Equipment state evaluation method and device based on artificial intelligence

Abstract

The invention relates to the technical field of equipment state analysis and evaluation, and discloses an artificial intelligence-based equipment state evaluation method and device, and the method comprises the steps: obtaining the target feature information of target equipment, executing the information preprocessing operation, and obtaining the model feature information and information weight parameters of the target feature information, determining an average prediction parameter of the target feature information in combination with a stochastic regression model; performing training operation on the initial evaluation model based on the average prediction parameter to obtain a model training result, generating updated feature information based on the model training result, and performing model iteration operation on the initial evaluation model based on the updated feature information to obtain a target evaluation model; and obtaining equipment operation information and inputting the information into the target evaluation model to obtain an equipment state evaluation result. Therefore, the equipment state can be evaluated in combination with artificial intelligence, the intelligence and efficiency of evaluating the equipment state can be improved, and the accuracy and reliability of evaluating the equipment state can be improved.

Description

Technical Field

[0001] This invention relates to the field of equipment condition analysis and evaluation technology, and in particular to an artificial intelligence-based equipment condition evaluation method and apparatus. Background Technology

[0002] Currently, most medical device assessments rely on manual statistics of usage frequency and duration. The device's condition is then evaluated based on usage frequency; more frequent and longer usage times indicate fatigue, while fewer and shorter usage times suggest better health. However, this method of assessment is entirely manual, resulting in low levels of intelligence and accuracy.

[0003] Therefore, it is particularly important to provide a new method for assessing equipment condition in order to improve the accuracy and intelligence of equipment condition assessment. Summary of the Invention

[0004] This invention provides an artificial intelligence-based equipment status assessment method and apparatus, which can combine artificial intelligence to assess equipment status, thereby improving the intelligence and efficiency of equipment status assessment, as well as the accuracy and reliability of equipment status assessment.

[0005] The first aspect of this invention discloses an artificial intelligence-based device status assessment method, the method comprising: Obtain target feature information of the target device, perform information preprocessing operation on the target feature information to obtain model characteristic information corresponding to the target feature information, and information weight parameters corresponding to each model characteristic information; Based on each of the model characteristic information and the information weight parameters corresponding to each of the model characteristic information, the average prediction parameters corresponding to the target feature information are determined in combination with a pre-determined random regression model. Based on the average prediction parameters, a training operation is performed on the initial evaluation model to obtain the model training result. Updated feature information is generated based on the model training result. Based on the updated feature information, a model iteration operation is performed on the initial evaluation model to obtain the target evaluation model. Obtain the device operation information of the target device, input the device operation information into the target evaluation model, and obtain the device status evaluation result corresponding to the target device.

[0006] As an optional implementation, in the first aspect of the present invention, after inputting the device operation information into the target evaluation model to obtain the device status evaluation result corresponding to the target device, the method further includes: Based on the equipment status assessment results and the pre-determined cross-validation optimization model, determine whether the target assessment model meets the preset model assessment conditions. When it is determined that the target evaluation model does not meet the preset model evaluation conditions, the model update parameters corresponding to the target evaluation model are determined, and a model update operation is performed on the target evaluation model based on the model update parameters to obtain an updated evaluation model.

[0007] As an optional implementation, in a first aspect of the present invention, generating updated feature information based on the model training results includes: Based on the model training results, the model training results are uploaded to a predetermined target cloud to obtain model feedback information from the target cloud. Based on the model feedback information, the model parameters to be adjusted in the initial evaluation model and the parameter adjustment factor corresponding to each model parameter to be adjusted are determined. Updated feature information is generated according to each model parameter to be adjusted and the parameter adjustment factor corresponding to each model parameter to be adjusted.

[0008] As an optional implementation, in a first aspect of the present invention, the method further includes: Obtain the rated operating parameters corresponding to the target device, and determine the operating procedure corresponding to the target device based on the device status assessment result and the rated operating parameters. Obtain the real-time operating information corresponding to the target device, and determine whether the real-time operating information matches the operating program; When it is determined that the real-time operation information does not match the operation program, the safe operation range parameters of the target device are determined based on the real-time operation information and the rated operation parameters. Based on the safe operation range parameters, operation warning information and operation warning control parameters corresponding to the target device are generated. The operation warning information is transmitted to the target terminal corresponding to the target device, and the target device is controlled to perform a warning control operation that matches the warning control parameters.

