included in each influencing factor classification interval and the number of second corrosion sample data included in each influencing factor classification interval, the following steps are also included:

[0020] Obtain the values ​​of the influencing factors contained within each range of influencing factor values;

[0021] Corrosion sample data corresponding to the values ​​of each influencing factor within each influencing factor value range are used as corrosion sample data within each influencing factor classification range.

[0022] In one embodiment, each corrosion sample data package contains the influence factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the actual corrosion results include a first corrosion result characterizing the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result characterizing the corrosion result corresponding to the corrosion sample data as uncorroded;

[0023] After determining the current corrosion identification conditions based on the current corrosion influencing factors, the following are included:

[0024] Obtain the conditional branch corresponding to the current corrosion identification condition;

[0025] Obtain the first number of influencing factor values ​​corresponding to the first corrosion result in the conditional branch, and the second number of influencing factor values ​​corresponding to the second corrosion result in the conditional branch, and obtain the sum of the first and second numbers;

[0026] If the first preset condition is met, the branch attribute of the condition branch is confirmed as the first branch attribute; the first preset condition is that the ratio of the first quantity to the sum of quantities is less than or equal to a preset ratio threshold, and the ratio of the second quantity to the sum of quantities is less than or equal to the ratio threshold.

[0027] If the second preset condition is met, the branch attribute of the condition branch is confirmed as the second branch attribute; the second preset condition is that the ratio of the first quantity to the sum of quantities is greater than the ratio threshold, or the ratio of the second quantity to the sum of quantities is greater than the ratio threshold.

[0028] Obtain the erosion identification results corresponding to each second branch attribute, including:

[0029] If the ratio of the first quantity to the sum of quantities is greater than the ratio threshold, the first corrosion result corresponding to the first quantity is taken as the corrosion identification result.

[0030] If the ratio of the second quantity to the sum of the quantities is greater than the ratio threshold, the second corrosion result corresponding to the second quantity is taken as the corrosion identification result.

[0031] In one embodiment, the actual corrosion result includes a first corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as uncorroded.

[0032] Based on the current corrosion identification conditions and the corrosion identification results corresponding to each second branch attribute, a corrosion identification model for the target ship component is obtained, including:

[0033] Based on the current corrosion identification conditions and the corrosion identification results corresponding to each second branch attribute, an initial corrosion identification model for ship components is generated.

[0034] Obtain multiple sets of corrosion verification data corresponding to the corrosion sample data of the sample ship components, and obtain the actual corrosion results corresponding to each corrosion verification data.

[0035] Based on the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data, the recognition accuracy of the initial ship component corrosion identification model is determined.

[0036] Multiple sets of corrosion verification data are input into the initial ship component corrosion identification model to obtain the corrosion identification results corresponding to each second branch attribute. Based on the number of corrosion identification results corresponding to each second branch attribute, and the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data, the identification accuracy of each second branch attribute is determined.

[0037] The verification results of each second branch attribute are determined by using the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model.

[0038] Based on the verification results of each second branch attribute, the verified second branch attributes are obtained, and based on the current corrosion identification conditions corresponding to the verified second branch attributes and the corrosion identification results corresponding to the verified second branch attributes, the corrosion identification model of the target ship component is obtained.

[0039] In one embodiment, the recognition accuracy of each second branch attribute is determined based on the number of corrosion identification results corresponding to each second branch attribute, and the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data, including:

[0040] Corrosion verification data that meets the current corrosion identification conditions corresponding to the current second branch attribute will be used as the corrosion verification data for the current second branch attribute; the current second branch attribute can be any second branch attribute.

[0041] If the corrosion identification result of the current second branch attribute is the first corrosion result, then the identification accuracy of the current second branch attribute is obtained based on the number of first corrosion results of the current second branch attribute and the number of first corrosion results in the actual corrosion results.

[0042] If the corrosion identification result of the current second branch attribute is the second corrosion result, then the identification accuracy of the current second branch attribute is obtained based on the number of second corrosion results of the current second branch attribute and the number of second corrosion results in the actual corrosion results.

[0043] In one embodiment, the verification result of each second branch attribute is determined using the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model, including:

[0044] If the current corrosion identification condition corresponding to the second branch attribute is the first current corrosion identification condition with the second branch attribute, and the identification accuracy of the second branch attribute is greater than the identification accuracy of the initial ship component corrosion identification model, then the verification result of the second branch attribute is determined to be verified.

[0045] If the current corrosion identification condition corresponding to the second branch attribute is not the first current corrosion identification condition with the second branch attribute, and the identification accuracy corresponding to the second branch attribute is greater than the identification accuracy corresponding to the second branch attribute of the previous current corrosion identification condition with the second branch attribute, then the verification result of the second branch attribute is determined to be verified.

[0046] Secondly, this application also provides a component-operating condition coupled environmental damage assessment device for the South China Sea marine atmospheric environment, which includes:

[0047] The sample data acquisition module is used to acquire multiple sets of corrosion sample data of the sample ship components, and to obtain the actual corrosion results corresponding to each corrosion sample data; among them, each corrosion sample data corresponds to different corrosion influencing factors;

[0048] The current corrosion factor acquisition module is used to determine the corrosion factor with the highest correlation with the actual corrosion results based on the correlation between each corrosion influencing factor and the actual corrosion results, and to determine the current corrosion identification conditions based on the current corrosion influencing factor.

[0049] The branch attribute module is used to, when the branch attribute of the condition branch corresponding to the current corrosion identification condition is the first branch attribute that cannot directly identify the corrosion result of the ship component, select the corrosion influencing factor with the highest correlation from the corrosion influencing factors other than the current corrosion identification condition as the new current corrosion influencing factor, and return to execute the step of determining the current corrosion identification condition based on the current corrosion influencing factor, until the branch attribute of the condition branch corresponding to the current corrosion identification condition is the second branch attribute that can directly identify the corrosion result of the ship component;

[0050] The identification model acquisition module is used to acquire the corrosion identification results corresponding to each second branch attribute, and based on each current corrosion identification condition and the corrosion identification results corresponding to each second branch attribute, obtain the corrosion identification model of the target ship component. The corrosion identification model of the target ship component is used to evaluate the corrosion evaluation data of the pre-acquired ship component.

[0051] Thirdly, this application also provides a computer device. This computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method described above.

[0052] The aforementioned method for evaluating environmental damage in the South China Sea marine atmospheric environment coupled with operational conditions involves acquiring multiple sets of corrosion sample data for the sampled ship components and obtaining the actual corrosion results corresponding to each corrosion sample. Each corrosion sample corresponds to different corrosion influencing factors. Based on the correlation between each corrosion influencing factor and the actual corrosion results, the corrosion influencing factor with the highest correlation is identified as the current corrosion influencing factor. The current corrosion identification condition is then determined based on this current corrosion influencing factor. If the branch attribute of the condition branch corresponding to the current corrosion identification condition is the first branch attribute that cannot directly identify the corrosion results of the ship components, then based on the correlation... From the corrosion influencing factors other than the current corrosion identification conditions, the corrosion influencing factor with the highest correlation is selected as the new current corrosion influencing factor. The process then returns to the step of determining the current corrosion identification conditions based on the current corrosion influencing factor, continuing until the branch attribute of the condition branch corresponding to the current corrosion identification condition is a second branch attribute that can directly identify the corrosion result of the ship component. The corrosion identification results corresponding to each second branch attribute are obtained, and based on each current corrosion identification condition and the corrosion identification results corresponding to each second branch attribute, a corrosion identification model for the target ship component is obtained. This target ship component corrosion identification model is then used to evaluate the pre-acquired corrosion evaluation data of the ship component. This application utilizes multiple sets of corrosion sample data, each with different corresponding corrosion influencing factors, and selects the corrosion influencing factor with the highest correlation as the current corrosion influencing factor. Furthermore, the current corrosion identification conditions can be determined based on the current corrosion influencing factor. The branch attribute of the condition branch of the current corrosion identification condition is used to construct a corrosion identification model for the target ship component. This target ship component corrosion identification model can then be used to identify the corrosion of ship components in a marine atmospheric environment-operating condition coupled environment, effectively and accurately obtaining the corrosion identification results for ship components in such environments. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating the environmental damage assessment method for the combined marine atmospheric environment and operating conditions of components in the South China Sea in one embodiment.

[0054] Figure 2 This is a flowchart illustrating the environmental damage assessment method for the combined marine atmospheric environment and operating conditions of components in another embodiment.

[0055] Figure 3 This is a flowchart illustrating the environmental damage assessment method for the component components in the South China Sea marine atmospheric environment-operating condition coupling in another embodiment.

[0056] Figure 4 This is a schematic diagram of a damage correlation decision tree for the coupled marine atmospheric environment and operating conditions in the South China Sea in one embodiment.

[0057] Figure 5 This is a structural block diagram of a component of a South China Sea marine atmospheric environment-operating condition coupled environmental damage assessment device in one embodiment.

[0058] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0060] Conducting correlation studies on damage in natural-laboratory environments represents a significant advancement in the technology of coupled testing of marine atmospheric environment and operating conditions in the South China Sea for components, and is a crucial step in evaluating the adaptability of components to the South China Sea marine environment. Coupled testing of the South China Sea atmospheric laboratory environment and operating conditions can acquire corrosion information in a short time. However, the service environment of equipment products is highly variable, and there are differences between the laboratory simulated environment and the actual service environment. Due to significant differences in atmospheric composition, content, humidity, and pollution levels across regions, the mechanisms of single-factor and multi-factor combined effects are not yet fully understood. The challenge of correlation research lies in establishing a correlation analysis method between the South China Sea atmospheric laboratory environment-operating condition coupled test data and the test data under natural marine atmospheric conditions, and accurately converting degradation corrosion information.

