Leather cultural relic microbial disease deterioration prediction method and device and computer equipment

Abstract

The invention relates to the technical field of microbial detection, and discloses a leather cultural relic microbial disease deterioration prediction method and device and computer equipment, and the method comprises the steps: obtaining the state data of each leather cultural relic in time sequence; according to a preset parameter threshold value of each microbial disease type, processing the state data of each leather cultural relic at each moment, and determining a comprehensive risk level of degradation of each leather cultural relic by the microbial disease at each moment; inputting the state data of each leather cultural relic in the time sequence and the comprehensive risk level into a leather cultural relic degradation risk assessment model to obtain a trained leather cultural relic degradation risk assessment model; and processing the state data of the leather cultural relics in the time sequence by using the trained leather cultural relic degradation risk assessment model, and predicting to obtain predicted risk levels of the leather cultural relics at multiple moments in the future. The method has the beneficial effect of improving the deterioration prediction accuracy of the leather cultural relics.

Description

Technical Field

[0001] This application relates to the field of microbial detection technology, and in particular to a method, apparatus and computer equipment for predicting the deterioration of leather artifacts by microbial diseases. Background Technology

[0002] Leather artifacts refer to historical relics made from animal hides and fur that possess historical, artistic, and scientific and technological value.

[0003] Currently, researchers typically use environmental monitoring equipment to record environmental parameters such as temperature, humidity, and light to predict the degree of deterioration of leather artifacts.

[0004] However, environment-based predictions are prone to inaccuracies. Summary of the Invention

[0005] This application provides a method, apparatus, and computer equipment for predicting the deterioration of leather artifacts due to microbial diseases, which solves the technical problem of inaccurate prediction of deterioration of leather artifacts and achieves the technical effect of improving the accuracy of prediction of deterioration of leather artifacts.

[0006] To achieve the above objectives, the main technical solutions adopted in this application include: In a first aspect, embodiments of this application provide a method for predicting the deterioration of leather artifacts due to microbial diseases, including: Obtain the temporal status data of each leather artifact; Based on the preset parameter thresholds for each type of microbial disease, the state data of each leather artifact at each time point are processed to determine the comprehensive risk level of each leather artifact being deteriorated by microbial diseases at each time point. The time-series state data of each leather artifact and the comprehensive risk level are input into the leather artifact deterioration risk assessment model to obtain the trained leather artifact deterioration risk assessment model. The trained leather artifact deterioration risk assessment model is used to process the time-series state data of the leather artifact to predict the predicted risk level of the leather artifact at multiple future moments.

[0007] In this embodiment, the risk of leather artifacts being deteriorated by microbial diseases is accurately assessed and predicted, thereby improving the efficiency of monitoring and managing leather artifacts. In one example, the status data includes multiple status indicators; based on preset parameter thresholds for each type of microbial disease, the status data of each leather artifact at each time point is processed to determine the comprehensive risk level of microbial disease deterioration of each leather artifact at each time point, including: Obtain the preset status indicators for each type of microbial disease from the status data; The status indicators are processed using preset parameter thresholds for each type of microbial disease to obtain a quantitative score for the type of microbial disease. Based on the quantitative scores of each of the aforementioned microbial disease types, the comprehensive risk level of the leather artifact being deteriorated by microbial diseases at that time is calculated.

[0008] This example demonstrates the ability to accurately assess the risk of leather artifacts deteriorating due to microbial damage at a specific time.

[0009] In one example, the parameter thresholds include an indicator threshold and a quantification threshold; when a desired state indicator is preset for the microbial disease type, the preset parameter thresholds for each microbial disease type are used to process the state indicator to obtain a quantification score for the microbial disease type, including: Based on the value of the state index, determine the corresponding index threshold and quantization threshold in the parameter threshold; Determine the proportion of the status indicator within the indicator threshold; The quantitative score of the microbial disease type is obtained based on the occupancy ratio and the quantification threshold.

[0010] This example demonstrates the ability to accurately quantify the impact of microbial diseases on leather artifacts.

[0011] In one example, the parameter thresholds include indicator thresholds and quantification thresholds; when multiple state indicators are preset for the microbial disease type, the preset parameter thresholds for each microbial disease type are used to process the state indicators to obtain a quantification score for the microbial disease type, including: Based on each state indicator and its value, determine the indicator threshold and quantization threshold corresponding to the state indicator in the parameter threshold; Determine the proportion of each of the aforementioned status indicators in the threshold values ​​corresponding to the status indicators; Based on the occupancy ratio and the quantification threshold corresponding to the status indicator, the indicator score corresponding to the status indicator is obtained; Based on the preset weights of each status indicator, the weighted sum of the scores of each indicator is calculated to obtain the quantitative score of the microbial disease type.

[0012] This example demonstrates the ability to accurately quantify the impact of microbial diseases on leather artifacts.

[0013] In one example, based on the quantitative scores of each of the aforementioned microbial disease types, a comprehensive risk level of microbial disease deterioration of the leather artifact at that time is calculated, including: Based on the mapping relationship between quantitative thresholds and risk levels, the risk level of each of the aforementioned microbial disease types is determined according to the quantitative scores of each type. The risk levels of each of the aforementioned microbial disease types are weighted and summed to obtain the comprehensive risk level of the leather artifact being deteriorated by microbial diseases at the stated time.

[0014] This example demonstrates the ability to accurately assess the overall risk level of microbial damage to leather artifacts at a specific time.

[0015] In one example, the temporal state data of each leather artifact and the comprehensive risk level are input into a leather artifact deterioration risk assessment model to obtain a trained leather artifact deterioration risk assessment model, including: Using a preset sliding window, the input data for the current iteration is obtained from the state data of each leather artifact in the time sequence; The aforementioned leather artifact deterioration risk assessment model is used to process the time-series state data to predict the predicted risk level for the next moment. Based on the comprehensive risk level at the next moment and the predicted risk level, the loss value of the current iteration is calculated using the loss function of the leather cultural relic deterioration risk assessment model. Based on the loss value, the leather artifact deterioration risk assessment model is optimized and trained.

[0016] This example demonstrates how to improve model training efficiency and model accuracy.

[0017] In one example, the method further includes: If the predicted risk level is greater than a preset threshold, simulated state data for the moment when the predicted risk level is greater than the preset threshold is generated based on the state data of the leather artifact. The simulated state data is input into the restoration suggestion model to generate restoration suggestions for the leather artifact; wherein, the restoration suggestion model is a large language model generated based on the state data and restoration strategies of leather artifacts in historical restoration cases.

[0018] In this example, the goal is to generate accurate restoration suggestions for leather artifacts at an early stage when abnormalities occur.

[0019] Secondly, embodiments of this application provide a device for predicting the deterioration of leather artifacts due to microbial diseases, the device comprising: The acquisition module is used to acquire the temporal status data of each leather artifact. The training module is used to process the state data of each leather artifact at each time step according to the preset parameter thresholds of each type of microbial disease, and determine the comprehensive risk level of each leather artifact being deteriorated by microbial diseases at each time step; the state data of each leather artifact in time step and the comprehensive risk level are input into the leather artifact deterioration risk assessment model to obtain the trained leather artifact deterioration risk assessment model. The prediction module is used to process the time-series state data of the leather artifact using the trained leather artifact deterioration risk assessment model, and to predict the predicted risk level of the leather artifact at multiple future moments.

[0020] Thirdly, embodiments of this application provide a computer device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the method described in any of the above embodiments.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer instructions, which are used to cause a computer to perform the method described in any one of the above embodiments.

[0022] Fifthly, embodiments of this application provide a computer program product, including computer instructions, which are used to cause a computer to perform the method described in any of the above embodiments. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 A flowchart illustrating a method for predicting the deterioration of leather artifacts due to microbial diseases, provided in this application embodiment; Figure 2 A flowchart illustrating a method for predicting the deterioration of leather artifacts due to microbial diseases, provided in this application embodiment; Figure 3 A schematic diagram of a predicted trend provided in an embodiment of this application; Figure 4 A schematic diagram of the structure of a microbial disease deterioration prediction system for leather cultural relics provided in this application embodiment; Figure 5A module diagram of a device for predicting the microbial disease deterioration of leather artifacts provided in this application embodiment; Figure 6 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] Leather artifacts refer to historical remains made from animal hides and fur, possessing historical, artistic, and scientific value. Compared to other types of organic artifacts, leather artifacts are more difficult to preserve due to their diverse materials, complex tanning processes, and unclear degradation mechanisms. The main component of leather is collagen, whose structure is highly susceptible to degradation under the influence of light, humidity, pH fluctuations, and microbial erosion. Especially in excavated or water-saturated environments, the reproductive activity of microorganisms has a particularly significant impact on leather degradation. Microbial-induced leather deterioration is one of the most common and major types of damage to excavated leather artifacts. Under water-saturated conditions, the degradation process of leather by microorganisms is more active and complex, and leather artifacts buried in damp environments for extended periods become even more vulnerable due to the proliferation of microorganisms in the water.