[0009] As an optional implementation, in a first aspect of the present invention, determining the average prediction parameter corresponding to the target feature information based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, combined with a pre-determined random regression model, includes: Based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, a feature subset partitioning operation is performed on the target feature information to obtain at least one feature information subset. Based on each subset of the feature information, a mapping relationship is generated between each model feature information and the information weight parameter corresponding to each model feature information. Based on the mapping relationship and the pre-determined random forest regression model, the average prediction parameter corresponding to the target feature information is determined. The model characteristic information includes equipment usage frequency characteristics, equipment workload characteristics, equipment operation habit characteristics, and equipment material characteristics.

[0010] As an optional implementation, in a first aspect of the present invention, the method further includes: Based on the equipment status assessment result corresponding to the target equipment, determine whether the equipment status assessment result is used to indicate that the target equipment does not meet the preset equipment operating conditions; When it is determined that the device status assessment result indicates that the target device does not meet the preset device operating conditions, real-time environmental information of the current environment where the target device is located is obtained; Based on the equipment status assessment results, generate equipment status early warning information for the target equipment, and generate target status early warning information for the target equipment based on the equipment status early warning information and the real-time environmental information, and determine the early warning method parameters corresponding to the target equipment based on the real-time environmental information. Based on the target status warning information and the warning method parameters, device warning parameters for the target device are generated, and the device warning parameters are transmitted to the warning receiving terminal corresponding to the target device.

[0011] As an optional implementation, in a first aspect of the present invention, the step of performing information preprocessing on the target feature information to obtain model characteristic information corresponding to the target feature information and information weight parameters corresponding to each of the model characteristic information includes: An information cleaning operation is performed on the target feature information to obtain a feature information cleaning result, and an information classification operation is performed on the feature information cleaning result to obtain at least one information category, wherein the information category includes a numerical information category and a label information category; Based on the information category, a target information category is determined, and model characteristic information matching the target information category is determined from all the feature information cleaning results included in the target information category. Local correlation information corresponding to the model characteristic information is extracted based on a pre-determined convolutional neural network. Based on the local correlation information and the model characteristic information, the feature temporal transformation correlation relationship is determined. Based on the correlation of the feature time-series transformation, determine the information weight parameter corresponding to each of the model characteristic information; The feature temporal transformation correlation relationship includes feature dynamic correlation relationship, temporal attenuation coefficient relationship, and material compatibility correlation relationship.

[0012] A second aspect of the present invention discloses an artificial intelligence-based device for assessing device status, the device comprising: The acquisition module is used to acquire target feature information of the target device; The processing module is used to perform information preprocessing operations on the target feature information to obtain model characteristic information corresponding to the target feature information, and information weight parameters corresponding to each model characteristic information; The determination module is used to determine the average prediction parameter corresponding to the target feature information based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, combined with a pre-determined random regression model. The training module is used to perform training operations on the initial evaluation model based on the average prediction parameters to obtain the model training results; The generation module is used to generate updated feature information based on the model training results; An iteration module is used to perform model iteration operations on the initial evaluation model based on the updated feature information to obtain the target evaluation model; The acquisition module is also used to acquire the device operation information of the target device; The input module is used to input the device operation information into the target evaluation model to obtain the device status evaluation result corresponding to the target device.

[0013] As an optional implementation, in a second aspect of the invention, the apparatus further includes: The judgment module is used to determine whether the target evaluation model meets the preset model evaluation conditions based on the device status evaluation results and the device status evaluation results, after the input module inputs the device operation information into the target evaluation model and obtains the device status evaluation results corresponding to the target device; The determining module is further configured to determine the model update parameters corresponding to the target evaluation model when the judging module determines that the target evaluation model does not meet the preset model evaluation conditions; The update module is also used to perform a model update operation on the target evaluation model based on the model update parameters to obtain an updated evaluation model.

[0014] As an optional implementation, in a second aspect of the present invention, the specific method by which the generation module generates updated feature information based on the model training results includes: Based on the model training results, the model training results are uploaded to a predetermined target cloud to obtain model feedback information from the target cloud. Based on the model feedback information, the model parameters to be adjusted in the initial evaluation model and the parameter adjustment factor corresponding to each model parameter to be adjusted are determined. Updated feature information is generated according to each model parameter to be adjusted and the parameter adjustment factor corresponding to each model parameter to be adjusted.

[0015] As an optional implementation, in a second aspect of the present invention, the acquisition module is further configured to acquire the rated operating parameters corresponding to the target device; The determining module is further configured to determine the operating procedure corresponding to the target device based on the device status assessment result corresponding to the target device and the rated operating parameters; The acquisition module is also used to acquire real-time operating information corresponding to the target device; The judgment module is also used to determine whether the real-time running information matches the running program; The determining module is further configured to determine the safe operating range parameters of the target device based on the real-time operating information and the rated operating parameters when the judging module determines that the real-time operating information does not match the operating program. The generation module is also used to generate operation warning information and operation warning control parameters corresponding to the target device based on the safe operation range parameters; The control module is used to transmit the operation warning information to the target terminal corresponding to the target device, and control the target device to perform warning control operations that match the warning control parameters.