[0061] Domestic research on damage assessment of the marine atmospheric environment in the South China Sea can currently be divided into the following three parts:

[0062] (1) Research on marine atmospheric corrosion tests and accelerated tests of various structural materials (aluminum alloys, pure zinc, etc.) and research on corrosion life assessment methods.

[0063] When metallic materials are corroded, their weight, thickness, mechanical properties, microstructure, and electrode processes all change. The rate of change of these physical or mechanical properties can be used to represent the degree of metal corrosion and serve as the basis for assessing metal corrosion damage. Various methods can be used to evaluate the degree of corrosion damage; traditional assessment methods include: mass change method, corrosion depth method, corrosion current method, and mechanical property degradation method. For the main structural materials used in equipment, their calendar life usually depends on the maximum corrosion depth. There are two methods to obtain the maximum corrosion depth of a material. One method applies the theoretical viewpoints and analytical methods of damage mechanics, establishing a quantitative damage mechanics model based on in-depth research into the damage mechanism of environmental factors. Some scholars have established quantitative damage mechanics models for typical aluminum alloys under different environmental corrosion and stress coupling. Due to the complexity of corrosion fatigue and its unclear fracture mechanism, some theoretical models of corrosion fatigue crack propagation are difficult to apply in engineering. Therefore, experimental methods are widely used: based on the equivalent relationship between laboratory accelerated corrosion tests and atmospheric exposure tests, laboratory accelerated corrosion tests are conducted using EXCO solution immersion / copper-accelerated acetic acid salt spray tests, followed by statistical processing of the test results. Regression methods are then used to fit the correlation between corrosion time and maximum corrosion depth, serving as the basis for evaluating corrosion damage and determining the calendar life of aircraft structures. In this research, a normal distribution can be used to describe the probability distribution of corrosion damage. Some studies utilize the excellent self-learning and nonlinear mapping capabilities of artificial neural networks to study material-level corrosion damage and predict calendar life. Regarding the environment-condition coupling problem, researchers at Northwestern Polytechnical University conducted salt spray accelerated corrosion tests on two types of high-strength aluminum alloy specimens with and without loading, studying the deteriorating effect of corrosion on material properties and the coupled accelerating effect of load and environment on material damage.

[0064] (2) Research on corrosion intensity assessment of ship structure.

[0065] Ships and marine structures are subjected to the combined effects of atmospheric corrosion and wave loads during their service life. Therefore, the influence of corrosion factors must be considered when assessing the fatigue damage of structures. For the problem of corrosion fatigue strength assessment of ship structures, a new corrosion model that considers mechanical factors based on actual sea corrosion data is often used, combined with spectral analysis and the Miner criterion to assess the fatigue strength of key nodes of the target ship.

[0066] Compared with conventional fatigue life studies, there is currently very little experimental data on fatigue life and corrosion damage of military equipment components under corrosive environments, and the distribution characteristics are still unclear. Further research is needed on their distribution characteristics and methods for estimating distribution parameters, as well as how to determine their optimal distribution based on small samples and improve the accuracy of parameter estimation. In addition, improving the damage detection technology for military equipment components and conducting accurate measurement and reliable assessment of corrosion damage are crucial for the study of the calendar life of military equipment components.

[0067] Currently, my country has obtained a wealth of valuable corrosion data on the main materials used in marine engineering projects in actual sea areas. However, directly applying this data to practical engineering projects still presents significant challenges. This is primarily because actual engineering structures are subject not only to atmospheric corrosion but also to the combined effects of alternating loads. Therefore, it is necessary to establish a corrosion mathematical model that considers mechanical factors.

[0068] (3) Research on corrosion intensity assessment of marine aircraft structures

[0069] Over 90% of the time, marine aircraft are parked. In the marine environment, these aircraft are exposed to high humidity, high temperature, and heavy salt spray for extended periods. Salt particles in the marine atmosphere deposit on structural surfaces, causing hygroscopic deliquescence and increasing the conductivity of the liquid film on the metal surface. Combined with the highly corrosive nature of chloride ions, this accelerates the corrosion of the airframe structure and systems. Common forms of corrosion include: typical corrosion at aircraft structural connections, aging and failure of protective coatings, typical corrosion of aluminum alloy structures, corrosion fatigue and stress corrosion of load-bearing bolts and support columns, and typical corrosion of steel structures. Through analysis of marine atmospheric corrosion environmental factors, a cyclic accelerated testing environment spectrum consisting of two sub-tests—ultraviolet irradiation and periodic immersion—was used to simulate the effects of light, temperature, humidity, and chloride ions in the marine atmospheric environment.

[0070] Corrosion damage to aviation products severely impacts aircraft reliability and safety, making the study of corrosion patterns a hot topic for scholars both domestically and internationally. Environmental damage assessment of military aircraft has been gradually established and improved based on empirical research methods, following thorough studies of environmental damage events (accidents) occurring during equipment use and maintenance. Due to the diversity and complexity of environments, aviation product corrosion exhibits complex variation patterns. Corrosion characterization and equivalent corrosion have become key technologies for studying the corrosion fatigue life and calendar life of aviation products under service conditions. Among these, the corrosion equivalent conversion method serves as a "bridge" between field use environments and laboratory verification tests for aviation products, providing effective data and methods for the environmental adaptability design, improvement, and verification of aviation products.

[0071] The ground parking environment spectrum describes the real-world environmental experiences of an aircraft during its entire calendar life. However, using this spectrum to conduct corrosion tests on aircraft structures in service environments is difficult due to time, cost, and technical limitations. To accurately determine the corrosive effects of real-world environments on aircraft structures, it is necessary to establish an equivalent relationship between the ground parking environment spectrum and the laboratory accelerated environment spectrum. This would allow for the rapid acquisition of corrosion damage patterns of aircraft materials, structures, and protection systems under real ground parking conditions, providing background data and design input for the engineering implementation of aircraft structural calendar life assessment.

[0072] Currently, there is a large body of research on equivalent environmental spectra or accelerated test spectra in existing technologies. For example, methods for compiling equivalent environmental spectra of metal structure corrosion using metal corrosion current as a metric have been developed. Other studies have explored methods for calculating equivalent corrosion damage from commonly used metallic materials, coatings, and critical structures in aerospace products, proposing methods for quantifying equivalent corrosion damage in metallic materials, coatings, composite materials, and structural components.

[0073] With the improvement of research and development capabilities, there is a need to enhance the environmental resistance of aviation products. This requires more reliable and accurate laboratory testing methods for verification, and further acquisition of corrosion patterns to guide environmental adaptability design and improvement. Based on the available data on the marine atmospheric environment spectrum of the South China Sea and the conversion relationship between the natural environment and accelerated environment to standard humid air, an equivalent acceleration relationship between the accelerated test environment spectrum and the typical marine atmospheric environment spectrum is established. This method is simple and feasible.

[0074] Corrosion damage equivalent conversion methods have become a key technology for aircraft calendar age assessment. Existing corrosion damage equivalent conversion methods have the following main shortcomings:

[0075] 1) The equivalent conversion of corrosion damage to metallic materials only considers the influence of environmental factors such as temperature and humidity on the corrosion of metallic materials;

[0076] 2) Insulation resistance alone cannot fully describe the corrosion damage pattern of the protective coating during field use, and the accuracy of the established equivalent relationship needs to be improved;

[0077] 3) There is a lack of research and data on the equivalent conversion of corrosion damage of composite materials. The equivalent conversion formula establishes the relationship between ultraviolet radiation, stress load and material mechanical properties, but does not take into account the influence of humidity environment factors on composite materials.

[0078] 4) The fatigue rating of details can only reflect the fatigue quality of structural details well within the medium life range, and the measured test data is limited. The obtained acceleration equivalent relationship needs to be corrected or improved after more data is added.

[0079] For the reasons mentioned above, this application provides a method for evaluating component damage in the South China Sea marine atmospheric-operating environment coupled environment, which can effectively assess corrosion identification of components in the marine atmospheric-operating environment coupled environment.

[0080] In one embodiment, such as Figure 1As shown, a method for evaluating the environmental damage of components in the South China Sea marine atmospheric environment coupled with operating conditions is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and to a system including both a terminal and a server, and can be implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0081] S102, acquire multiple sets of corrosion sample data for the sample ship components, and obtain the actual corrosion results corresponding to each corrosion sample data; among them, each corrosion sample data corresponds to different corrosion influencing factors.

[0082] Among them, the sample ship components can be typical components in the marine environment, such as rotating components, electronic module components, and optical cable components; typical components can be used as research objects to obtain sample ship components.

[0083] Corrosion sample data can be obtained from test data conducted on sample ship components in a marine environment. It can also be training data derived from test data. For example, a large amount of test data obtained from the marine environment can be randomly divided into a training set and a validation set using multiple test data points and a pre-defined ratio. The training set data can then be used as corrosion sample data. For instance, 2000 test data points can be randomly divided into a training set and a validation set in a 3:1 ratio, with the training set containing 1500 test data points and the validation set containing 500 test data points.