[0027] Current research generally lacks a systematic microbial characterization database for leather artifacts, resulting in poor data comparability and insufficient traceability between different samples. Furthermore, it is currently impossible to establish a quantitative correlation between microbial communities and leather degradation characteristics. In addition, there is no comprehensive database or intelligent analysis model specifically for leather artifacts, making it difficult to assess the degree of deterioration under microbial influence, perform visual diagnosis, and predict trends. Therefore, there is an urgent need to establish a basic database of microbial diseases affecting leather artifacts, integrating microbial communities, physicochemical characteristics, and environmental factors, and constructing a visual deterioration diagnosis and predictive analysis model.

[0028] To address the aforementioned issues, this application proposes a method for predicting the deterioration of leather artifacts due to microbial diseases. By analyzing the condition data of leather artifacts, the method can determine the correlation between the condition data and microbial diseases, thereby revealing the mechanism by which microorganisms affect the deterioration of leather artifacts. This enables quantitative assessment of diseases and precise preventive protection, providing a scientific basis and technical support for the protection of cultural relics.

[0029] According to an embodiment of this application, a method for predicting the deterioration of leather artifacts due to microbial diseases is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed on a computer device via a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here. The computer device can be a mobile terminal, a personal computer, a server, etc.

[0030] Figure 1 A flowchart illustrating a method for predicting microbial deterioration of leather artifacts provided in this application embodiment is shown below. Figure 1 As shown, with a computer device as the execution subject, the process includes the following steps: S101. Obtain the temporal status data of each leather artifact.

[0031] For example, the computer device first obtains daily records of the status data of each leather artifact, thus obtaining the status data of each leather artifact in time sequence.

[0032] In one implementation, the status data of each leather artifact can be recorded in a database. Alternatively, the collected data can be recorded in a basic database.

[0033] In one implementation, the status data may include information on the microbial community of the leather artifact's surface and surrounding environment. For example, it may include the types, quantities, and activity indicators of the microorganisms.

[0034] In one implementation, the state data may include physical and chemical characteristics. For example, it may include collagen structure, humidity, pH value, spectral information, etc.

[0035] In one implementation, the state data may include environmental parameters. For example, it may include temperature, humidity, light intensity, and historical storage conditions.

[0036] In one implementation, the status data provides information describing the surface condition, thermal properties, and chemical composition of the leather artifact, comprehensively reflecting its deterioration.

[0037] In one implementation, computer equipment can collect various characteristic indicators of leather artifacts at certain time intervals using sensors, high-definition scanners, etc., thereby obtaining state data of leather artifacts at different times.

[0038] In one implementation, computer equipment can also directly test leather artifacts using specialized testing equipment, such as spectrometers and tensile testing machines, to obtain their specific physicochemical characteristics.

[0039] In one implementation, the computer equipment can also use high-definition image acquisition equipment to periodically capture images of the appearance of leather artifacts and use image analysis technology to obtain information on changes in their appearance.

[0040] In one implementation, the computer device can also acquire state data of the leather artifacts recorded during the simulated aging process. Optionally, the acquired data can be recorded in a simulation database.

[0041] In one implementation, the timing sequence can be multiple time points with a fixed time interval. Optionally, the time interval can be a sampling period.

[0042] S102. Based on the preset parameter thresholds for each type of microbial disease, process the state data of each leather artifact at each time point to determine the comprehensive risk level of each leather artifact being deteriorated by microbial diseases at each time point.

[0043] For example, the computer device can acquire expert experience information and, based on that information, compile preset parameter thresholds for each type of microbial disease. The computer device can then input the state data of the leather artifacts at various times into the calculation model for each type of microbial disease, and, based on the parameter thresholds, calculate the comprehensive risk level of each leather artifact's deterioration due to microbial diseases at each time point.

[0044] In one implementation, the type of microbial disease is a disease that may be caused by microorganisms and may damage leather artifacts.

[0045] In one implementation, each type of microbial disease can have its corresponding computational model. The threshold parameter can be a value required by the computational model.

[0046] In one implementation, the computer device can process the state data one by one.

[0047] In one implementation, each type of microbial disease may correspond to one or more status indicators in the status data. A computer device can retrieve the required status indicators for each type of microbial disease from the status data. The computer device can then use these status indicators to calculate the overall risk level of microbial disease deterioration of the leather artifact at the corresponding time point in the status data.

[0048] In one implementation, a computer device can calculate the risk level of the microbial disease type based on the status indicator, using parameter thresholds and a preset calculation model. This risk level indicates the likelihood of the occurrence of that microbial disease type. Alternatively, the risk level indicates the severity of the microbial disease type that has already occurred. Furthermore, the computer device can calculate a comprehensive risk level of microbial disease deterioration for the leather artifact based on the risk levels of each microbial disease type.

[0049] In one implementation, a computer device can calculate a quantitative score for the microbial disease type based on the status indicator, using parameter thresholds and a preset calculation model. This quantitative score is used to determine the risk of the occurrence of the microbial disease type within a fixed quantitative range. Alternatively, the quantitative score can be used to determine the severity of the already occurring microbial disease type within a fixed quantitative range. The computer device can have a preset mapping relationship between values ​​within the quantitative range and risk levels. Based on this mapping relationship, the computer device can determine the risk level of the microbial disease type.

[0050] In one implementation, computer equipment can generate a comprehensive risk level based on the status data of each leather artifact at a given moment.

[0051] In one implementation, the computer device can acquire expert experience information and, based on this information, determine multiple parameter thresholds in a risk level assessment algorithm for state data. Furthermore, the computer device can determine the comprehensive risk level corresponding to each piece of state data based on the risk level assessment algorithm that combines expert experience information to determine the parameter thresholds.

[0052] In one implementation, the expert experience information may include the types of damage to the leather artifacts. Furthermore, the expert experience information may also include criteria for determining the leather damage level corresponding to each type of damage. A computer device can determine a parameter threshold for a type of damage based on these criteria.

[0053] In one implementation, computer equipment can use a combination of hierarchical quantification and weighted fusion to construct a risk level assessment algorithm.

[0054] In one implementation, the computer device can quantitatively calculate the risk level of each disease type corresponding to each state data point. Then, based on the risk levels of each disease type, a weighted and fused calculation is performed to obtain the final comprehensive risk level.

[0055] S103. Input the temporal status data and comprehensive risk level of each leather artifact into the leather artifact deterioration risk assessment model to obtain a trained leather artifact deterioration risk assessment model.

[0056] For example, a computer device can combine a state data point and its corresponding comprehensive risk level into sample data. The computer device can also combine time-series sample data of various leather artifacts into training data. This training data can then be input into a leather artifact deterioration risk assessment model for training, resulting in a well-trained model.

[0057] In one implementation, the leather artifact deterioration risk assessment model can be constructed by combining a Long Short-Term Memory (LSTM) network with a Transformer network.

[0058] In one implementation, the computer device can use a sliding window to collect sample data of a preset length from the training data as input for the current iteration.

[0059] In one implementation, a computer device can use the leather artifact deterioration risk assessment model to predict the risk level for the next moment based on the current state data. Furthermore, the computer device can input the predicted risk level for the next moment and the comprehensive risk level for the next moment into a loss function to train the leather artifact deterioration risk assessment model.

[0060] In one implementation, a computer device can use the training data to train a risk assessment model for the deterioration of leather artifacts. This risk assessment model can be a deep learning model or a neural network model. For example, the risk assessment model can be constructed by combining LSTM and Transformer networks.

[0061] S104. Using the trained leather artifact deterioration risk assessment model, process the temporal state data of the leather artifacts to predict the predicted risk level of the leather artifacts at multiple future moments.

[0062] For example, a computer device can input the temporal state data of leather artifacts into a trained leather artifact deterioration risk assessment model. Based on this existing state data, the model can predict the comprehensive risk level of the leather artifact at future moments and output the predicted risk level of the leather artifact at multiple future moments.

[0063] In one implementation, the number of state data points input into the leather artifact deterioration risk assessment model can be determined based on the sliding window size. Furthermore, the state data input into the leather artifact deterioration risk assessment model is the state data closest to the current moment.

[0064] In one implementation, the leather artifact deterioration risk assessment model is used to predict the state changes of leather artifacts at multiple future moments based on the temporal patterns learned internally, and output the corresponding predicted risk level.

[0065] In one implementation, multiple future moments refer to several points in time after the current point in time, and the specific time intervals and number can be set according to actual needs.