[0016] As an optional implementation, in a second aspect of the present invention, the specific method by which the determining module determines the average prediction parameter corresponding to the target feature information based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, combined with a pre-determined random regression model, includes: Based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, a feature subset partitioning operation is performed on the target feature information to obtain at least one feature information subset. Based on each subset of the feature information, a mapping relationship is generated between each model feature information and the information weight parameter corresponding to each model feature information. Based on the mapping relationship and the pre-determined random forest regression model, the average prediction parameter corresponding to the target feature information is determined. The model characteristic information includes equipment usage frequency characteristics, equipment workload characteristics, equipment operation habit characteristics, and equipment material characteristics.

[0017] As an optional implementation, in a second aspect of the present invention, the judgment module is further configured to determine, based on the equipment status evaluation result corresponding to the target equipment, whether the equipment status evaluation result is used to indicate that the target equipment does not meet the preset equipment operating conditions; The acquisition module is further configured to acquire real-time environmental information of the current environment where the target device is located when the judgment module determines that the device status evaluation result indicates that the target device does not meet the preset device operating conditions; The generation module is also used to generate device status warning information for the target device based on the device status assessment results, and to generate target status warning information for the target device based on the device status warning information and the real-time environmental information. The determining module is further configured to determine the early warning method parameters corresponding to the target device based on the real-time environmental information; The generation module is further configured to generate device warning parameters for the target device based on the target status warning information and the warning method parameters, and transmit the device warning parameters to the warning receiving terminal corresponding to the target device.

[0018] As an optional implementation, in a second aspect of the present invention, the specific method by which the processing module performs information preprocessing operations on the target feature information to obtain model characteristic information corresponding to the target feature information, and information weight parameters corresponding to each of the model characteristic information, includes: An information cleaning operation is performed on the target feature information to obtain a feature information cleaning result, and an information classification operation is performed on the feature information cleaning result to obtain at least one information category, wherein the information category includes a numerical information category and a label information category; Based on the information category, a target information category is determined, and model characteristic information matching the target information category is determined from all the feature information cleaning results included in the target information category. Local correlation information corresponding to the model characteristic information is extracted based on a pre-determined convolutional neural network. Based on the local correlation information and the model characteristic information, the feature temporal transformation correlation relationship is determined. Based on the correlation of the feature time-series transformation, determine the information weight parameter corresponding to each of the model characteristic information; The feature temporal transformation correlation relationship includes feature dynamic correlation relationship, temporal attenuation coefficient relationship, and material compatibility correlation relationship.

[0019] A third aspect of the present invention discloses another device for evaluating the condition of an artificial intelligence-based device, the device comprising: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute some or all of the steps in the artificial intelligence-based device status assessment method according to any of the first aspects of the present invention.

[0020] The fourth aspect of the present invention discloses a computer storage medium storing computer instructions, which, when invoked, are used to execute some or all of the steps in the artificial intelligence-based device status assessment method described in any of the first aspects of the present invention.

[0021] Compared with the prior art, the present invention has the following beneficial effects: In this embodiment of the invention, target feature information of the target device is acquired and information preprocessing is performed to obtain model characteristic information and information weight parameters of the target feature information. The average prediction parameters of the target feature information are determined by combining a stochastic regression model. Based on the average prediction parameters, an initial evaluation model is trained to obtain the model training result. Updated feature information is generated based on the model training result. Based on the updated feature information, an iterative model operation is performed on the initial evaluation model to obtain the target evaluation model. Finally, device operating information is acquired and input into the target evaluation model to obtain the device status evaluation result. Therefore, implementing this invention can combine artificial intelligence to evaluate device status, which is beneficial to improving the intelligence and efficiency of device status evaluation, as well as the accuracy and reliability of device status evaluation. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating an artificial intelligence-based device status assessment method disclosed in an embodiment of the present invention; Figure 2 This is a flowchart illustrating another artificial intelligence-based device status assessment method disclosed in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an artificial intelligence-based device for evaluating equipment status, as disclosed in an embodiment of the present invention. Figure 4This is a schematic diagram of another device status assessment device based on artificial intelligence disclosed in an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of another artificial intelligence-based device for evaluating equipment status disclosed in an embodiment of the present invention. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or end that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or ends.