[0084] The actual corrosion results refer to the actual corrosion results obtained by typical components during the test, including both corroded and uncorroded results. Corrosion influencing factors can be the main influencing factors that cause corrosion of components in the coupled marine atmospheric environment and operating conditions, as identified through research.

[0085] For example, users conducting experiments can identify the main influencing factors causing corrosion of components in the coupled marine atmospheric environment and operating conditions of the South China Sea. According to the investigation results, the main influencing factors of the natural environment under the coupled marine atmospheric environment and operating conditions of the South China Sea include: temperature, humidity, vibration, solar radiation, and salt spray. The main influencing factors of the operating conditions of the components include: potential, polarization, and impedance.

[0086] For example, the server can obtain multiple corrosion sample data points input by the user, as well as the actual corrosion results of typical components corresponding to each corrosion sample data point. Each corrosion sample data point corresponds to different corrosion influencing factors. For example, each corrosion sample data point corresponds to corrosion influencing factors such as temperature, humidity, vibration, solar radiation, salt spray, potential, polarization, and impedance.

[0087] S104. Based on the correlation between each corrosion influencing factor and the actual corrosion result, the corrosion influencing factor with the highest correlation is determined as the current corrosion influencing factor, and the current corrosion identification conditions are determined based on the current corrosion influencing factor.

[0088] The degree of relevance can be information entropy. The current corrosion influencing factors can be the corrosion factors determined at the current node.

[0089] For example, for each corrosion influencing factor, the server can calculate the information entropy corresponding to that factor and use this information entropy as the correlation between the factor and the actual corrosion result. Furthermore, the corrosion influencing factor with the highest correlation can be selected as the current corrosion influencing factor; for example, the corrosion influencing factor with the lowest information entropy can be selected as the current corrosion influencing factor. Furthermore, the current corrosion identification conditions can be determined using the current corrosion influencing factor.

[0090] S106, if the branch attribute of the condition branch corresponding to the current corrosion identification condition is the first branch attribute that cannot directly identify the corrosion result of the ship component, then, based on the degree of relevance, select the corrosion influencing factor with the highest degree of relevance from the corrosion influencing factors other than the current corrosion identification condition as the new current corrosion influencing factor, and return to execute the step of determining the current corrosion identification condition based on the current corrosion influencing factor, until the branch attribute of the condition branch corresponding to the current corrosion identification condition is the second branch attribute that can directly identify the corrosion result of the ship component.

[0091] Here, a conditional branch can be any one of the current corrosion identification conditions. Branch attributes refer to the properties of the conditional branch; for example, a conditional branch's attribute might indicate that corrosion results for ship components cannot be directly identified, or it might indicate that corrosion results for ship components can be directly identified. The first branch attribute is the branch attribute for which corrosion results for ship components cannot be directly identified. The second branch attribute is the branch attribute for which corrosion results for ship components can be directly identified.

[0092] For example, if the branch attribute of the condition branch corresponding to the current corrosion identification condition is a first branch attribute that cannot directly identify the corrosion result of the ship component, then under this condition branch, a new current corrosion identification condition is obtained. Specifically, from multiple corrosion sample data belonging to this branch condition, the correlation degree of other current corrosion influencing factors besides the current corrosion influencing factor corresponding to the current corrosion identification condition can be determined, and the corrosion influencing factor with the highest correlation degree is taken as the new current corrosion influencing factor. Further, the step of determining the current corrosion identification condition based on the current corrosion influencing factor can be returned until the branch attribute of the condition branch corresponding to the current corrosion identification condition is a second branch attribute that can directly identify the corrosion result of the ship component. For example, until the branch attribute of the condition branch corresponding to the current corrosion identification condition is a second branch attribute that can directly identify the corrosion result of the ship component, a complete decision tree model can be obtained.

[0093] S108. Obtain the corrosion identification results corresponding to each second branch attribute, and based on each current corrosion identification condition and the corrosion identification results corresponding to each second branch attribute, obtain the corrosion identification model of the target ship component. Use the corrosion identification model of the target ship component to evaluate the corrosion evaluation data of the pre-obtained ship component.

[0094] The corrosion identification model for the target ship component can be a decision tree model. The corrosion identification result refers to the predicted corrosion identification result in the second branch attribute. The corrosion evaluation data can be the data of the ship component to be evaluated.

[0095] For example, a decision tree model can be generated based on various corrosion identification conditions and the corrosion identification results corresponding to the second branch attributes that can directly identify the corrosion results of ship components. This decision tree model can be used to assess whether components are corroded in a coupled marine atmospheric environment and operating conditions. For instance, based on the current corrosion identification conditions and the corrosion identification results of each second branch attribute corresponding to the current corrosion identification conditions, the corrosion identification conditions and the corrosion identification results corresponding to the corrosion identification conditions at each level node in the decision tree model can be constructed. Thus, a complete decision tree model for assessing whether components are corroded in a coupled marine atmospheric environment and operating conditions can be obtained. Furthermore, this decision tree model can be used to evaluate pre-acquired corrosion evaluation data of ship components.

[0096] In this embodiment, by utilizing multiple sets of corrosion sample data, each with different corrosion influencing factors, the corrosion influencing factor with the highest correlation is selected as the current corrosion influencing factor. Furthermore, the current corrosion identification conditions can be determined based on the current corrosion influencing factor. By using the branch attributes of the condition branches of the current corrosion identification conditions, a corrosion identification model for the target ship component can be constructed. Furthermore, the corrosion identification model for the target ship component can be used to identify the corrosion of ship components in the marine atmospheric environment-operating condition coupled environment, which can effectively and accurately obtain the corrosion identification results of ship components in the marine atmospheric environment-operating condition coupled environment.

[0097] In one embodiment, the actual corrosion result includes a first corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as uncorroded.

[0098] like Figure 2 As shown, before determining the corrosion influencing factor with the highest correlation as the current corrosion influencing factor based on the correlation between each corrosion influencing factor and the actual corrosion results, the following steps are also taken:

[0099] S202, obtain the target corrosion influencing factors, as well as multiple influencing factor classification intervals pre-defined for the target corrosion influencing factors; the target corrosion influencing factor is any one of the various corrosion influencing factors.

[0100] The target corrosion influencing factor can be any one of the various corrosion influencing factors; that is, any corrosion influencing factor can be processed using the method described in this embodiment. The influencing factor classification range can be a range for classifying corrosion influencing factors.

[0101] For example, for each corrosion influencing factor, multiple pre-defined influencing factor classification intervals are obtained. For instance, the influencing factor classification intervals can be obtained by classifying the influencing factor values ​​corresponding to the corrosion influencing factor, and the influencing factor values ​​corresponding to the corrosion influencing factor are discretized using these intervals. Multiple features corresponding to the corrosion influencing factor can be obtained based on the influencing factor classification intervals.

[0102] S204. Based on the number of first corrosion sample data contained in each influencing factor classification interval and the number of second corrosion sample data contained in each influencing factor classification interval, the sub-correlation degree between each influencing factor classification interval and the actual corrosion result is obtained.

[0103] The actual corrosion results include a first corrosion result that characterizes the corrosion sample data as corroded, and a second corrosion result that characterizes the corrosion sample data as uncorroded.

[0104] The first corrosion sample data refers to the corrosion sample data within each influencing factor classification interval, where the actual corrosion result is the first corrosion result. The second corrosion sample data refers to the corrosion sample data within each influencing factor classification interval, where the corresponding actual corrosion result is the second corrosion result.

[0105] For example, for each influencing factor classification interval, the information entropy corresponding to the influencing factor classification interval can be calculated based on the number of first corrosion sample data whose actual corrosion result is corrosion and the number of second corrosion sample data whose actual corrosion result is not corrosion. This information entropy is used as the sub-correlation degree between the influencing factor classification interval and the actual corrosion result.

[0106] S206, based on the correlation degree of each sub-correlation, obtain the correlation degree between the target corrosion influencing factors and the actual corrosion results.

[0107] For example, for each influencing factor classification interval, the total information entropy of the corrosion influencing factor can be obtained based on the information entropy corresponding to each influencing factor classification interval. This total information entropy is then used as the correlation between the target corrosion influencing factor and the actual corrosion result.

[0108] In this embodiment, the sub-correlation degree between each influencing factor classification interval and the actual corrosion result is obtained by using the number of first corrosion sample data contained in each influencing factor classification interval and the number of second corrosion sample data contained in each influencing factor classification interval. Furthermore, the correlation degree between the target corrosion influencing factor and the actual corrosion result is obtained through each sub-correlation degree. This can effectively determine the correlation degree of each corrosion influencing factor and further accurately obtain the maximum correlation degree, which is beneficial for constructing a corrosion identification model for target ship components. This enables accurate and effective identification of corrosion of ship components and other components in the marine atmospheric environment-operating condition coupled environment.

[0109] In one embodiment, taking "temperature" as an example, temperature can be divided into "temperature = high temperature", "temperature = normal temperature", and "temperature = low temperature" according to the classification range of influencing factors. Among the 1500 data points in the training set, 367 data points are "temperature = high temperature", 858 data points are "temperature = normal temperature", and 275 experimental data points are "temperature = low temperature".