[0066] In one implementation, the predicted risk level is a comprehensive risk level of the leather artifacts that may be deteriorated by microbial diseases in the future.

[0067] In one implementation, the leather artifact deterioration risk assessment model can predict and output the predicted risk level at multiple subsequent time points in a step-by-step prediction manner.

[0068] In this embodiment, by acquiring the temporal state data of leather artifacts and calculating the comprehensive risk level corresponding to each state data based on the parameter threshold of the microbial disease type, and then using the state data and the comprehensive risk level to train a leather artifact deterioration risk assessment model, the leather artifact deterioration risk assessment model is used to predict the future risk level based on the state data, thereby achieving accurate assessment and prediction of the risk of leather artifacts being deteriorated by microbial diseases, and improving the efficiency of monitoring and management of leather artifacts.

[0069] In one example, before obtaining the temporal status data of each leather artifact in step S101 above, the computer device can also generate a leather artifact database.

[0070] For example, a computer device can collect microbial information, physicochemical characteristics, and environmental parameters of a leather artifact over time to form state data of the artifact at various points in time. The computer device can then write this state data into a database to generate a leather artifact database.

[0071] In one implementation, the computer equipment can also collect state data of the leather artifact at various moments during the simulated aging experiment. The computer equipment can then write the simulated data of the leather artifact into a leather artifact database.

[0072] In one implementation, the database may specifically include two parts: a basic database and a simulation database. The basic database can contain state data of the leather artifacts obtained through actual collection at various times. The simulation database can contain state data of the leather artifacts obtained during simulated aging processes at various times.

[0073] In one implementation, a sample in the database can consist of a set of state data of a leather artifact over a certain period of time.

[0074] In one implementation, the computer device can perform statistics and analysis on the data in the database through the database management module.

[0075] In this example, by collecting data on the state of leather artifacts under real and simulated aging scenarios over time and constructing a database containing basic and simulated components, the data is stored in sample form and statistical analysis is performed using the management module to achieve the effect of comprehensively integrating leather artifact data and deeply exploring its changing patterns.

[0076] In one example, building the underlying database includes: For example, computer devices can periodically collect status data of leather artifacts.

[0077] In one implementation, the periodicity can be a day, a week, a month, etc.

[0078] In one implementation, the status data may include microbial information about the surface of the leather artifact and its surrounding environment. This microbial information may include the types, quantities, and activity levels of the microorganisms.

[0079] In one implementation, the state data may further include the physicochemical characteristics of the artifact. These physicochemical characteristics may include collagen structure, humidity, pH value, spectral information, etc.

[0080] In one implementation, the status data may also include environmental parameters of the artifact's location. These environmental parameters may include temperature, humidity, lighting, and historical storage conditions.

[0081] In one implementation, the microbial information, physicochemical characteristics, and environmental parameters can also be written into a basic database from the status data.

[0082] In this example, by constructing a basic database of cultural relics, the data of leather cultural relics at the current moment is obtained, providing a data foundation for the subsequent training of the leather cultural relics deterioration risk assessment model.

[0083] In one example, building a simulated database includes: For example, during the simulation, the computer device can determine the sampling time interval according to a preset sampling period, and acquire the simulated state data based on this time interval.

[0084] In one implementation, the simulated aging experiment can be achieved by adding damp heat cycling, microbial simulation, or other methods.

[0085] In one implementation, the simulated state data may include microbial information, physicochemical characteristics, and environmental information.

[0086] In this example, by constructing a simulated database of cultural relics, data on the deterioration of leather cultural relics is obtained, providing a data foundation for the subsequent training of a risk assessment model for the deterioration of leather cultural relics.

[0087] In one example, the status data may include multiple status indicators. For example, the microbial species can be one status indicator, and pH value can be another. In step S102 above, determining the comprehensive risk level corresponding to each status data point includes: S1021. Obtain the preset state indicators for each type of microbial disease from the state data.

[0088] For example, the computer device may be pre-set with state indicators required for calculating various types of microbial diseases. The computer device can first obtain the state indicator corresponding to each type of microbial disease from the currently processed state data.

[0089] In one implementation, the computer device can determine the type of microbial disease requiring calculation based on expert experience. Furthermore, the computer device can determine, based on expert experience, the necessary indicators for judging whether the type of microbial disease is likely to occur, or the severity of its occurrence. For example, for some microorganisms severely affected by acidic or alkaline environments, pH value needs to be used as a required state indicator.

[0090] S1022. Using the preset parameter thresholds for each type of microbial disease, process the status indicators to obtain a quantitative score for the type of microbial disease.

[0091] For example, the computer device may store preset parameter thresholds for each type of microbial disease. These parameter thresholds are used to indicate the parameter thresholds related to each microbial disease type that need to be used in the calculation formula for that specific microbial disease type.

[0092] Computer equipment can process various state indicators based on this parameter threshold using a preset calculation formula to obtain a quantitative score. This quantitative score is a value within a fixed quantitative range. The quantitative score can indicate the probability or severity of the occurrence of a type of microbial disease.

[0093] In one implementation, the fixed quantization range of the quantized score can be [0, 100]. That is, the computer device can perform percentage-based quantization.

[0094] In one implementation, the higher the quantitative score of the disease type, the higher the risk level. A higher risk level indicates a higher probability of the disease type existing. Alternatively, a higher risk level indicates a higher probability of the disease type occurring.

[0095] In one implementation, when the type of microbial disease corresponds to a status index, the computer device can directly quantify the status index to obtain the final quantitative score.

[0096] In one implementation, when the microbial disease type corresponds to multiple status indicators, the computer device can directly quantify the multiple status indicators, and then weight and fuse the scores of the quantified indicators to obtain the final quantified score.

[0097] S1023. Based on the quantitative scores of each type of microbial disease, the comprehensive risk level of leather cultural relics being deteriorated by microbial diseases at any given time is calculated.

[0098] For example, the computer device comprehensively considers the quantitative scores of various microbial disease types, integrates these scores using a suitable fusion algorithm, and determines the comprehensive risk level of the leather artifact at that moment based on the integrated result.

[0099] In one implementation, the comprehensive risk level is an overall assessment of the likelihood of leather artifacts being deteriorated by all kinds of microbial diseases at a given moment. For example, this comprehensive risk level can be divided into four levels: 0-III. Alternatively, it can be divided into three levels: low risk, medium risk, and high risk.

[0100] In one implementation, the computer device can first map the risk level of each microbial disease based on the quantitative scores of each type of microbial disease. Then, the computer device can integrate and calculate the risk levels of each microbial disease to obtain a comprehensive risk level of deterioration caused by the microbial disease.

[0101] In another implementation, the computer device can first calculate the quantitative scores of each microbial disease to obtain a fusion score. Then, the computer device can map the fusion score to obtain a comprehensive risk level.

[0102] In one implementation, the computer device can determine the status indicators in the status data required to calculate each disease type based on expert experience. Then, using a percentage-based quantification strategy, the status indicators are quantified into a unified range to obtain a quantified score. Subsequently, the computer device can determine the risk level of the disease type based on this quantified score and a preset mapping relationship.

[0103] In this example, by calculating the quantitative scores of each type of microbial disease based on the status indicators in the status data, and then determining the comprehensive risk level based on the quantitative scores, the risk of leather artifacts being deteriorated by microbial diseases at a specific time can be accurately assessed.

[0104] In one example, the parameter threshold may include an index threshold and a quantification threshold. When a desired state index is preset for the microbial disease type, the process of calculating the quantification score in step S1022 above includes: Step 221: Based on the values ​​of the status indicators, determine the corresponding indicator thresholds and quantization thresholds in the parameter thresholds.

[0105] For example, after determining the status index corresponding to the currently calculated microbial disease type, the computer device can obtain the value corresponding to that status index. Since the status index corresponding to that microbial disease type can correspond to at least one set of index thresholds and quantization thresholds within the parameter thresholds, the computer device can determine its corresponding set of index thresholds and quantization thresholds based on the value of that status index.

[0106] In one implementation, for a state indicator of a microbial disease type, a set of indicator thresholds and quantification thresholds can be preset. The computer device can then directly obtain these indicator thresholds and quantification thresholds.

[0107] In another implementation, for a state indicator of a microbial disease type, multiple sets of indicator thresholds and quantization thresholds can be preset. The computer device can then determine the corresponding indicator thresholds and quantization thresholds based on the value of the state indicator. For example, when the state indicator is less than t1, it can correspond to one set of indicator thresholds and quantization thresholds; when the state indicator is greater than t1, it can correspond to another set of indicator thresholds and quantization thresholds.