[0026] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0027] This invention discloses a method for acquiring target feature information of a target device and performing information preprocessing operations to obtain model characteristic information and information weight parameters of the target feature information. The method then combines a stochastic regression model to determine the average prediction parameters of the target feature information. Based on the average prediction parameters, an initial evaluation model is trained to obtain the model training result. Updated feature information is generated based on the model training result, and iterative model operations are performed on the initial evaluation model based on the updated feature information to obtain the target evaluation model. Finally, device operating information is acquired and input into the target evaluation model to obtain the device status evaluation result. This method combines artificial intelligence to evaluate device status, which is beneficial for improving the intelligence and efficiency of device status evaluation, as well as its accuracy and reliability. Detailed explanations follow.

[0028] Example 1 Please see Figure 1 , Figure 1 This is a flowchart illustrating an artificial intelligence-based device status assessment method disclosed in an embodiment of the present invention. Figure 1 The described AI-based device status assessment method can be applied to AI-based device status assessment devices or to the device itself. The AI-based device status assessment device can be integrated into a cloud server or a local server; this embodiment of the invention does not impose any limitations. Figure 1 As shown, this AI-based device status assessment method may include the following operations: 101. Obtain the target feature information of the target device, perform information preprocessing operation on the target feature information to obtain the model characteristic information corresponding to the target feature information, and the information weight parameter corresponding to each model characteristic information.

[0029] In this embodiment of the invention, optionally, the target device may include a medical device; further, the device may include instruments used in oral root canal treatment. For example, the target device may include one or more of root canal files, reamers, etc.

[0030] In this embodiment of the invention, optionally, the target feature information of the target device may include one or more of the following: the cumulative number of times the target device has been used, the longest single working time, the frequency of torque peak occurrence, and doctor's operating habit tags.

[0031] 102. Based on the characteristic information of each model and the information weight parameters corresponding to each characteristic information, the average prediction parameters corresponding to the target feature information are determined in combination with the pre-determined random regression model.

[0032] In this embodiment of the invention, optionally, the pre-determined random regression model may include a random forest regression model, wherein the random forest regression model comprises an ensemble of 100 decision trees, and each tree learns a mapping rule of "feature combination - remaining lifetime" based on different feature subsets, ultimately outputting an average predicted value; furthermore, the random forest regression model is a regression method based on ensemble learning, belonging to the Bagging (Bootstrap Aggregating) family. It improves the accuracy and robustness of the model by constructing multiple decision trees and averaging their prediction results.

[0033] 103. Based on the average prediction parameters, perform training operations on the initial evaluation model to obtain the model training results. Based on the model training results, generate updated feature information. Based on the updated feature information, perform model iteration operations on the initial evaluation model to obtain the target evaluation model.

[0034] In this embodiment of the invention, optionally, the above-mentioned training operation on the initial evaluation model based on the average prediction parameters to obtain the model training result may include: Based on the average prediction parameters and the target feature information of the target device, an information partitioning operation is performed on the target feature information to obtain a training dataset and a training and validation set. The data feature vectors corresponding to the training dataset are input into the initial evaluation model, and the model parameters are optimized by gradient descent to obtain a preliminary optimized model. The training and validation set is then input into the preliminary optimized model, the validation data results of the training and validation set are calculated, and the model training results are generated based on the validation data results.

[0035] In this embodiment of the invention, optionally, the above-mentioned process of performing model iteration operations on the initial evaluation model based on updated feature information to obtain the target evaluation model may include: Based on the updated feature information, an updated data training set is generated and input into the preliminary evaluation model, and it is determined whether the preliminary evaluation model meets the preset model generation conditions. When it is determined that the preliminary evaluation model meets the preset model generation conditions, the target evaluation model is generated based on the preliminary evaluation model; when it is determined that the preliminary evaluation model does not meet the preset model generation conditions, the model adjustment parameters are applied to the preliminary evaluation model to make the preliminary evaluation model meet the preset model generation conditions, and the target evaluation model is generated based on the preliminary evaluation model.

[0036] In this embodiment of the invention, optionally, for example, after each use of the model, the terminal device uploads new working data (such as the duration of this use and the peak torque) to the cloud, and the model automatically updates the feature values ​​and recalculates the remaining lifespan; Model iteration: every 1000 accumulated usage data, the model is triggered to retrain (retaining historical parameters as initial values ​​and reducing the amount of computation through incremental learning), so that the model can adapt to the material differences of different batches of instruments (such as the fatigue characteristic fluctuations of NiTi alloy), so as to realize real-time equipment status assessment and intelligent equipment status assessment based on artificial intelligence model.