[0110] The probability of "temperature = high temperature" is R_high temperature = 367 / 1500. Of the 367 data points where "temperature = high temperature", 300 data points have a "test result = corrosion", and 67 data points have a "test result = no corrosion". Therefore, the information entropy under the characteristic condition of "temperature = high temperature" can be calculated using the formula: H_high temperature = -a*log2a - b*log2b, where a = 300 / (300+67) and b = 67 / (300+67).

[0111] Using the same method, we can obtain the probability of "temperature = room temperature" as R_room temperature = 858 / 1500. Among all 858 data points of "temperature = room temperature", 120 data points have "experiment result = corrosion", and 738 data points have "experiment result = no corrosion". Then, according to the formula: H_room temperature = -a*log2a - b*log2b, we can calculate the information entropy under the characteristic condition of "temperature = room temperature", where a = 120 / (120+738) and b = 738 / (120+738).

[0112] Using the same method, we can obtain the probability of "temperature = low temperature" occurring as R_low temperature = 275 / 1500, and the information entropy H_low temperature under the characteristic condition of "temperature = low temperature". Finally, we calculate the total entropy under the temperature characteristic value: H_temperature = R_high temperature * H_high temperature + R_normal temperature * H_normal temperature + R_low temperature * H_low temperature.

[0113] In one embodiment, determining the current corrosion identification conditions based on current corrosion influencing factors includes:

[0114] Obtain multiple classification intervals of influencing factors corresponding to the current corrosion influencing factors;

[0115] Based on the classification intervals of multiple influencing factors, multiple corrosion identification conditions corresponding to the current corrosion influencing factor are obtained; the current corrosion identification condition is any one of the multiple corrosion identification conditions corresponding to the current corrosion influencing factor.

[0116] For example, for each influencing factor classification interval, a corrosion identification condition corresponding to the current corrosion influencing factor can be formed based on the influencing factor classification interval; any one of multiple corrosion identification conditions can be used as the current corrosion identification condition.

[0117] In this embodiment, multiple corrosion identification conditions corresponding to the current corrosion influencing factors can be obtained based on multiple influencing factor classification intervals, which is beneficial to the construction of the decision tree model. At the same time, using corrosion influencing factors to construct the decision tree model can improve the identification of corrosion results of ship components and other components in the coupled marine atmospheric environment-operating condition environment.

[0118] In one embodiment, such as Figure 3As shown, each corrosion sample data package contains the influencing factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the influencing factor classification intervals are characterized by the influencing factor value intervals.

[0119] Before determining the sub-correlation degree between each influencing factor classification interval and the actual corrosion results, based on the number of first corrosion sample data included in each influencing factor classification interval and the number of second corrosion sample data included in each influencing factor classification interval, the following steps are also included:

[0120] S302, obtain the values ​​of the influencing factors included in the value range of each influencing factor;

[0121] S304. The corrosion sample data corresponding to the influencing factor values ​​contained in each influencing factor value interval shall be used as the corrosion sample data contained in each influencing factor classification interval.

[0122] Among them, the classification interval of influencing factors is represented by the range of influencing factor values. For example, the range of influencing factor values ​​can be divided according to the influencing factor values, and the range of influencing factor values ​​can be used to represent the classification interval of influencing factors.

[0123] For example, for each range of influencing factor values, the influencing factor values ​​contained in the range can be obtained, and the corrosion sample data corresponding to the influencing factor values ​​contained in the range can be used as the corrosion sample data contained in the influencing factor classification range corresponding to the range.

[0124] In this embodiment, the corrosion sample data corresponding to the influencing factor values ​​contained in each influencing factor value interval will be used as the corrosion sample data contained in each influencing factor classification interval. This can accurately obtain the number of corrosion sample data contained in each influencing factor classification interval, and further use this number to determine the information entropy, thereby improving the accuracy of the decision tree model.

[0125] In one embodiment, the experimental data undergoes preliminary processing by discretizing the continuous experimental data. Temperature data can be categorized into high temperature (T > 30°C), normal temperature (30°C > T > 5°C), and low temperature (5°C > T). Humidity data can be categorized into high humidity (W > 80% RH), relatively high humidity (80% RH > T > 60% RH), medium humidity (60% RH > T > 40% RH), relatively low humidity (40% RH > T > 20% RH), and low humidity (20% RH > T). Other corrosion-influencing factors are processed similarly.

[0126] For example, taking the test data of rotating parts as an example, the test data is presented in the following table:

[0127]

[0128] The experimental data were discretized to obtain the following table:

[0129]

[0130]

[0131] In one embodiment, each corrosion sample data package contains the influence factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the actual corrosion results include a first corrosion result characterizing the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result characterizing the corrosion result corresponding to the corrosion sample data as uncorroded.

[0132] After determining the current corrosion identification conditions based on the current corrosion influencing factors, the following are included:

[0133] Obtain the conditional branch corresponding to the current corrosion identification condition.

[0134] Obtain the first number of influencing factor values ​​corresponding to the first corrosion result in the conditional branch, and the second number of influencing factor values ​​corresponding to the second corrosion result in the conditional branch, and obtain the sum of the first and second numbers.

[0135] If the first preset condition is met, the branch attribute of the conditional branch is confirmed as the first branch attribute; the first preset condition is that the ratio of the first quantity to the sum of quantities is less than or equal to a preset ratio threshold, and the ratio of the second quantity to the sum of quantities is less than or equal to the ratio threshold.

[0136] If the second preset condition is met, the branch attribute of the condition branch will be confirmed as the second branch attribute; the second preset condition is that the ratio of the first quantity to the sum of quantities is greater than the ratio threshold, or the ratio of the second quantity to the sum of quantities is greater than the ratio threshold.

[0137] For example, the condition branch corresponding to the current corrosion identification condition is determined, and the first number of influencing factor values ​​corresponding to the first corrosion result in the condition branch is obtained, the second number of influencing factor values ​​corresponding to the second corrosion result in the condition branch is obtained, and the first number and the second number are summed to obtain the sum of the numbers.

[0138] If the ratio of the first quantity to the sum of quantities is less than or equal to a preset ratio threshold, and the ratio of the second quantity to the sum of quantities is less than or equal to the ratio threshold, it indicates that the conditional branch cannot directly identify the corrosion result of the ship component. That is, the branch attribute of the conditional branch is confirmed as the first branch attribute.

[0139] If the ratio of the first quantity to the sum of quantities is greater than the ratio threshold, or the ratio of the second quantity to the sum of quantities is greater than the ratio threshold, it indicates that the conditional branch can directly identify the corrosion result of the ship component, that is, the branch attribute of the conditional branch is confirmed as the second branch attribute.

[0140] Obtain the erosion identification results corresponding to each second branch attribute, including:

[0141] If the ratio of the first quantity to the sum of quantities is greater than the ratio threshold, the first corrosion result corresponding to the first quantity is taken as the corrosion identification result.

[0142] If the ratio of the second quantity to the sum of the quantities is greater than the ratio threshold, the second corrosion result corresponding to the second quantity is taken as the corrosion identification result.

[0143] For example, if it is determined that the conditional branch can directly identify the corrosion result of the ship component, that is, if the branch attribute of the conditional branch is confirmed as the second branch attribute, then if the ratio of the first quantity to the sum of quantities is greater than a ratio threshold, it means that the first corrosion result corresponding to the first quantity can be used as the corrosion identification result of the conditional branch. If the ratio of the second quantity to the sum of quantities is greater than a ratio threshold, it means that the second corrosion result corresponding to the second quantity can be used as the corrosion identification result of the conditional branch.

[0144] In this embodiment, the ratio of the first quantity to the sum of quantities, the ratio of the second quantity to the sum of quantities, and the ratio threshold determine the attributes of the conditional branches. This can accurately and effectively obtain the branch attributes of the conditional branches corresponding to the identification conditions, thereby improving the accuracy of establishing a corrosion identification model for target ship components.

[0145] In one embodiment, "vibration" is the corrosion influencing factor with the strongest correlation to whether corrosion occurs. In this case, if the percentage of data where "vibration = high" indicates either "experimental result = corrosion" or "experimental result = no corrosion" exceeds 80% (this 80% threshold can be adjusted for better training results), then "vibration = high" is considered to necessarily result in either "experimental result = corrosion" or "experimental result = no corrosion," and this conditional branch no longer considers other characteristic factors. If the percentage of either "experimental result = corrosion" or "experimental result = no corrosion" does not exceed this threshold, then other corrosion influencing factors such as "temperature," "humidity," and "solar radiation" need to be further selected from the "vibration = high" data to obtain new current corrosion influencing factors for further classification.

[0146] In one embodiment, the actual corrosion result includes a first corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as uncorroded.

[0147] Based on the current corrosion identification conditions and the corrosion identification results corresponding to each second branch attribute, a corrosion identification model for the target ship component is obtained, including:

[0148] Based on the current corrosion identification conditions and the corrosion identification results corresponding to each second branch attribute, an initial corrosion identification model for ship components is generated.

[0149] For example, an initial decision tree model for assessing whether components are corroded in a marine atmospheric environment-operating condition coupled environment can be generated based on various corrosion identification conditions and the corrosion identification results corresponding to the second branch attributes that can directly identify the corrosion results of ship components.

[0150] Multiple sets of corrosion verification data corresponding to the corrosion sample data of the sample ship components were obtained, and the actual corrosion results corresponding to each corrosion verification data were obtained.