[0108] In one implementation, the status indicator is specific data used to describe the condition of the leather artifact. It is typically a value within the status data, such as the leather's humidity, pH level, or the number of microbial colonies. Optionally, when this status data is recorded as a table in the database, the status indicator is a field.

[0109] In one implementation, the parameter threshold is a series of pre-set critical values ​​for assessing the degree of microbial damage to leather artifacts.

[0110] In one implementation, the indicator threshold is a range critical value set for a single state indicator.

[0111] Optionally, when an indicator corresponds to an indicator threshold, the indicator threshold can be the normal range of values ​​for that state indicator. For example, a computer device can determine the threshold range based on the maximum and minimum values ​​of that state indicator determined from all the data in the database.

[0112] Optionally, when an indicator corresponds to multiple indicator thresholds, the indicator threshold can be a segmented value of the state indicator.

[0113] In one implementation, the threshold for this indicator can be determined based on expert experience.

[0114] In one implementation, the quantization threshold is the range of values ​​for the quantized score when converting the value of the state indicator into a quantized score. This quantization threshold is usually a fixed range. For example, in percentage-based quantization, the quantization threshold is [0, 100].

[0115] In one implementation, the quantization threshold can be determined based on computational needs.

[0116] Step 222: Determine the proportion of status indicators in the indicator thresholds.

[0117] For example, after obtaining the current value of a status indicator, the computer device compares it with a predetermined range of indicator thresholds, and calculates the position of the value within the indicator threshold range to determine its proportion within the indicator threshold.

[0118] For example, if the threshold range of a certain status indicator is 0 to 100, and the current value is 60, then its proportion in the threshold is 60%.

[0119] In one implementation, the calculation formula can be: in, This represents the percentage of space occupied. This is a status indicator. and These are the maximum and minimum values ​​of the threshold for this indicator.

[0120] Step 223: Based on the occupancy ratio and quantification threshold, obtain the quantitative score of the microbial disease type.

[0121] For example, the computer device calculates the occupancy ratio of the status indicator and performs a calculation with a pre-set quantification threshold, converting the occupancy ratio into a specific numerical value according to a specific quantification rule. This numerical value is the quantification score of the microbial disease type on that status indicator.

[0122] In one implementation, the quantitative score is a numerical representation of the degree to which leather artifacts are affected by specific microbial diseases; the higher the score, the more severe the impact.

[0123] In one implementation, the formula for calculating the quantitative score can be: in, To quantify the score. and This represents a unified range of quantitative scores corresponding to this disease type. For example, It is 100. It is 0. This represents the percentage of space occupied.

[0124] In this example, by obtaining the status indicators and then using indicator thresholds and quantification thresholds to calculate and obtain a quantification score, the effect of accurately quantifying the degree of impact of microbial diseases on leather artifacts is achieved.

[0125] In one example, the parameter threshold includes an indicator threshold and a quantification threshold. When multiple state indicators are preset for a microbial disease type, the process of calculating the quantification score in step S1022 above includes: Step 224: Based on each state indicator and its value, determine the indicator threshold and quantization threshold corresponding to the state indicator in the parameter threshold.

[0126] For example, after determining multiple state indicators corresponding to the currently calculated microbial disease type, the computer device can obtain the value of each state indicator. Then, based on this value, the computer device can select one set of indicator thresholds and quantization thresholds from multiple sets of indicator thresholds and quantization thresholds included in the parameter thresholds for calculating the state indicator.

[0127] In one implementation, for a state indicator of a microbial disease type, a set of indicator thresholds and quantification thresholds can be preset. The computer device can then directly obtain these indicator thresholds and quantification thresholds.

[0128] In another implementation, for a state indicator of a microbial disease type, multiple sets of indicator thresholds and quantization thresholds can be preset. The computer device can then determine the corresponding indicator thresholds and quantization thresholds based on the value of the state indicator. For example, when the state indicator is less than t1, it can correspond to one set of indicator thresholds and quantization thresholds; when the state indicator is greater than t1, it can correspond to another set of indicator thresholds and quantization thresholds.

[0129] Step 225: Determine the proportion of each status indicator in the corresponding indicator threshold.

[0130] For example, after acquiring the value of each status indicator, the computer device can compare it in detail with a pre-determined threshold range corresponding to that status indicator. The computer device can determine the proportion of the value within the corresponding threshold by accurately calculating its position within the threshold range.

[0131] For example, if the threshold range of a certain status indicator is 50 to 70, and the current value is 60, then its proportion in the threshold range is 50%.

[0132] In one implementation, the occupancy ratio is a key value used to measure the degree to which the condition indicator value deviates from the indicator threshold. This occupancy ratio can intuitively reflect the likelihood that leather artifacts are affected by microbial diseases in terms of that condition indicator.

[0133] In one implementation, a state indicator for a microbial disease type can have multiple preset threshold values ​​and quantification thresholds. Assume that the threshold value corresponding to a state indicator is... The formula for calculating this occupancy ratio can be: in, This represents the percentage of space occupied. This is a status indicator. and These are the upper and lower limits of the indicator threshold.

[0134] Step 226: Based on the occupancy ratio and the quantification threshold corresponding to the status indicator, obtain the indicator score corresponding to the status indicator.

[0135] For example, the computer device calculates the occupancy ratio of the status indicator and performs a calculation with a pre-set quantization threshold corresponding to the status indicator. According to a specific quantization rule, the occupancy ratio is converted into a specific numerical value. This numerical value is the indicator score corresponding to the status indicator.

[0136] In one implementation, the index score is a numerical representation of the degree to which leather artifacts are affected by microbial diseases in a specific condition indicator. A higher score indicates a more severe impact on that indicator.

[0137] In one implementation, a quantization threshold is an important parameter used to map the occupancy ratio to an indicator score. The quantization threshold specifies the score value corresponding to different occupancy ratio ranges.

[0138] In one implementation, when different values ​​of the state indicator correspond to different quantization thresholds, the formula for calculating the indicator score can be: in, These are the boundary data for the threshold values ​​of each state indicator. These are the boundary values ​​of the quantization threshold corresponding to the values ​​of each state indicator. This represents the maximum value of this status indicator in the database. The score of the indicator is calculated for this status indicator.

[0139] Step 227: Calculate the weighted sum of the scores of each indicator according to the preset weights of each status indicator to obtain the quantitative score of the microbial disease type.

[0140] For example, the computer device first assigns a preset weight to each status indicator. The computer device can then use this preset weight to perform a weighted summation of the scores for each status indicator to obtain a quantitative score for that type of microbial disease.

[0141] In one implementation, the quantization score is the score of each indicator calculated in step 226 above after being quantized to a preset range.

[0142] In one implementation method, the preset weights are determined based on factors such as the criticality and sensitivity of each state indicator to the impact of microbial diseases on leather artifacts.

[0143] Optionally, the computer equipment can use expert experience to determine the preset weights of multiple status indicators included in each type of microbial disease. Alternatively, the computer equipment can use the analytic hierarchy process (AHP) to construct a hierarchical model, compare each status indicator pairwise, and calculate the relative importance weight of each indicator. Alternatively, the computer equipment can also use principal component analysis to automatically determine the weights of each status indicator based on the correlation between each status indicator and the severity of the microbial disease in historical data.

[0144] In one implementation, the computer device can calculate the state index scores of multiple state indicators separately, and then perform a weighted fusion of these multiple state index scores to obtain the final quantitative score. The fusion formula can be: in, For the first The weighting coefficients of each state indicator. For the first The status indicator score for each status indicator. This represents the total number of status indicators that need to be obtained for this type of disease. To quantify the score.

[0145] In this example, by calculating the index scores of multiple status indicators corresponding to each type of microbial disease, and then merging the scores of multiple indicators to obtain a quantitative score, the effect of accurately quantifying the impact of microbial diseases on leather artifacts is achieved.

[0146] In one example, in step S1023 above, based on the quantitative scores of each type of microbial disease, the comprehensive risk level of the leather artifact being deteriorated by microbial diseases at a given time is calculated, including: Step 231: Based on the mapping relationship between the quantitative threshold and the risk level, determine the risk level of each microbial disease type according to the quantitative score of each microbial disease type.

[0147] For example, the computer device stores a pre-defined mapping table between quantitative thresholds and risk levels. This table specifies in detail the risk levels corresponding to different quantitative score ranges.

[0148] The computer equipment can compare the calculated quantitative scores of each microbial disease type with the quantitative score intervals in the mapping table based on this mapping relationship. By precisely searching, the computer equipment determines the specific interval in which each quantitative score falls, and thus determines the risk level of the microbial disease type corresponding to that quantitative score.

[0149] In one implementation, the risk level is a classification representing the severity of the impact of a certain microbial disease on leather artifacts. For example, it could include four levels from 0 to III. Alternatively, it could include three levels: low risk, medium risk, and high risk.