[0037] 104. Obtain the equipment operation information of the target equipment, input the equipment operation information into the target evaluation model, and obtain the equipment status evaluation result corresponding to the target equipment.

[0038] In this embodiment of the invention, optionally, the device operation information of the target device may include one or more of the following: current usage count information, current usage duration information, cumulative usage duration information, current peak torque, current average speed, and historical maximum speed.

[0039] In this embodiment of the invention, optionally, the process of inputting device operation information into the target evaluation model to obtain the device status evaluation result corresponding to the target device may include: The equipment operation information is input into the target evaluation model, which performs analysis operations on the equipment operation, obtains the model output results, and determines the equipment status evaluation result corresponding to the target equipment based on the model output results.

[0040] In this embodiment of the invention, optionally, the equipment status assessment result corresponding to the target device may include the life prediction parameters of the target device; further, if the equipment status assessment result of the target device is in a normal state, it is determined that the wear and tear of the target device is low, the remaining life is sufficient, and it can be used safely; if the equipment status assessment result of the target device is in a warning state, the device has entered the middle stage of wear and tear, the remaining life is limited, and attention should be paid to the operating procedures; if the equipment status assessment result of the target device is in a warning state, the device is close to being scrapped and there is a risk of breakage, so it should be used with caution; if the equipment status assessment result of the target device is in an emergency state, the device is currently in use at the risk of overload and may break immediately.

[0041] It is evident that implementation Figure 1 The described AI-based equipment status assessment method can acquire target feature information of the target equipment and perform information preprocessing operations to obtain model characteristic information and information weight parameters of the target feature information. It then combines this with a stochastic regression model to determine the average prediction parameters of the target feature information. Based on the average prediction parameters, it performs training operations on the initial assessment model to obtain the model training results. Based on the model training results, it generates updated feature information, and based on the updated feature information, it performs model iteration operations on the initial assessment model to obtain the target assessment model. Finally, it acquires equipment operating information and inputs it into the target assessment model to obtain the equipment status assessment result. This method can comprehensively determine the model feature information and corresponding information weight parameters based on multiple indicator dimensions of the target feature information, which is beneficial for improving the subsequent determination of the average prediction parameters and generation of target feature information. This approach improves the accuracy and reliability of the target-based assessment model, thereby enhancing the accuracy and reliability of equipment condition assessment results derived from the target assessment, as well as the intelligence and efficiency of such assessments. Each model characteristic corresponds to a specific information weight parameter, and the model corrects prediction results through high-weight features, making the average prediction parameters closer to actual losses. This contributes to providing accurate and reliable equipment condition assessment results. Furthermore, combining this model with a stochastic regression model further improves prediction accuracy and reliability, thereby increasing the precision of equipment condition assessment. Finally, combining this model with artificial intelligence for equipment condition assessment enhances the intelligence and efficiency of equipment condition assessment, and ultimately improves its accuracy and reliability.

[0042] Example 2 Please see Figure 2 , Figure 2This is a flowchart illustrating another artificial intelligence-based device status assessment method disclosed in an embodiment of the present invention. Figure 2 The described AI-based device status assessment method can be applied to AI-based device status assessment devices or to the device itself. The AI-based device status assessment device can be integrated into a cloud server or a local server; this embodiment of the invention does not impose any limitations. Figure 2 As shown, this AI-based device status assessment method may include the following operations: 201. Obtain the target feature information of the target device, perform information preprocessing operation on the target feature information to obtain the model characteristic information corresponding to the target feature information, and the information weight parameter corresponding to each model characteristic information.

[0043] 202. Based on the characteristic information of each model and the information weight parameters corresponding to each characteristic information, the average prediction parameters corresponding to the target feature information are determined in combination with the pre-determined random regression model.

[0044] 203. Based on the average prediction parameters, perform training operations on the initial evaluation model to obtain the model training results. Based on the model training results, generate updated feature information. Based on the updated feature information, perform model iteration operations on the initial evaluation model to obtain the target evaluation model.

[0045] 204. Obtain the equipment operation information of the target equipment, input the equipment operation information into the target evaluation model, and obtain the equipment status evaluation result corresponding to the target equipment.

[0046] In this embodiment of the invention, for a detailed description of steps 201-204, please refer to the other descriptions of steps 101-104 in Embodiment 1. This embodiment of the invention will not repeat them.

[0047] 205. Based on the equipment status assessment results and the pre-determined cross-validation optimization model, determine whether the target assessment model meets the preset model assessment conditions.