[0151] For example, corrosion validation data can be experimental data used to validate the model. For instance, a large amount of experimental data obtained from the marine environment can be randomly divided into training and validation sets by multiple experimental data points and a pre-defined ratio; the validation set data can then be used as corrosion validation data. For example, 2000 experimental data points can be randomly divided into a training set and a validation set in a 3:1 ratio, with the training set containing 1500 experimental data points and the validation set containing 500 experimental data points.

[0152] Based on the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data, the recognition accuracy of the initial ship component corrosion identification model is determined.

[0153] For example, in the training set of 1500 data points, 569 data points were rated as "corroded" and 931 data points were rated as "not corroded". The number of "not corroded" data points is greater than the number of "corroded" data points. Therefore, the initial ship component corrosion identification model considers that "all data points in the validation set are not corroded".

[0154] The initial model, which assumes all data is uncorrupted, was applied to a validation set of 500 data points (300 uncorrupted and 200 corrupted). The accuracy of the initial model was calculated to be 60%.

[0155] Multiple sets of corrosion verification data are input into the initial ship component corrosion identification model to obtain the corrosion identification results corresponding to each second branch attribute. Based on the number of corrosion identification results corresponding to each second branch attribute, and the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data, the identification accuracy of each second branch attribute is determined.

[0156] For example, corrosion identification is performed using the current corrosion identification conditions at each level to obtain the corrosion identification results corresponding to each second branch attribute. For each first branch attribute, the identification accuracy of each second branch attribute is determined based on the number of corrosion identification results corresponding to that second branch attribute, and the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data.

[0157] The verification results of each second branch attribute are determined by using the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model.

[0158] Based on the verification results of each second branch attribute, the verified second branch attributes are obtained, and based on the current corrosion identification conditions corresponding to the verified second branch attributes and the corrosion identification results corresponding to the verified second branch attributes, the corrosion identification model of the target ship component is obtained.

[0159] For example, the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model are used to determine the verification results of each second branch attribute. Based on the verified second branch attributes, the initial decision tree model for evaluating whether the components are corroded in the marine atmospheric environment-operating condition coupled environment is improved to obtain the target decision tree model for evaluating whether the components are corroded in the marine atmospheric environment-operating condition coupled environment.

[0160] In this embodiment, by verifying the initial ship component corrosion identification model, and based on the current corrosion identification conditions corresponding to the second branch attribute after verification, and the corrosion identification results corresponding to the verified second branch attribute, a target ship component corrosion identification model is generated. This improves the accuracy of establishing the target ship component corrosion identification model, and further improves the accuracy of corrosion identification of ship components and other components in the marine atmospheric environment-operating condition coupled environment.

[0161] In one embodiment, the recognition accuracy of each second branch attribute is determined based on the number of corrosion identification results corresponding to each second branch attribute, and the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data, including:

[0162] Corrosion verification data that meets the current corrosion identification conditions corresponding to the current second branch attribute will be used as the corrosion verification data for the current second branch attribute; the current second branch attribute can be any second branch attribute.

[0163] If the corrosion identification result of the current second branch attribute is the first corrosion result, then the identification accuracy of the current second branch attribute is obtained based on the number of first corrosion results of the current second branch attribute and the number of first corrosion results in the actual corrosion results.

[0164] If the corrosion identification result of the current second branch attribute is the second corrosion result, then the identification accuracy of the current second branch attribute is obtained based on the number of second corrosion results of the current second branch attribute and the number of second corrosion results in the actual corrosion results.

[0165] For example, corrosion verification data that meets the current corrosion identification conditions corresponding to the current second branch attribute can be used as the corrosion verification data for the current second branch attribute. For instance, if the current corrosion identification condition is the corrosion identification condition of the first child node, then the corrosion verification data for the current second branch attribute must simultaneously satisfy the corrosion identification conditions of the root node, the first child node, and the second child node.

[0166] If the current corrosion identification result of the second branch attribute is the first corrosion result that has been corroded, then the identification accuracy of the current second branch attribute is obtained based on the number of corroded results obtained from the current identification of the second branch attribute and the number of corroded results in the actual corrosion results.

[0167] If the current corrosion identification result of the second branch attribute is an uncorroded second corrosion result, then the identification accuracy of the current second branch attribute is obtained based on the number of uncorroded results of the current second branch attribute and the number of uncorroded results in the actual corrosion results.

[0168] For example, the number of corrosion results is used to determine the recognition accuracy of the second branch attribute by comparing the ratio of the number of identified corrosion results to the actual number of corrosion results.

[0169] In this embodiment, the identification accuracy of the second branch attribute can be obtained by the ratio of the number of identified corrosion results to the actual number of corrosion results. This can effectively verify various corrosion identification conditions and improve the accuracy of corrosion identification of ship components and other components in the marine atmospheric environment-operating condition coupled environment.

[0170] In one embodiment, the verification result of each second branch attribute is determined using the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model, including:

[0171] If the current corrosion identification condition corresponding to the second branch attribute is the first current corrosion identification condition with the second branch attribute, and the identification accuracy of the second branch attribute is greater than the identification accuracy of the initial ship component corrosion identification model, then the verification result of the second branch attribute is determined to be verified.

[0172] If the current corrosion identification condition corresponding to the second branch attribute is not the first current corrosion identification condition with the second branch attribute, and the identification accuracy corresponding to the second branch attribute is greater than the identification accuracy corresponding to the second branch attribute of the previous current corrosion identification condition with the second branch attribute, then the verification result of the second branch attribute is determined to be verified.

[0173] For example, for the first current corrosion identification condition with a second branch attribute, the accuracy of the trained model after identification using a validation set containing 500 data points is calculated. If the accuracy of the trained model for the current corrosion identification condition is >60%, the current corrosion identification condition is retained to improve the model's accuracy. If the accuracy of the trained model for the current corrosion identification condition is ≤60%, the classification variable is considered ineffective and is not classified.

[0174] For each current corrosion identification condition that has a second branch attribute, except for the first one, the current corrosion identification condition is identified and verified in turn. If the identification accuracy of the second branch attribute corresponding to the current corrosion identification condition is higher than that of the previous current corrosion identification condition, the current corrosion identification condition is retained; if the accuracy is lower, it is not classified. This process continues until the entire decision tree is perfected.

[0175] In this embodiment, if the recognition accuracy of the current corrosion recognition condition is higher than that of the previous corrosion recognition condition, the current corrosion recognition condition is retained, thereby improving the accuracy of the model.

[0176] In one embodiment, a method for generating a damage decision tree model for components in a coupled marine atmospheric environment-operating condition environment in the South China Sea is provided. The method includes the following steps:

[0177] The main influencing factors causing corrosion of components in the coupled marine-atmospheric environment and operating conditions of the South China Sea were identified. According to the investigation results, the main influencing factors of the natural environment under the coupled marine-atmospheric environment and operating conditions of the South China Sea include: temperature, humidity, vibration, solar radiation, and salt spray. The main influencing factors of the operating conditions include: potential, polarization, and impedance.

[0178] Define the evaluation indicators for damage.

[0179] According to the investigation results, corrosion damage is mainly manifested in changes in the surface morphology of the test samples. Specific evaluation indicators include: whether the surface macro / micro morphology is corroded, corrosion area, and corrosion depth. This embodiment uses "whether the surface macro / micro morphology is corroded" as the evaluation indicator for subsequent method descriptions. Corrosion area and corrosion depth are similar and will not be repeated.

[0180] Typical components (rotating components, electronic module components, and optical cable components) were selected as research objects to carry out various South China Sea marine atmospheric environment-operating condition coupling tests and collect a large amount of test data.

[0181] For example, taking the test data of rotating parts as an example, the test data is presented in the following table 1:

[0182] Table 1. Corrosion Technical Indicators and Damage Evaluation Analysis

[0183]

[0184]

[0185] The experimental data underwent preliminary processing. First, the continuous experimental data was discretized, as shown in Table 2:

[0186] Table 2 Corrosion Technical Indicators and Damage Evaluation Analysis After Discretization Treatment

[0187]

[0188] Temperature data can be categorized into high temperature (T > 30℃), normal temperature (30℃ > T > 5℃), and low temperature (5℃ > T). Humidity data can be categorized into high humidity (W > 80% RH), relatively high humidity (80% RH > T > 60% RH), medium humidity (60% RH > T > 40% RH), relatively low humidity (40% RH > T > 20% RH), and low humidity (20% RH > T). Other corrosion-influencing factors are similarly categorized.

[0189] The 2000 experimental data points were randomly divided into a training set and a validation set in a 3:1 ratio. The training set contained 1500 experimental data points, and the validation set contained 500 experimental data points.

[0190] Training was conducted using 1500 experimental data points in the training set. First, the actual corrosion results were statistically analyzed: 569 experimental data points showed "corrosion," and 931 showed "no corrosion." The initial entropy H_start of the training set was calculated using the formula: H_start = -a*log₂a - b*log₂b, where a = 569 / 1500 and b = 931 / 1500.

[0191] For each corrosion-influencing factor, such as "temperature", "humidity", and "vibration", the conditional entropy of each corrosion-influencing factor is calculated in turn.

[0192] The probability of "temperature = high temperature" is R_high temperature = 367 / 1500. Of the 367 data points where "temperature = high temperature", 300 data points have a "test result = corrosion", and 67 data points have a "test result = no corrosion". Therefore, the information entropy under the characteristic condition of "temperature = high temperature" can be calculated using the formula: H_high temperature = -a*log2a - b*log2b, where a = 300 / (300+67) and b = 67 / (300+67).