[0150] In one implementation, the range of the quantization score can be a fixed range; that is, the quantization threshold can be a fixed value. Optionally, when multiple sets of quantization thresholds are included, these multiple sets of quantization thresholds can have the same maximum range. For example, the minimum value among these multiple sets of quantization thresholds can be 0, and the maximum value can be 100.

[0151] In one implementation, the mapping table can be a single table applicable to all types of microbial diseases. For example, when there are preset risk levels 0–III, corresponding to a quantitative score range of [0, 100], the score ranges corresponding to these four risk levels can be [0, 25], [26, 50], [51, 75], [76, 100]. Similarly, when there are preset three risk levels, corresponding to a quantitative score range of [0, 100], the score ranges corresponding to these three risk levels can be [0, 60], [60, 90], [90, 100].

[0152] Step 232: Weighted summation of the risk levels of each type of microbial disease to obtain the comprehensive risk level of the leather artifact being deteriorated by microbial diseases at a given time.

[0153] For example, the computer device may have preset weights for each type of microbial disease. These preset weights reflect the relative importance of different types of microbial diseases to the deterioration of leather artifacts. Based on the preset weights for each type of microbial disease, the computer device can perform a weighted summation of the risk levels of the microbial disease types to obtain the final comprehensive risk level.

[0154] In one implementation, the preset weights are determined comprehensively based on factors such as the impact mechanism, frequency of occurrence, and severity of damage of each type of microbial disease on the deterioration of leather artifacts. Optionally, the computer equipment can determine the preset weights corresponding to each type of microbial disease based on expert experience information. Alternatively, the computer equipment can use the analytic hierarchy process (AHP) to construct a hierarchical model, compare each type of microbial disease pairwise, and calculate the relative importance weight of each type. Alternatively, the computer equipment can automatically determine the weights of each type of microbial disease by analyzing the contribution of different types of microbial diseases to the deterioration of leather artifacts in historical data.

[0155] In one implementation, the computer equipment can calculate a comprehensive risk level based on the risk level of each disease type, using a weighted average. The calculation formula can be: in, For the first The weighting coefficients for each disease type. For the first Risk levels for each disease type. This represents the total number of disease types. Indicates rounding down. The overall risk level is determined by the risk assessment criteria.

[0156] In this example, by calculating the risk level of each type of microbial disease and weighting and fusing the risk levels to obtain a comprehensive risk level, the method of accurately assessing the comprehensive risk level of leather artifacts deteriorating due to microbial diseases at a specific time is achieved.

[0157] In one example, the computer device can also weight and fuse the quantitative scores corresponding to multiple disease types to obtain a final comprehensive quantitative score. The calculation formula can be: in, For the first The weighting coefficients for each disease type. For the first Quantitative scores for each disease type. This represents the total number of disease types. This is a comprehensive quantitative score.

[0158] The computer device can map the comprehensive quantitative score to the corresponding comprehensive risk level. In one implementation, each comprehensive risk level can correspond to a score range of comprehensive quantitative scores. For example, when there are preset comprehensive risk levels 0–Ⅲ, and the comprehensive quantitative score range corresponding to the four comprehensive risk levels is [0, 25], [26, 50], [51, 75], [76, 100].

[0159] In this example, by first calculating the comprehensive quantitative score and then calculating the comprehensive risk level, the effect of accurately assessing the comprehensive risk level of leather artifacts being deteriorated by microbial diseases at a specific time is achieved.

[0160] In one example, in step S103 above, the temporal state data and comprehensive risk level of each leather artifact are input into the leather artifact deterioration risk assessment model to obtain a trained leather artifact deterioration risk assessment model, including: S1031. Using a preset sliding window, obtain the input data for the current iteration from the temporal state data of each leather artifact.

[0161] For example, a computer device has a pre-defined sliding window. This sliding window has two parameters: window size and sliding step size. The window size setting limits the amount of time-series state data acquired in each iteration. The sliding step size setting limits the distance the sliding window moves between two iterations.

[0162] The computer device moves a sliding window sequentially across the temporal state data sequence of the leather artifacts after each iteration, according to a set sliding step size. After each move, the data segment covered by the window becomes the input data for the current iteration.

[0163] In one implementation, the temporal state data refers to data recorded at various points in time to describe the state of the leather artifact. For example, this state data may include information such as leather temperature, humidity, pH level, and the number of microbial colonies. Each piece of state data reflects the state of the leather artifact at a given time.

[0164] In one implementation, the computer device can use a sliding window to obtain the input data for the current training from the time-series state data of each leather artifact.

[0165] In one implementation, the window size of the preset sliding window can be determined by the size of the input data of the leather artifact deterioration risk assessment model.

[0166] In one implementation, the sliding step size of the preset sliding window can be determined empirically.

[0167] In one implementation, the computer device can select multiple leather artifacts from the database before acquiring data in each iteration, and use a sliding window to select a portion of continuous state data from the temporal state data of the multiple leather artifacts as input data for that iteration.

[0168] In one implementation, the computer device first processes the state data in the database to generate a training dataset. Each training data point in this dataset may include a sample data point and its label data. The sample data point consists of a set of state data for a leather artifact over a certain period of time. The label data consists of the comprehensive risk level of the leather artifact across all its state data points over that period of time.

[0169] S1032. Use the leather artifact deterioration risk assessment model to process the temporal state data and predict the predicted risk level for the next moment.

[0170] For example, the computer device inputs the acquired input data of the current iteration into a pre-built leather artifact deterioration risk assessment model. Each time, the computer device can input a series of state data for a leather artifact into the model. The model can analyze the data relationships between these state data and output the predicted risk level for each state data point. This predicted risk level is the comprehensive risk level for the next time step predicted based on each state data point.

[0171] In one implementation, the leather artifact deterioration risk assessment model is a model that can estimate and predict the risk level of a leather artifact's susceptibility to microbial diseases in the future based on its current state data. The predicted risk level output by this leather artifact deterioration risk assessment model corresponds to the comprehensive risk level.

[0172] In one implementation approach, the risk assessment model for the deterioration of leather artifacts can typically be constructed using machine learning or deep learning models. Examples include decision trees, support vector machines, and random forests. Deep learning models such as neural networks, recurrent neural networks, long short-term memory networks, and gated recurrent units are also suitable.

[0173] In one implementation, the leather artifact deterioration risk assessment model is constructed using an LSTM and Transformer architecture, combining the advantages of local temporal correlation and global attention mechanisms. This includes an input embedding layer, an LSTM layer, a Transformer layer, and a fully connected output layer.

[0174] In one implementation, the input embedding layer is used to input state data from a time series. Optionally, the input data used by the input embedding layer can be feature-processed state data.

[0175] In one implementation, the LSTM layer is used to capture local temporal dependencies.

[0176] In one implementation, the Transformer layer is used to capture long-term dependencies through a self-attention mechanism.

[0177] In one implementation, a fully connected layer is used to output the overall risk level for the next T periods. The overall risk level for these T periods can form a risk trend sequence.

[0178] In one implementation, the leather artifact deterioration risk assessment model can learn the correlation between multiple state data points over time by training on a large amount of leather artifact state data with known risk levels. Furthermore, this model can also determine the mapping relationship between state data and comprehensive risk levels.

[0179] S1033. Based on the comprehensive risk level and predicted risk level at the next moment, the loss value of the current iteration is calculated using the loss function of the leather cultural relic deterioration risk assessment model.

[0180] For example, the computer device can obtain the comprehensive risk level for the next time step corresponding to each state data point contained in the input data. The computer device can then use the comprehensive risk level for the next time step and the predicted risk level for the next time step as inputs, substituting them into a predefined loss function of the leather artifact deterioration risk assessment model. Based on the preset loss function, the computer device can calculate a numerical value according to the degree of difference between the comprehensive risk level and the predicted risk level. This numerical value is the loss value for the current iteration.

[0181] In one implementation, the loss value reflects the magnitude of the deviation between the model's prediction and the actual situation. The smaller the loss value, the more accurate the model's prediction.

[0182] In one implementation, the loss function can be a mean squared error loss function, a cross-entropy loss function, or the like.

[0183] In one implementation, the loss function of the leather artifact deterioration risk assessment model is the weighted mean-square error (MSE). Its formula is: Where N is the number of state data inputs in the current iteration. The first in the current input data Individual status data. These are weighting coefficients. To predict the overall risk level for the next moment, This represents the overall risk level for the next time step corresponding to this state data.

[0184] S1034. Based on the loss value, optimize the training of the leather cultural relic deterioration risk assessment model.

[0185] For example, the computer device uses the calculated loss value of the current iteration as feedback information and adjusts the parameters of the leather artifact deterioration risk assessment model using an optimization algorithm.