[0048] In this embodiment of the invention, optionally, the pre-determined cross-validation optimization model may include a 5-fold cross-validation optimization model. The 5-fold cross-validation optimization model is a framework for model evaluation and hyperparameter tuning, which divides the training set evenly into 5 parts (non-overlapping); each time, one part is used as the validation set, and the remaining 4 parts are used as the training set; this is repeated 5 times, each time using a different fold for validation; finally, the average of the 5 validation results is taken as the model performance index (such as accuracy, F1, AUC, etc.). For example, by collecting the full lifecycle data (from first use to breakage / discard) of 500+ similar devices, labeling them with "feature value - actual remaining uses", and using the 5-fold cross-validation optimization model, the prediction error can be ≤ ±1 (for devices with a rated lifespan of 20 uses) to determine whether the target evaluation model meets the preset model evaluation conditions.

[0049] In this embodiment of the invention, optionally, the above-mentioned determination of whether the target evaluation model meets the preset model evaluation conditions based on the equipment status evaluation results and the pre-determined cross-validation optimization model may include: Based on the equipment status assessment results, the model validation dataset is determined and divided into multiple equal quantum sets. Data validation operations are performed on each equal quantum set to obtain the data deviation parameters corresponding to each equal quantum set. Based on all data deviation parameters, a target deviation value is generated, and it is determined whether the target deviation value is greater than or equal to the deviation threshold corresponding to the preset model evaluation conditions. When the target deviation value is determined to be greater than or equal to the deviation threshold corresponding to the preset model evaluation conditions, it is determined that the target evaluation model does not meet the preset model evaluation conditions; when the target deviation value is determined to be less than the deviation threshold corresponding to the preset model evaluation conditions, it is determined that the target evaluation model meets the preset model evaluation conditions.

[0050] 206. When it is determined that the target evaluation model does not meet the preset model evaluation conditions, determine the model update parameters corresponding to the target evaluation model, and perform a model update operation on the target evaluation model based on the model update parameters to obtain an updated evaluation model.

[0051] In this embodiment of the invention, optionally, the process can be terminated when it is determined that the target evaluation model meets the preset model evaluation conditions.

[0052] In this embodiment of the invention, optionally, the process of determining the model update parameters corresponding to the target evaluation model and performing a model update operation on the target evaluation model based on the model update parameters to obtain an updated evaluation model may include: The reasons why the target evaluation model does not meet the preset model evaluation conditions are identified. Based on the reasons, the model update parameters that match the reasons in the target evaluation model are determined. Based on the model update parameters, the model update operation is performed on the target evaluation model to update the target evaluation model and obtain the updated evaluation model.

[0053] In this embodiment of the invention, the reasons for the target model may optionally include one or more of the following: low prediction accuracy and large error fluctuation.

[0054] In this embodiment of the invention, the model update parameters may optionally include one or more of the following: feature weight adjustment parameters, decision tree number adjustment parameters, and model architecture adjustment parameters.

[0055] It is evident that implementation Figure 2 The described AI-based equipment status assessment method can determine whether the target assessment model meets the preset model assessment conditions based on the equipment assessment results and cross-validation optimization model. If not, it determines the corresponding model update parameters and performs a model update operation on the target assessment model to obtain an updated assessment model. This method can perform intelligent assessment operations on the equipment assessment results through cross-validation optimization, which helps improve the accuracy and reliability of updating the target assessment model. Furthermore, it can perform update operations on the target assessment model based on the determined model update parameters to improve the intelligence and efficiency of model adjustment and updating. When the model does not meet the assessment conditions, it can specifically correct accuracy defects through model update parameters, further improving the intelligence and efficiency of model adjustment and updating. It can also dynamically update the target assessment model based on the equipment assessment results to obtain an updated assessment model, which helps improve the model adaptability and real-time performance of the target assessment model. This enables intelligent and real-time equipment status assessment, and by combining AI with equipment status assessment, it improves the intelligence and efficiency of equipment status assessment, as well as its accuracy and reliability.

[0056] In an optional embodiment, generating updated feature information based on the model training results includes: Based on the model training results, the model training results are uploaded to the predetermined target cloud to obtain the model feedback information from the target cloud; Based on model feedback information, the parameters to be adjusted in the initial evaluation model and the parameter adjustment factor corresponding to each parameter to be adjusted are determined. Updated feature information is then generated based on each parameter to be adjusted and the parameter adjustment factor corresponding to each parameter to be adjusted.