[0193] Using the same method, we can obtain the probability of "temperature = room temperature" as R_room temperature = 858 / 1500. Among all 858 data points of "temperature = room temperature", 120 data points have "experiment result = corrosion", and 738 data points have "experiment result = no corrosion". Then, according to the formula: H_room temperature = -a*log2a - b*log2b, we can calculate the information entropy under the characteristic condition of "temperature = room temperature", where a = 120 / (120+738) and b = 738 / (120+738).

[0194] Using the same method, we can obtain the probability of "temperature = low temperature" occurring as R_low temperature = 275 / 1500, and the information entropy H_low temperature under the characteristic condition of "temperature = low temperature". Finally, we calculate the total entropy under the temperature characteristic value: H_temperature = R_high temperature * H_high temperature + R_normal temperature * H_normal temperature + R_low temperature * H_low temperature.

[0195] For each corrosion-influencing factor such as "temperature", "humidity", and "vibration", the corresponding total entropy is calculated, and the minimum value is selected. That is, among the calculated results of H_temperature, H_humidity, and H_vibration, the feature item corresponding to the minimum value is selected as the first corrosion identification condition.

[0196] For example, if H_vibration is at its minimum, it indicates that "vibration," as a corrosion influencing factor, has the strongest correlation with whether corrosion occurs. In this case, examining the data where "vibration = high," if the percentage of data where "experimental result = corrosion" or "experimental result = no corrosion" exceeds 80% (this 80% threshold can be adjusted for better training results), then it is assumed that "vibration = high" will necessarily result in either "experimental result = corrosion" or "experimental result = no corrosion," and this conditional branch will no longer consider other feature factors. If the percentage of "experimental result = corrosion" or "experimental result = no corrosion" does not exceed this threshold, then it is necessary to further select other corrosion influencing factors such as "temperature," "humidity," and "solar radiation" from the "vibration = high" data to obtain new current corrosion influencing factors for continued classification and identification.

[0197] The entropy (e.g., H_vibration) calculated after each classification and recognition is compared with the initial entropy H_start before classification and recognition. If the entropy does not decrease, it means that the classification is invalid and no further classification is performed.

[0198] The selection principle for the second corrosion identification condition is the same as that for the first corrosion identification condition: in the current dataset, find the minimum information entropy H of each corrosion influencing factor.

[0199] For datasets where "vibration = low", process them in the same way as datasets where "vibration = high".

[0200] Using the validation set, the accuracy of the initial decision tree model trained for assessing damage to components in the coupled marine-atmospheric environment and operating conditions in the South China Sea was calculated.

[0201] In the training set of 1500 data points, 569 data points showed "corrosion" and 931 data points showed "no corrosion". The number of "no corrosion" data points was greater than the number of "corrosion" data points. Therefore, the initial ship component corrosion identification model considered that "all data points in the validation set were no corrosion".

[0202] The initial model, which assumes all data is uncorrupted, was applied to a validation set of 500 data points (300 uncorrupted and 200 corrupted). The accuracy of the initial model was calculated to be 60%.

[0203] Then, using the same method and a validation set containing 500 data points, the accuracy of the training model after identification under the first corrosion identification condition is calculated. If the accuracy of the training model under the current corrosion identification condition is >60%, the first corrosion identification condition is retained to improve the model's accuracy. If the accuracy of the training model under the first corrosion identification condition is ≤60%, the classification variable is considered ineffective and is not classified.

[0204] The accuracy of the classification model is checked sequentially for the second corrosion identification condition, the third corrosion identification condition, and so on. The accuracy of each corrosion identification condition is compared with that of the previous level corrosion identification condition. If the accuracy is higher, the classification is retained; if the accuracy is lower, the classification is not performed, until the entire decision tree is perfected.

[0205] Based on the training results of the decision tree algorithm, a damage decision tree model for components in the coupled marine atmospheric environment and operating conditions of the South China Sea was obtained.

[0206] Taking a rotating component sample as an example, the damage correlation decision tree of its South China Sea marine atmospheric environment-operating condition coupled environment is as follows: Figure 4 As shown.

[0207] Corrosion influencing factors with strong damage correlation in the coupled marine atmospheric environment and working conditions of the South China Sea were screened out.

[0208] This embodiment proposes a standardized process for damage evaluation of components in a coupled marine atmospheric environment and operating conditions in the South China Sea. This process addresses the shortcomings of existing damage evaluation methods in terms of versatility, fills a gap in the industry, and provides a means for related experimental design verification and test evaluation. It can guide the acceleration analysis of component testing in the coupled marine atmospheric environment and operating conditions in the South China Sea.

[0209] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0210] Based on the same inventive concept, this application also provides a component-operating-condition coupled environmental damage assessment device for implementing the aforementioned component-operating-condition coupled environmental damage assessment method in the South China Sea. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more component-operating-condition coupled environmental damage assessment device embodiments provided below can be found in the limitations of the component-operating-condition coupled environmental damage assessment method in the South China Sea described above, and will not be repeated here.

[0211] In one embodiment, such as Figure 5 As shown, a component-operating-condition coupled environmental damage assessment device for the South China Sea marine atmospheric environment is provided, comprising: a sample data acquisition module 510, a current corrosion factor acquisition module 520, a branch attribute module 530, and a recognition model acquisition module 540, wherein:

[0212] The sample data acquisition module 510 is used to acquire multiple sets of corrosion sample data of the sample ship components and to acquire the actual corrosion results corresponding to each corrosion sample data; among them, each corrosion sample data corresponds to different corrosion influencing factors.

[0213] The current corrosion factor acquisition module 520 is used to determine the corrosion factor with the highest correlation with the actual corrosion results based on the correlation between each corrosion influencing factor and the actual corrosion results, and to determine the current corrosion identification conditions based on the current corrosion influencing factor.

[0214] The branch attribute module 530 is used to, when the branch attribute of the condition branch corresponding to the current corrosion identification condition is the first branch attribute that cannot directly identify the corrosion result of the ship component, select the corrosion influencing factor with the highest correlation from the corrosion influencing factors other than the current corrosion identification condition as the new current corrosion influencing factor, and return to execute the step of determining the current corrosion identification condition based on the current corrosion influencing factor, until the branch attribute of the condition branch corresponding to the current corrosion identification condition is the second branch attribute that can directly identify the corrosion result of the ship component.

[0215] The identification model acquisition module 540 is used to acquire the corrosion identification results corresponding to each second branch attribute, and based on each current corrosion identification condition and the corrosion identification results corresponding to each second branch attribute, obtain the corrosion identification model of the target ship component, and use the corrosion identification model of the target ship component to evaluate the corrosion evaluation data of the pre-acquired ship component.

[0216] In one embodiment, the actual corrosion result includes a first corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as corroded, and a second corrosion result that characterizes the corrosion result corresponding to the corrosion sample data as uncorroded.

[0217] The device also includes a target influencing factor acquisition module, a sub-correlation degree acquisition module, and a correlation degree acquisition module.

[0218] The target influencing factor acquisition module is used to acquire the target corrosion influencing factors and multiple pre-defined influencing factor classification intervals for the target corrosion influencing factors; the target corrosion influencing factor is any one of the various corrosion influencing factors. The sub-correlation degree acquisition module is used to obtain the sub-correlation degree between each influencing factor classification interval and the actual corrosion result based on the number of first corrosion sample data and the number of second corrosion sample data contained in each influencing factor classification interval; the first corrosion sample data refers to the corrosion sample data in each influencing factor classification interval whose corresponding actual corrosion result is the first corrosion result; the second corrosion sample data refers to the corrosion sample data in each influencing factor classification interval whose corresponding actual corrosion result is the second corrosion result. The correlation degree acquisition module is used to obtain the correlation degree between the target corrosion influencing factor and the actual corrosion result based on each sub-correlation degree.

[0219] In one embodiment, the current corrosion factor acquisition module includes a classification interval unit and a corrosion identification condition unit.

[0220] The classification interval unit is used to obtain multiple classification intervals for the current corrosion influencing factor. The corrosion identification condition unit is used to obtain multiple corrosion identification conditions for the current corrosion influencing factor based on the multiple influencing factor classification intervals; the current corrosion identification condition is any one of the multiple corrosion identification conditions for the current corrosion influencing factor.

[0221] In one embodiment, each corrosion sample data package contains the influencing factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the influencing factor classification interval is characterized by the influencing factor value interval;

[0222] The device also includes an influencing factor value module and a corrosion sample data module.

[0223] The Influence Factor Values ​​module is used to obtain the influence factor values ​​contained in each influence factor value range. The Corrosion Sample Data module is used to take the corrosion sample data corresponding to the influence factor values ​​contained in each influence factor value range as the corrosion sample data contained in each influence factor classification range.

[0224] In one embodiment, each corrosion sample data package contains the influence factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the actual corrosion results include a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded.

[0225] The device also includes a conditional branch acquisition module, a quantity and acquisition module, a first branch attribute confirmation module, and a second branch attribute confirmation module. The identification model acquisition module includes a first corrosion identification result unit and a second corrosion result unit.