[0186] In one implementation, the optimization algorithm determines the direction and magnitude of model parameter adjustment based on the magnitude and trend of the loss value, so that the model can gradually reduce the loss value and improve the accuracy of prediction in subsequent iterative training.

[0187] In one implementation, by repeating steps S1031 to S1034 above, the computer device can train and iteratively optimize the risk assessment model for the deterioration of cultural relics.

[0188] In one implementation, the computer device can continuously optimize the model's parameters through iterative training until a preset training stopping condition is reached. Optionally, the stopping condition of the model can be that the loss value is less than a certain threshold, the number of iterations reaches a maximum value, etc.

[0189] In one implementation, the optimization algorithm is an algorithm used to adjust the model parameters to minimize the loss function. For example, the optimization algorithm can be stochastic gradient descent, batch gradient descent, mini-batch gradient descent, Adam algorithm, etc.

[0190] In this example, the model for assessing the deterioration risk of leather artifacts is trained using time-series state data. The model loss value is calculated using the predicted risk level and comprehensive risk level output by the model. Based on the loss value, the model is optimized to improve training efficiency and accuracy.

[0191] In one example, the computer device can also analyze the output predicted risk level after training the leather artifact deterioration risk assessment model. The method also includes: S105. If the predicted risk level is greater than the preset threshold, simulate the state data of the moment when the predicted risk level is greater than the preset threshold based on the state data of the leather artifacts.

[0192] For example, the computer device first performs a comprehensive check of the predicted risk levels of each leather artifact. Once a predicted risk level is found to exceed a preset threshold, the computer device can determine that the leather artifact poses a risk. Furthermore, the computer device can record the moment when the predicted risk level exceeds the preset threshold.

[0193] At this point, the computer equipment uses a specific simulation algorithm to simulate the state data of the leather artifact and obtain simulated state data at the moment when the predicted risk level exceeds a pre-set threshold.

[0194] In one implementation, the preset threshold is a level threshold determined comprehensively based on multiple factors such as the preservation requirements of leather artifacts, their historical deterioration, and expert experience. When the predicted risk level is higher than this preset threshold, the leather artifact can be determined to be in a high-risk deterioration state.

[0195] In one implementation, a computer device can generate simulated state data based on the state data of the leather artifact using an established mathematical model.

[0196] In another implementation, computer equipment can use a trained machine learning model to simulate and generate simulated state data that predicts when the risk level exceeds a pre-set threshold.

[0197] S106. Input the simulated state data into the restoration suggestion model to generate restoration suggestions for the leather artifacts. The restoration suggestion model is a large language model generated based on the state data and restoration strategies of leather artifacts from historical restoration cases.

[0198] For example, the computer device can organize and convert the simulated state data of the leather artifact obtained in the previous steps when the predicted risk level exceeds a preset threshold, according to the input format required by the restoration suggestion model. Then, the computer device can input this organized simulated state data into the pre-trained restoration suggestion model. After receiving the input data, the restoration suggestion model uses the mapping relationship between the state data of leather artifacts and restoration strategies in historical restoration cases that it has learned internally to analyze and process the currently input simulated state data, and finally generate restoration suggestions suitable for the leather artifact.

[0199] In one implementation, the status data of leather artifacts in historical restoration cases records various characteristic information of leather artifacts under different degrees of deterioration.

[0200] In one implementation, the repair strategy includes specific repair measures and methods to be taken for different states.

[0201] In one implementation, the large language model is a natural language processing model based on deep learning. It learns rich language knowledge and semantic understanding capabilities by pre-training on massive amounts of text data.

[0202] In one implementation, when constructing the restoration suggestion model, the state data and restoration strategies from historical restoration cases are used as training data, allowing the large language model to learn how to generate reasonable restoration suggestions based on the state of the leather artifacts.

[0203] In this example, by comparing the predicted risk level with a preset threshold, the moment when the predicted risk level exceeds the preset threshold is determined, and the simulated state data at that moment is obtained. This data is then input into a large language restoration suggestion model built based on historical cases to obtain modification suggestions. This method aims to generate accurate restoration suggestions for leather artifacts in the early stages when abnormalities occur.

[0204] In one example, after predicting the risk level of the leather artifact at multiple future times in step S104 above, the computer equipment can also generate a risk trend diagram, including: S105. Based on the risk assessment model for the deterioration of leather cultural relics, output the predicted risk levels at multiple future time points and form a risk trend.

[0205] For example, after obtaining the leather artifact deterioration risk assessment model, the computer device can input multiple temporal state data of leather artifacts from the basic database into the model. This model can then output predicted risk levels for multiple future time points based on a preset quantity.

[0206] Computer equipment can generate risk trend charts based on the predicted risk levels at multiple future points in time.

[0207] In one implementation, the risk trend chart can be as follows: Figure 2 As shown in the diagram. Trend changes can be represented by color gradients and arrow directions.

[0208] In this example, multiple predicted risk levels are obtained by forecasting, and a risk trend chart is drawn accordingly to visually demonstrate the future trend of the deterioration risk of leather cultural relics.

[0209] In one example, the computer device can also generate prevention and restoration strategies for leather artifacts based on a large AI model. These strategies specifically include: S106. The computer equipment can access a strategy knowledge database composed of human expert knowledge and historical restoration cases. This strategy knowledge database includes effective inhibition methods for different types of microorganisms, suitable restoration materials, and recommendations for protective environmental parameters.

[0210] For example, disease characteristics may include parameters such as localized punctate colonies and microbial damage area between 2% and 30%. Protection methods may include placing the affected area in a BioQuell isolation chamber (ozone concentration 0.1ppm), cleaning with ethanol / water (7:3) swabs, and applying a nano-silver solution (particle size 20nm) locally.

[0211] S107. Computer equipment can use basic attributes of cultural relics such as age, material, and tanning type, environmental conditions such as temperature, humidity, and pH, risk level results, and future trend data as input features.

[0212] S108. Invoke the AI ​​large model, set the corresponding user prompt and system prompt, generate preventive strategy text based on the large language model inference, and perform structured output in combination with rule constraints. The computer device can automatically generate preventive protection strategies and remedial protection strategies corresponding to the risk level.

[0213] S109. The computer device can display the output strategy in a visual interface for users to manually confirm and correct.

[0214] S110. Computer equipment can automatically incorporate the correction results into the database and use them for continuous learning and optimization of large language models.

[0215] In this example, a strategy knowledge database containing human expert knowledge and historical restoration cases is constructed. Using data such as the basic attributes of cultural relics, environmental conditions, and risk levels as input features, an AI model is invoked and corresponding prompt words are set to generate prevention and restoration strategy text. After being structured and constrained by rules, and confirmed and corrected by humans, the correction results are incorporated into the database for continuous learning and optimization of the model. This achieves the effect of generating scientific and dynamically optimizable prevention and restoration strategies for leather cultural relics.

[0216] Figure 3 This is a schematic diagram of the structure of a microbial disease deterioration prediction system for leather cultural relics provided in an embodiment of this application. Figure 1 and Figure 2 Based on the illustrated embodiments, as Figure 3 As shown, the microbial disease deterioration prediction system for leather cultural relics can include three parts: database management, diagnostic analysis, and prevention and modification strategies.

[0217] The database management system includes a statistical analysis module, a basic database, and a simulation database. The statistical analysis module primarily collects information on leather artifacts from multiple aspects, such as region, age, material, and tanning agents, and writes this information into the basic database. The basic artifact database records basic information and attribute descriptions of leather artifacts collected at various points in time. These basic attributes include the artifact's microstructure, chemical composition, physical properties, structural analysis, and environmental factors. All data in this basic database for each leather artifact at a given moment constitutes the state data for that leather artifact. The simulation database manages the basic data input from simulation experiments. Data in both the basic and simulation databases are used for training subsequent models to predict the microbial deterioration of leather artifacts.

[0218] The diagnostic analysis section implements the risk assessment algorithm. This diagnostic section can extract key quantitative indicators of damage to leather artifacts from multimodal state data, and uses an adaptive weighted feature fusion strategy to optimize the fusion of the quantified state indicators. This achieves a multi-indicator, multi-level risk assessment mechanism, improving the efficiency and accuracy of quantitative analysis of artifact damage and providing a scientific basis for restoration and protection strategies.

[0219] The prevention and restoration strategy section primarily utilizes AI-assisted models to generate conservation strategies for leather artifacts. Based on the condition of leather damage and deterioration, this AI-driven section, powered by large language model technology, provides clear, scientifically sound, and highly operable restoration and preventative conservation strategies for leather artifacts. It takes artifact damage observation data, historical restoration records, and standardized conservation guidelines as input, and employs automated information matching and natural language processing techniques to provide artifact management personnel with well-structured, scientifically sound, and highly operable restoration and preventative conservation strategies.