[0057] In this optional embodiment, the pre-determined target cloud may include a cloud server corresponding to the target device, and the cloud server has the functions of massive data storage, distributed computing, model performance analysis, and parameter optimization suggestion generation, and can pre-store the basic parameter library, historical model training dataset, and clinical use scenario label library of different models of medical devices.

[0058] In this optional embodiment, the model training results may include the data output after performing training operations based on the initial evaluation model.

[0059] In this optional embodiment, the updated feature information may include optimized feature data obtained by performing iterative operations on the initial evaluation model, so as to ensure that the iterated model can more accurately adapt to the actual wear patterns of the machinery.

[0060] In this optional embodiment, the process of uploading the model training results to a predetermined target cloud to obtain model feedback information from the target cloud may include: Based on the model training results, the model training results are uploaded to the predetermined target cloud, so that the target cloud can perform analysis operations on the model training results based on the predetermined cloud database and generate model feedback information to obtain the model feedback information fed back by the target cloud. The model feedback information provided by the target cloud may include one or more of the following: initial model performance evaluation feedback, parameter adjustment feedback, and data supplementation feedback.

[0061] In this optional embodiment, optionally, the above-mentioned determination of the model parameters to be tuned in the initial evaluation model and the parameter adjustment factor corresponding to each model parameter to be tuned based on model feedback information, and the generation of updated feature information based on each model parameter to be tuned and the parameter adjustment factor corresponding to each model parameter to be tuned, may include: Based on the model feedback information, at least one model parameter to be tuned that matches the model feedback information is identified in the initial evaluation model; For each model parameter to be tuned, the parameter adjustment factor corresponding to the model parameter is determined based on the model feedback information, and updated feature information is generated based on each model parameter to be tuned and the parameter adjustment factor corresponding to each model parameter to be tuned. The parameters of the model to be adjusted may include one or more of the following: equipment load characteristic weight parameters, scene sample weight parameters, material fatigue weight parameters, and user operation weight parameters.

[0062] In this optional embodiment, for example, after each use of the model, the terminal device uploads new working data (such as the duration of this use and the peak torque) to the cloud, and the model automatically updates the feature values ​​and recalculates the remaining lifespan; every 1000 uses of data are accumulated, the model is triggered to retrain (historical parameters are retained as initial values, and the amount of computation is reduced through incremental learning) so that the model can adapt to the material differences of different batches of instruments (such as the fatigue characteristics fluctuation of NiTi alloy).

[0063] As can be seen, implementing this optional embodiment can upload the model training results to the target cloud to obtain model feedback information from the target cloud. Based on the model feedback information, the parameters to be tuned for the initial evaluation model and the parameter adjustment factors corresponding to each parameter to be tuned are determined. Updated feature information is generated according to each parameter to be tuned and its corresponding parameter adjustment factor. It can combine the determination of the parameters to be tuned and their corresponding parameter adjustment factors with the target cloud, and can transfer resource-intensive operations such as feature importance analysis and scenario error distribution to the cloud, thereby freeing up local resources and memory. This is beneficial to improving the intelligence and efficiency of model optimization. Furthermore, by uniformly processing multi-terminal data through the cloud, it ensures that the model adjustment standards of all devices are consistent, which is beneficial to improving the consistency and objectivity of subsequent performance evaluation of the target device. It can also optimize the model smoothly through parameter adjustment factors, avoiding over-adjustment that leads to model instability, ensuring that the model is adapted to diverse clinical scenarios, which is beneficial to improving the adaptability and flexibility of the model. It can achieve intelligent and real-time evaluation of device status, and further, it can combine artificial intelligence to evaluate device status, which is beneficial to improving the intelligence and efficiency of device status evaluation, as well as the accuracy and reliability of device status evaluation.

[0064] In another alternative embodiment, the method further includes: Obtain the rated operating parameters of the target equipment, and determine the operating procedures of the target equipment based on the equipment status assessment results and rated operating parameters. Obtain the real-time operating information corresponding to the target device and determine whether the real-time operating information matches the operating program; When it is determined that the real-time operation information does not match the operation program, the safe operation range parameters of the target device are determined based on the real-time operation information and the rated operation parameters. Based on the safe operation range parameters, the corresponding operation warning information and operation warning control parameters of the target device are generated. The operation warning information is transmitted to the target terminal corresponding to the target device, and the target device is controlled to perform the warning control operation that matches the warning control parameters.

[0065] In this optional embodiment, the rated operating parameters corresponding to the target device may include one or more of the target device's rated speed, rated torque, and maximum rated operating time. For example, the rated operating parameters corresponding to the target device may include a recommended speed of 300-500 rpm and a maximum torque of 2.5 N•cm.