[0226] The branch acquisition module is used to acquire the condition branches corresponding to the current corrosion identification conditions. The quantity and sum acquisition module is used to acquire the first quantity of the influencing factor values ​​corresponding to the first corrosion result in the condition branch, and the second quantity of the influencing factor values ​​corresponding to the second corrosion result in the condition branch, and acquire the sum of the first quantity and the second quantity. The first branch attribute confirmation module is used to confirm the branch attribute of the condition branch as the first branch attribute if a first preset condition is met; the first preset condition is that the ratio of the first quantity to the sum of quantities is less than or equal to a preset ratio threshold, and the ratio of the second quantity to the sum of quantities is less than or equal to the ratio threshold. The second branch attribute confirmation module is used to confirm the branch attribute of the condition branch as the second branch attribute if a second preset condition is met; the second preset condition is that the ratio of the first quantity to the sum of quantities is greater than the ratio threshold, or the ratio of the second quantity to the sum of quantities is greater than the ratio threshold. The first corrosion identification result unit is used to take the first corrosion result corresponding to the first quantity as the corrosion identification result when the ratio of the first quantity to the sum of quantities is greater than the ratio threshold. The second corrosion result unit is used to take the second corrosion result corresponding to the second quantity as the corrosion identification result when the ratio of the second quantity to the sum of quantities is greater than the ratio threshold.

[0227] In one embodiment, the actual corrosion result includes a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded.

[0228] The recognition model acquisition module includes an initial recognition model generation unit, an experimental data result unit, an initial accuracy unit, an initial model recognition unit, a verification result determination unit, and a target recognition model generation unit.

[0229] The initial identification model generation unit generates an initial corrosion identification model for ship components based on the current corrosion identification conditions and the corrosion identification results corresponding to each second branch attribute. The experimental data result unit acquires multiple sets of corrosion verification data corresponding to the corrosion sample data of the sample ship components, and obtains the actual corrosion results corresponding to each corrosion verification data set. The initial accuracy unit determines the identification accuracy of the initial ship component corrosion identification model based on the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each corrosion verification data set. The initial model identification unit inputs multiple sets of corrosion verification data into the initial ship component corrosion identification model to obtain the corrosion identification results corresponding to each second branch attribute, and determines the identification accuracy verification result determination unit based on the number of corrosion identification results corresponding to each second branch attribute and the number of first and second corrosion results in the actual corrosion results corresponding to each corrosion verification data set. The verification result determination unit uses the identification accuracy of each second branch attribute and the identification accuracy of the initial ship component corrosion identification model to determine the verification result of each second branch attribute. The target identification model generation unit is used to obtain the verified second branch attributes based on the verification results of each second branch attribute, and to obtain the corrosion identification model of the target ship component based on the current corrosion identification conditions corresponding to the verified second branch attributes and the corrosion identification results corresponding to the verified second branch attributes.

[0230] In one embodiment, the initial model identification unit includes a verification data determination unit, a first result accuracy unit, and a second result accuracy unit.

[0231] The verification data determination unit is used to determine the corrosion verification data that meets the current corrosion identification conditions corresponding to the current second branch attribute, and use this as the corrosion verification data for the current second branch attribute; the current second branch attribute can be any one of the second branch attributes. The first result accuracy unit is used to calculate the identification accuracy of the current second branch attribute if the corrosion identification result of the current second branch attribute is a first corrosion result, based on the number of first corrosion results in the current second branch attribute and the number of first corrosion results in the actual corrosion results. The second result accuracy unit is used to calculate the identification accuracy of the current second branch attribute if the corrosion identification result of the current second branch attribute is a second corrosion result, based on the number of second corrosion results in the current second branch attribute and the number of second corrosion results in the actual corrosion results.

[0232] In one embodiment, the initial model identification unit includes a first second branch verification unit and a non-first second branch verification unit.

[0233] The first second-branch verification unit is used when the current corrosion identification condition corresponding to the second-branch attribute is the first current corrosion identification condition with a second-branch attribute. If the identification accuracy of the second-branch attribute is greater than the identification accuracy of the initial ship component corrosion identification model, then the verification result of the second-branch attribute is determined to be verified. The non-first second-branch verification unit is used when the current corrosion identification condition corresponding to the second-branch attribute is not the first current corrosion identification condition with a second-branch attribute. If the identification accuracy of the second-branch attribute is greater than the identification accuracy of the second-branch attribute corresponding to the previous current corrosion identification condition with a second-branch attribute, then the verification result of the second-branch attribute is determined to be verified.

[0234] Each module in the aforementioned South China Sea marine atmospheric environment-operating condition coupled environmental damage assessment device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0235] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores corrosion sample data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for evaluating the environmental damage of components in the South China Sea marine atmospheric environment-operating conditions coupled environment.

[0236] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0237] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0238] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0239] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0240] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0241] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0242] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for evaluating the coupled environmental damage of components in the South China Sea marine atmospheric environment and operating conditions, characterized in that, The method includes: Multiple sets of corrosion sample data for sample ship components were obtained, and the actual corrosion results corresponding to each corrosion sample data were obtained; wherein, each corrosion sample data corresponds to different corrosion influencing factors; Based on the correlation between each corrosion influencing factor and the actual corrosion result, the corrosion influencing factor with the highest correlation is determined as the current corrosion influencing factor. The current corrosion identification condition is then determined based on the current corrosion influencing factor, including: obtaining multiple influencing factor classification intervals corresponding to the current corrosion influencing factor; obtaining multiple corrosion identification conditions corresponding to the current corrosion influencing factor based on the multiple influencing factor classification intervals; the current corrosion identification condition is any one of the multiple corrosion identification conditions corresponding to the current corrosion influencing factor; each corrosion sample data package contains influencing factor values ​​for the sample ship components corresponding to different corrosion influencing factors; the influencing factor classification intervals are characterized by influencing factor value intervals. If the branch attribute of the condition branch corresponding to the current corrosion identification condition is a first branch attribute that cannot directly identify the corrosion result of the ship component, then, based on the degree of correlation, the corrosion influencing factor with the highest degree of correlation is selected from the corrosion influencing factors other than the current corrosion identification condition as the new current corrosion influencing factor, and the step of determining the current corrosion identification condition based on the current corrosion influencing factor is returned to be executed until the branch attribute of the condition branch corresponding to the current corrosion identification condition is a second branch attribute that can directly identify the corrosion result of the ship component. Obtain the corrosion identification results corresponding to each of the second branch attributes, and based on each of the current corrosion identification conditions and the corrosion identification results corresponding to each of the second branch attributes, obtain a corrosion identification model for the target ship component, including: taking corrosion verification data that meets the current corrosion identification conditions corresponding to the current second branch attribute as the corrosion verification data of the current second branch attribute; the current second branch attribute is any one of the second branch attributes; if the corrosion identification result of the current second branch attribute is a first corrosion result, then based on the number of first corrosion results of the current second branch attribute and the number of first corrosion results in the actual corrosion results, obtain the identification accuracy of the current second branch attribute; if the current corrosion identification result of the second branch attribute is a first corrosion result, then obtain the identification accuracy of the current second branch attribute based on the number of first corrosion results of the current second branch attribute and the number of first corrosion results in the actual corrosion results .... If the corrosion identification result of the second branch attribute is the second corrosion result, then the identification accuracy of the current second branch attribute is obtained based on the number of second corrosion results of the current second branch attribute and the number of second corrosion results in the actual corrosion results. Using the identification accuracy of each second branch attribute and the identification accuracy of the initial ship component corrosion identification model, the verification result of each second branch attribute is determined. Based on the verification results of each second branch attribute, the verified second branch attributes are obtained, and based on the current corrosion identification conditions corresponding to the verified second branch attributes and the corrosion identification results corresponding to the verified second branch attributes, the target ship component corrosion identification model is obtained. The corrosion identification model for the target ship components is used to evaluate the corrosion evaluation data of the pre-acquired ship components. Before determining the corrosion influencing factor with the highest correlation as the current corrosion influencing factor based on the correlation between each corrosion influencing factor and the actual corrosion result, the method further includes: taking the corrosion sample data corresponding to the influencing factor values ​​contained in each influencing factor value interval as the corrosion sample data contained in each influencing factor classification interval; obtaining the sub-correlation degree between each influencing factor classification interval and the actual corrosion result based on the number of first corrosion sample data contained in each influencing factor classification interval and the number of second corrosion sample data contained in each influencing factor classification interval; the first corrosion sample data is the corrosion sample data in each influencing factor classification interval whose corresponding actual corrosion result is the first corrosion result; the second corrosion sample data is the corrosion sample data in each influencing factor classification interval whose corresponding actual corrosion result is the second corrosion result; the actual corrosion result includes a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded.

2. The method according to claim 1, characterized in that, Before determining the corrosion influencing factor with the highest correlation as the current corrosion influencing factor based on the correlation between each corrosion influencing factor and the actual corrosion result, the method further includes: The target corrosion influencing factors are obtained, along with multiple influencing factor classification intervals pre-defined for the target corrosion influencing factors; the target corrosion influencing factor is any one of the corrosion influencing factors. Based on the correlation degree of each sub-correlation, the correlation degree between the target corrosion influencing factor and the actual corrosion result is obtained.