[0220] In this embodiment, by constructing a system that includes database management, diagnostic analysis, and prevention and repair strategies, a system for predicting the deterioration of leather cultural relics due to microbial diseases is established, thereby achieving the effects of comprehensive data management, accurate assessment of disease risks, and scientific formulation of repair and protection strategies.

[0221] In one example, the database design described above includes the following specific details: This database stores the status data of leather artifacts. This status data can be stored in an artifact information table. The artifact information table can include fields such as artifact number, name, material, age, and status. It can also include artifact classification tables, artifact image tables, etc.

[0222] In one implementation, the database can be a MySQL database.

[0223] In one implementation, the computer device can provide an interface for adding, deleting, modifying, and querying information on a cultural relic information table. The computer device can query the status data of one or more leather cultural relics. The computer device can also filter queries based on specific conditions.

[0224] In one implementation, computer equipment can be configured with a cultural relic classification management module to enable the addition, deletion, modification, and retrieval of cultural relic classifications.

[0225] In one implementation, the computer device can design a cultural relic image management module to upload and delete cultural relic images. Furthermore, the computer device can use this cultural relic image management module to associate the cultural relic image with its status information.

[0226] In one implementation, the computer device can return the result of database processing by encapsulating a generic response result model. When the result indicates an exception in database processing, the computer device can also generate an exception message.

[0227] In one implementation, computer devices can achieve single sign-on by accessing an authentication system.

[0228] In one implementation, computer equipment can monitor the use of the microbial disease deterioration prediction system for the leather artifact by adding functions such as logging and monitoring.

[0229] In one implementation, computer devices can improve the performance of the microbial disease deterioration prediction system for leather artifacts by using Redis for caching.

[0230] In this example, by setting up a database, cultural relic data can be used to participate in the training and validation of deep learning models, supporting continuous system optimization and dynamic updates, and improving the reliability of prediction and strategy generation.

[0231] In one example, in the diagnostic analysis section described above, the computer device can determine the overall risk level through degradation diagnosis by fusing multimodal state data. Specifically, this may include: S211. Computer equipment can integrate microbial damage-related characteristic data of the same category from the basic database of cultural relics according to different leather types and materials, so as to realize the integration of multimodal state data.

[0232] S212. Computer equipment can collect the criteria for judging the level of leather damage from experts, summarize the parameter thresholds of key indicators under different degrees of deterioration, and integrate expert experience information.

[0233] S213. Computer equipment can be used to train a model for predicting the deterioration of leather artifacts by microbial diseases by combining state data from a simulated aging database.

[0234] S214. The computer equipment can determine the required state indicators for calculation based on different types of microbial diseases. For a single state indicator, the computer equipment can output a quantitative score and risk level for that microbial disease type based on expert experience and quantitative standards. For multiple state indicators, the computer equipment can output the corresponding quantitative score and risk level using an adaptive weighted fusion method based on expert experience and quantitative standards.

[0235] In this example, the fusion diagnosis of multimodal state data enables a comprehensive assessment of single and multi-indicator diseases, accurately identifies leather deterioration, and improves the accuracy and scientific nature of cultural relic deterioration diagnosis.

[0236] In one example, the computer device can also generate a predicted risk level for leather artifacts based on deep learning-based risk trend prediction, including: S221. Computer equipment can generate sequential data in chronological order based on state data in a database. This state data may include information such as historical microbial characteristics, chemical spectral characteristics, microscopic image characteristics, and environmental parameters. This chronologically ordered sequential data can be referred to as historical data sequence construction.

[0237] S222. Computer equipment can construct deep sequence prediction models based on LSTM and Transformer. The computer equipment can input historical sequences into the network and learn the relationship between features and risk indices over time. This model is a prediction model for the microbial disease deterioration of leather artifacts.

[0238] S223. Computer equipment can combine simulated aging test data and cultural relic historical databases to optimize model parameters and structure, ensure the accuracy of future risk prediction, and obtain the final prediction model for microbial disease deterioration of leather cultural relics.

[0239] S224. The computer equipment can output a comprehensive quantitative score and comprehensive risk level of microbial deterioration at multiple future time points through the microbial disease deterioration prediction model for leather cultural relics. The computer equipment can generate a future deterioration trend chart based on the comprehensive quantitative score and comprehensive risk level.

[0240] In this example, the construction and training of a microbial disease deterioration prediction model for leather cultural relics based on LSTM and Transformer enable the prediction of future deterioration trends of cultural relics, realize forward-looking risk prediction capabilities, provide a quantitative risk index, and provide a basis for preventive protection decisions.

[0241] In one example, the aforementioned prevention strategy can be generated using AI-assisted prevention and remediation strategies based on a large language model. The specific steps are as follows: S231. Computer equipment can generate a strategy database corresponding to the level of microbial disease based on expert advice for prevention and remediation strategies.

[0242] S232. Computer equipment can use risk assessment results, future trends, leather material, disease indicators and environmental parameters as model inputs to provide specific prevention and remediation strategies through a large AI model.

[0243] S233. Computer equipment can generate personalized prevention and remediation plans for single or multiple indicator risks through large AI models, and prioritize them based on risk level and resource constraints.

[0244] S234. The computer equipment can generate structured strategy reports. These reports may include risk levels, recommended measures, and implementation priorities. Furthermore, the computer equipment can dynamically adjust strategies based on actual monitoring data.

[0245] This example combines AI big data models with expert experience to output personalized and dynamically optimized prevention and restoration strategies, enabling intelligent generation of protection strategies and improving the pertinence and feasibility of protection measures. Furthermore, a real-time visualization interface intuitively presents diagnostic results, risk indices, and future trends, helping managers quickly understand the status of cultural relics and protection priorities.

[0246] Figure 4 This is a schematic diagram of the structure of a microbial disease deterioration prediction system for leather cultural relics provided in an embodiment of this application. Figures 1 to 3 Based on the illustrated embodiments, as Figure 4 As shown, the microbial disease deterioration prediction system for leather cultural relics can realize a multi-layer layout of data acquisition layer, data storage layer, data analysis layer, service layer and application layer.

[0247] The data acquisition layer can acquire cultural relic data by collecting various types of data, such as text data, images, and spectral data. This layer can collect both cultural relic data and simulation data. It can also collect information such as material type, tanning type, microstructure, chemical composition, and physical properties. Furthermore, it can collect structural parameters, degradation parameters, disease types, degradation products, degradation characteristics, and disease parameters.

[0248] The data storage layer can be used to store various types of data, including structured data, unstructured data, semi-structured data, hierarchical data, data tags, pattern classification, and unified data. This data storage layer can be divided into two parts: a basic cultural relics database and a simulated sample database.

[0249] The data analysis layer can include data processing such as text and image feature extraction, multimodal data fusion, and weighted adaptive adjustment.

[0250] The service layer can include two parts: data services and system services. Data services are used to implement data management, indicator analysis, call comparison, and status identification. System services can include standard interfaces, model management, unified indicators, and tag management.

[0251] The application layer can be used to realize statistical analysis of cultural relics, deterioration assessment, trend prediction, and generation of prevention and protection strategies.

[0252] Figure 5 A structural diagram of a device for predicting microbial disease deterioration of leather artifacts provided in this application embodiment is shown below. Figure 5 As shown, the leather artifact microbial disease deterioration prediction device 500 includes: The acquisition module 501 is used to acquire the temporal status data of each leather artifact.

[0253] Training module 502 is used to process the state data of each leather artifact at each time step based on preset parameter thresholds for each type of microbial disease, and to determine the comprehensive risk level of microbial disease deterioration of each leather artifact at each time step. The time-series state data and comprehensive risk level of each leather artifact are input into the leather artifact deterioration risk assessment model to obtain the trained leather artifact deterioration risk assessment model.

[0254] The prediction module 503 is used to process the temporal state data of leather artifacts using a trained leather artifact deterioration risk assessment model, and to predict the predicted risk level of the leather artifacts at multiple future moments.

[0255] In one example, the state data includes multiple state metrics. Training module 502 is used for: Obtain the preset status indicators for each type of microbial disease from the status data.

[0256] By using preset parameter thresholds for each type of microbial disease, the status indicators are processed to obtain a quantitative score for each type of microbial disease.

[0257] Based on the quantitative scores of each type of microbial disease, the comprehensive risk level of leather artifacts deteriorating due to microbial diseases at any given time is calculated.

[0258] In one example, the parameter thresholds include an index threshold and a quantification threshold. When a predefined state index is required for a microbial disease type, the training module 502 is used to: Based on the values ​​of the status indicators, determine the corresponding indicator thresholds and quantization thresholds in the parameter thresholds.