3. The method according to claim 1, characterized in that, Each corrosion sample data package contains the influence factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the actual corrosion results include a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded; After determining the current corrosion identification conditions based on the current corrosion influencing factors, the process includes: Obtain the conditional branch corresponding to the current corrosion identification condition; Obtain the first number of influencing factor values ​​corresponding to the first corrosion result in the conditional branch, and the second number of influencing factor values ​​corresponding to the second corrosion result in the conditional branch, and obtain the sum of the first number and the second number; If the first preset condition is met, the branch attribute of the condition branch is confirmed as the first branch attribute; the first preset condition is that the ratio of the first quantity to the sum of the quantities is less than or equal to a preset ratio threshold, and the ratio of the second quantity to the sum of the quantities is less than or equal to the ratio threshold. If the second preset condition is met, the branch attribute of the condition branch is confirmed as the second branch attribute; the second preset condition is that the ratio of the first quantity to the sum of the quantities is greater than the ratio threshold, or the ratio of the second quantity to the sum of the quantities is greater than the ratio threshold. The step of obtaining the corrosion identification results corresponding to each of the second branch attributes includes: If the ratio of the first quantity to the sum of the quantities is greater than the ratio threshold, the first corrosion result corresponding to the first quantity is taken as the corrosion identification result. If the ratio of the second quantity to the sum of the quantities is greater than the ratio threshold, the second corrosion result corresponding to the second quantity is taken as the corrosion identification result.

4. The method according to claim 1, characterized in that, The actual corrosion results include a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded. The corrosion identification model for the target ship component is obtained based on each of the current corrosion identification conditions and the corrosion identification results corresponding to each of the second branch attributes, including: Based on the current corrosion identification conditions and the corrosion identification results corresponding to each of the second branch attributes, an initial corrosion identification model for ship components is generated. Multiple sets of corrosion verification data corresponding to corrosion sample data of sample ship components are obtained, and the actual corrosion results corresponding to each of the corrosion verification data are obtained respectively. Based on the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each of the corrosion verification data, the recognition accuracy of the initial ship component corrosion identification model is determined. The multiple sets of corrosion verification data are input into the initial ship component corrosion identification model to obtain the corrosion identification results corresponding to each of the second branch attributes.

5. The method according to claim 4, characterized in that, The determination of the verification result of each second branch attribute using the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model includes: If the current corrosion identification condition corresponding to the second branch attribute is the first current corrosion identification condition with the second branch attribute, and if the identification accuracy corresponding to the second branch attribute is greater than the identification accuracy of the initial ship component corrosion identification model, then the verification result of the second branch attribute is determined to be verified. If the current corrosion identification condition corresponding to the second branch attribute is not the first current corrosion identification condition with the second branch attribute, and the identification accuracy corresponding to the second branch attribute is greater than the identification accuracy corresponding to the second branch attribute of the previous current corrosion identification condition with the second branch attribute, then the verification result of the second branch attribute is determined to be verified.

6. A component-operating condition coupled environmental damage assessment device for the South China Sea marine atmospheric environment, characterized in that, The device includes: The sample data acquisition module is used to acquire multiple sets of corrosion sample data of the sample ship components, and to acquire the actual corrosion results corresponding to each corrosion sample data; wherein, each corrosion sample data corresponds to different corrosion influencing factors; The current corrosion factor acquisition module is used to determine the corrosion factor with the highest correlation with the actual corrosion result from the corrosion influencing factors as the current corrosion influencing factor, and to determine the current corrosion identification condition based on the current corrosion influencing factor. Specifically, it is used to acquire multiple influencing factor classification intervals corresponding to the current corrosion influencing factor; and to obtain multiple corrosion identification conditions corresponding to the current corrosion influencing factor based on the multiple influencing factor classification intervals. The current corrosion identification condition is any one of the multiple corrosion identification conditions corresponding to the current corrosion influencing factor. Each corrosion sample data package contains the influencing factor values ​​of the sample ship component corresponding to different corrosion influencing factors. The influencing factor classification interval is characterized by an influencing factor value interval. The branch attribute module is used to, when the branch attribute of the condition branch corresponding to the current corrosion identification condition is a first branch attribute that cannot directly identify the corrosion result of the ship component, select the corrosion influencing factor with the highest correlation from the corrosion influencing factors other than the current corrosion identification condition as the new current corrosion influencing factor, and return to execute the step of determining the current corrosion identification condition based on the current corrosion influencing factor, until the condition branch corresponding to the current corrosion identification condition is a second branch attribute that can directly identify the corrosion result of the ship component; The identification model acquisition module is used to acquire the corrosion identification results corresponding to each of the second branch attributes, and based on each of the current corrosion identification conditions and the corrosion identification results corresponding to each of the second branch attributes, to obtain the corrosion identification model of the target ship component. Specifically, it is used to take the corrosion verification data that meets the current corrosion identification conditions corresponding to the current second branch attribute as the corrosion verification data of the current second branch attribute; the current second branch attribute is any second branch attribute; if the corrosion identification result of the current second branch attribute is a first corrosion result, then the identification accuracy of the current second branch attribute is obtained according to the number of first corrosion results of the current second branch attribute and the number of first corrosion results in the actual corrosion results; if the corrosion identification result of the current second branch attribute is a first corrosion result, then the identification accuracy of the current second branch attribute is obtained according to the number of first corrosion results of the current second branch attribute and the number of first corrosion results in the actual corrosion results. If the corrosion result is the second corrosion result, then based on the number of second corrosion results of the current second branch attribute and the number of second corrosion results in the actual corrosion results, the recognition accuracy of the current second branch attribute is obtained; using the recognition accuracy of each second branch attribute and the recognition accuracy of the initial ship component corrosion recognition model, the verification result of each second branch attribute is determined; based on the verification results of each second branch attribute, the verified second branch attributes are obtained, and based on the current corrosion recognition conditions corresponding to the verified second branch attributes and the corrosion recognition results corresponding to the verified second branch attributes, the target ship component corrosion recognition model is obtained; using the target ship component corrosion recognition model, the pre-acquired corrosion evaluation data of the ship components is evaluated; It also includes: a sub-correlation degree acquisition module, used to take the corrosion sample data corresponding to the influence factor values ​​contained in each influence factor value interval as the corrosion sample data contained in each influence factor classification interval; based on the number of first corrosion sample data contained in each influence factor classification interval and the number of second corrosion sample data contained in each influence factor classification interval, to obtain the sub-correlation degree between each influence factor classification interval and the actual corrosion result; the first corrosion sample data is the corrosion sample data in each influence factor classification interval whose corresponding actual corrosion result is the first corrosion result; the second corrosion sample data is the corrosion sample data in each influence factor classification interval whose corresponding actual corrosion result is the second corrosion result; the actual corrosion result includes a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded.

7. The apparatus according to claim 6, characterized in that, The device further includes: a target influencing factor acquisition module and a correlation degree acquisition module; The target influencing factor acquisition module is used to acquire the target corrosion influencing factor and multiple influencing factor classification intervals pre-defined for the target corrosion influencing factor; the target corrosion influencing factor is any one of the corrosion influencing factors. The correlation degree acquisition module is used to obtain the correlation degree between the target corrosion influencing factor and the actual corrosion result based on each of the sub-correlation degrees.

8. The apparatus according to claim 6, characterized in that, Each corrosion sample data package contains the influence factor values ​​of different corrosion influencing factors corresponding to the sample ship components; the actual corrosion results include a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded; The device further includes: a branch acquisition module, a quantity and acquisition module, a first branch attribute confirmation module, and a second branch attribute confirmation module; The branch acquisition module is used to acquire the condition branch corresponding to the current corrosion identification condition; The quantity and acquisition module is used to acquire a first quantity of the influencing factor values ​​corresponding to the first corrosion result in the conditional branch, and a second quantity of the influencing factor values ​​corresponding to the second corrosion result in the conditional branch, and to acquire the sum of the first quantity and the second quantity; The first branch attribute confirmation module is used to confirm the branch attribute of the conditional branch as the first branch attribute if a first preset condition is met; the first preset condition is that the ratio of the first quantity to the sum of the quantities is less than or equal to a preset ratio threshold, and the ratio of the second quantity to the sum of the quantities is less than or equal to the ratio threshold. The second branch attribute confirmation module is used to confirm the branch attribute of the conditional branch as the second branch attribute if a second preset condition is met; the second preset condition is that the ratio of the first quantity to the sum of the quantities is greater than the ratio threshold, or the ratio of the second quantity to the sum of the quantities is greater than the ratio threshold. The recognition model acquisition module is used for: If the ratio of the first quantity to the sum of the quantities is greater than the ratio threshold, the first corrosion result corresponding to the first quantity is taken as the corrosion identification result. If the ratio of the second quantity to the sum of the quantities is greater than the ratio threshold, the second corrosion result corresponding to the second quantity is taken as the corrosion identification result.

9. The apparatus according to claim 6, characterized in that, The actual corrosion results include a first corrosion result indicating that the corrosion result corresponding to the corrosion sample data is corroded, and a second corrosion result indicating that the corrosion result corresponding to the corrosion sample data is not corroded. The recognition model acquisition module is used for: Based on the current corrosion identification conditions and the corrosion identification results corresponding to each of the second branch attributes, an initial corrosion identification model for ship components is generated. Multiple sets of corrosion verification data corresponding to corrosion sample data of sample ship components are obtained, and the actual corrosion results corresponding to each of the corrosion verification data are obtained respectively. Based on the number of first corrosion results and the number of second corrosion results in the actual corrosion results corresponding to each of the corrosion verification data, the recognition accuracy of the initial ship component corrosion identification model is determined. The multiple sets of corrosion verification data are input into the initial ship component corrosion identification model to obtain the corrosion identification results corresponding to each of the second branch attributes.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

Images