[0259] Determine the proportion of status indicators in the indicator thresholds.

[0260] Based on the occupancy ratio and quantification threshold, a quantitative score for the type of microbial disease is obtained.

[0261] In one example, the parameter threshold includes an index threshold and a quantification threshold. When multiple state indicators are preset for a microbial disease type, the training module 502 is used for: Based on each state indicator and its value, determine the indicator threshold and quantization threshold corresponding to the state indicator in the parameter threshold.

[0262] Determine the proportion of each status indicator within the corresponding indicator threshold.

[0263] Based on the occupancy ratio and the corresponding quantitative threshold of the status indicator, the indicator score corresponding to the status indicator is obtained.

[0264] Based on the preset weights of each status indicator, the weighted sum of the scores of each indicator is calculated to obtain the quantitative score of the microbial disease type.

[0265] In one example, training module 502 is used for: Based on the mapping relationship between quantitative thresholds and risk levels, the risk level of each microbial disease type is determined according to the quantitative score of each type.

[0266] By weighted summing of the risk levels of each type of microbial disease, the comprehensive risk level of leather artifacts deteriorating due to microbial diseases at any given time is obtained.

[0267] In one example, prediction module 503 is used for: Using a preset sliding window, the input data for the current iteration is obtained from the temporal state data of each leather artifact.

[0268] The leather artifact deterioration risk assessment model is used to process the time-series state data and predict the risk level for the next moment.

[0269] Based on the comprehensive risk level and predicted risk level at the next moment, the loss value for the current iteration is calculated using the loss function of the leather cultural relic deterioration risk assessment model.

[0270] Based on the loss value, optimize the training of the leather artifact deterioration risk assessment model.

[0271] In one example, prediction module 503 is used for: If the predicted risk level is greater than the preset threshold, simulated state data for the moment when the predicted risk level is greater than the preset threshold is generated based on the state data of the leather artifacts.

[0272] The simulated state data is input into the restoration suggestion model to generate restoration suggestions for leather artifacts. The restoration suggestion model is a large language model generated based on the state data and restoration strategies of leather artifacts from historical restoration cases.

[0273] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0274] In this embodiment, the microbial disease deterioration prediction device for leather artifacts is presented in the form of a functional unit. Here, a unit refers to an application-specific integrated circuit (ASIC), a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0275] Figure 6 A structural diagram of a computer device provided in an embodiment of this application, such as... Figure 6 As shown, the computer device 600 includes one or more processors 601, memory 602, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 6 Take the 601 processor as an example.

[0276] Processor 601 may be a central processing unit, a network processor, or a combination thereof. Processor 601 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0277] The memory 602 stores instructions executable by at least one processor 601 to cause the at least one processor 601 to perform the method shown in the above embodiments.

[0278] Memory 602 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, memory 602 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, memory 602 may optionally include memory remotely located relative to processor 601, and this remote memory may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0279] Memory 602 may include volatile memory, such as random access memory; memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; memory 602 may also include combinations of the above types of memory.

[0280] The computer device also includes a communication interface 603 for communicating with other devices or communication networks.

[0281] This application also provides a computer-readable storage medium. The methods described in this application can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code downloaded over a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and subsequently stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the methods shown in the above embodiments are implemented.

[0282] This application provides a computer program product including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the method of any embodiment of this application.

[0283] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and all such modifications and variations fall within the scope defined by the appended claims.

[0284] The apparatus, module, or unit described in the above embodiments can be implemented by a computer chip or entity, or by a product having a certain function. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0285] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0286] Those skilled in the art will understand that embodiments of this application can be provided as methods, computer devices, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0287] This application is described with reference to flowchart illustrations and / or block diagrams of methods, computer devices, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0288] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0289] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0290] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0291] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the computer device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0292] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

[0293] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for predicting the deterioration of leather artifacts due to microbial diseases, characterized in that, The method includes: Obtain the temporal status data of each leather artifact; Based on the preset parameter thresholds for each type of microbial disease, the state data of each leather artifact at each time point are processed to determine the comprehensive risk level of each leather artifact being deteriorated by microbial diseases at each time point. The time-series state data of each leather artifact and the comprehensive risk level are input into the leather artifact deterioration risk assessment model to obtain the trained leather artifact deterioration risk assessment model. The trained leather artifact deterioration risk assessment model is used to process the time-series state data of the leather artifact to predict the predicted risk level of the leather artifact at multiple future moments.

2. The method according to claim 1, characterized in that, The status data includes multiple status indicators; based on preset parameter thresholds for each type of microbial disease, the status data of each leather artifact at each time point is processed to determine the comprehensive risk level of microbial disease deterioration of each leather artifact at each time point, including: Obtain the preset status indicators for each type of microbial disease from the status data; The status indicators are processed using preset parameter thresholds for each type of microbial disease to obtain a quantitative score for the type of microbial disease. Based on the quantitative scores of each of the aforementioned microbial disease types, the comprehensive risk level of the leather artifact being deteriorated by microbial diseases at that time is calculated.

3. The method according to claim 2, characterized in that, The parameter thresholds include index thresholds and quantification thresholds; when a desired state index is preset for a microbial disease type, the preset parameter thresholds for each microbial disease type are used to process the state index to obtain a quantification score for the microbial disease type, including: Based on the value of the state index, determine the corresponding index threshold and quantization threshold in the parameter threshold; Determine the proportion of the status indicator within the indicator threshold; The quantitative score of the microbial disease type is obtained based on the occupancy ratio and the quantification threshold.

4. The method according to claim 2, characterized in that, The parameter thresholds include indicator thresholds and quantification thresholds; when multiple state indicators are preset for a microbial disease type, the preset parameter thresholds for each microbial disease type are used to process the state indicators to obtain a quantification score for the microbial disease type, including: Based on each state indicator and its value, determine the indicator threshold and quantization threshold corresponding to the state indicator in the parameter threshold; Determine the proportion of each of the aforementioned status indicators in the threshold values ​​corresponding to the status indicators; Based on the occupancy ratio and the quantification threshold corresponding to the status indicator, the indicator score corresponding to the status indicator is obtained; Based on the preset weights of each status indicator, the weighted sum of the scores of each indicator is calculated to obtain the quantitative score of the microbial disease type.

5. The method according to claim 2, characterized in that, Based on the quantitative scores of each of the aforementioned microbial disease types, the comprehensive risk level of the leather artifact's deterioration due to microbial diseases at that time is calculated, including: Based on the mapping relationship between quantitative thresholds and risk levels, the risk level of each of the aforementioned microbial disease types is determined according to the quantitative scores of each type. The risk levels of each of the aforementioned microbial disease types are weighted and summed to obtain the comprehensive risk level of the leather artifact being deteriorated by microbial diseases at the stated time.

6. The method according to any one of claims 1-5, characterized in that, The time-series state data and comprehensive risk level of each leather artifact are input into the leather artifact deterioration risk assessment model to obtain a trained leather artifact deterioration risk assessment model, including: Using a preset sliding window, the input data for the current iteration is obtained from the state data of each leather artifact in time sequence; The aforementioned leather artifact deterioration risk assessment model is used to process the time-series state data to predict the predicted risk level for the next moment. Based on the comprehensive risk level at the next moment and the predicted risk level, the loss value of the current iteration is calculated using the loss function of the leather cultural relic deterioration risk assessment model. Based on the loss value, the leather artifact deterioration risk assessment model is optimized and trained.

7. The method according to any one of claims 1-5, characterized in that, The method further includes: If the predicted risk level is greater than a preset threshold, simulated state data for the moment when the predicted risk level is greater than the preset threshold is generated based on the state data of the leather artifact. The simulated state data is input into the restoration suggestion model to generate restoration suggestions for the leather artifact; wherein, the restoration suggestion model is a large language model generated based on the state data and restoration strategies of leather artifacts in historical restoration cases.

8. A device for predicting the deterioration of leather artifacts due to microbial diseases, characterized in that, The device includes: The acquisition module is used to acquire the temporal status data of each leather artifact. The training module is used to process the state data of each leather artifact at each time step according to the preset parameter thresholds of each type of microbial disease, and determine the comprehensive risk level of each leather artifact being deteriorated by microbial diseases at each time step; the state data of each leather artifact in time step and the comprehensive risk level are input into the leather artifact deterioration risk assessment model to obtain the trained leather artifact deterioration risk assessment model. The prediction module is used to process the time-series state data of the leather artifact using the trained leather artifact deterioration risk assessment model, and to predict the predicted risk level of the leather artifact at multiple future moments.

9. A computer device, characterized in that, include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the method of any one of claims 1 to 7.

11. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the method of any one of claims 1 to 7.

Images