[0066] In this optional embodiment, the equipment status assessment result corresponding to the target device may optionally include the remaining lifetime count information and the wear level information of the target device.

[0067] In this optional embodiment, the above-mentioned determination of the operating procedure corresponding to the target device based on the device status assessment results and rated operating parameters may include: Based on the equipment status assessment results and rated operating parameters of the target equipment, the equipment operating parameters of the target equipment are determined, and the equipment operating parameters are sent to the target components of the target equipment so that the target components can determine the operating procedures corresponding to the equipment operating parameters. The operating procedure is a set of clinical operation instructions adapted to the current state and rated parameters of the target device, and is used to control the target device to perform the corresponding operating operations.

[0068] In this optional embodiment, the real-time operating information corresponding to the target device may include one or more of the following: real-time rotational speed, real-time torque, real-time operating duration, and real-time operating status of the target device.

[0069] In this optional embodiment, the determination of whether the real-time running information matches the running program may include: The real-time operating status of the target device is determined based on real-time operating information, and the expected operating status of the target device is determined based on the operating procedure. It is then determined whether the real-time operating status matches the expected operating status. When it is determined that the real-time operating status matches the expected operating status, it is determined that the real-time operating information matches the operating program; when it is determined that the real-time operating status does not match the expected operating status, it is determined that the real-time operating information does not match the operating program.

[0070] In this optional embodiment, it is further possible to terminate the process when it is determined that the real-time running information matches the running program.

[0071] In this optional embodiment, the safe operating range parameter of the target device may include one or more of the following: safe operating speed range, safe operating torque range, and safe operating duration range.

[0072] In this optional embodiment, the operational warning information corresponding to the target device may include warning prompts for informing the operator of the risk type, current status, and recommended operations of the target device, and the operational warning control parameters corresponding to the target device may include instruction parameters for automatically controlling the target device to adjust its status.

[0073] In this optional embodiment, the process of generating operational warning information and operational warning control parameters corresponding to the target device based on safe operating range parameters, transmitting the operational warning information to the target terminal corresponding to the target device, and controlling the target device to execute warning control operations matching the warning control parameters may include: Based on the safe operating range parameters, determine the equipment warning parameters corresponding to the target equipment, and generate operating warning information based on the equipment warning parameters. The equipment warning parameters include one or more of the following: warning level, risk type, and current status parameters of the target equipment. Based on the safe operating range parameters, the equipment safety adjustment parameters corresponding to the target equipment are determined, and operating early warning control parameters are generated based on the equipment safety adjustment parameters. The equipment safety adjustment parameters may include one or more of the target equipment's torque adjustment parameters and speed adjustment parameters. The operating early warning information is transmitted to the target terminal corresponding to the target equipment so as to provide early warning prompts to the operator corresponding to the target equipment through the target terminal.

[0074] As can be seen, implementing this optional embodiment can obtain the rated operating parameters of the target device and determine the corresponding operating procedure based on the device status assessment results and the rated operating parameters. It obtains the real-time operating information of the target device to determine whether it matches the operating procedure. If they do not match, it determines the safe operating range parameters based on the real-time operating information and the rated operating parameters. Based on the safe operating range parameters, it generates operating warning information and operating warning control information for the target device and transmits the operating warning information to the target terminal corresponding to the target device. It also controls the target device to execute warning control operations that match the warning control parameters. This comprehensive approach, based on the device status assessment results and the rated operating parameters, helps improve the accuracy and reliability of determining the corresponding operating procedure for the target component and improves the efficiency of determining the operating procedure. The accuracy and reliability of the program can be improved by using real-time data collection of speed, torque, and working time from the root canal motor to determine the match between the data and the running program. This improves the accuracy of judging whether real-time operating information matches the running program. Furthermore, it can comprehensively determine safe operating range parameters based on real-time operating information and rated operating parameters, enhancing the accuracy and reliability of determining these parameters, as well as their intelligence and efficiency. This, in turn, improves the accuracy and reliability of generating operating warning information and operating warning control parameters, and enhances the safety and reliability of the target equipment. Finally, it can be combined with artificial intelligence to assess equipment status, improving the intelligence and efficiency of equipment status assessment, as well as its accuracy and reliability.

[0075] In another optional embodiment, the average prediction parameter corresponding to the target feature information is determined based on each model feature information and the information weight parameter corresponding to each model feature information, combined with a pre-determined random regression model, including: Based on each model feature information and the information weight parameter corresponding to each model feature information, a feature subset partitioning operation is performed on the target feature information to obtain at least one feature information subset. Based on each subset of feature information, a mapping relationship is generated between each model feature information and the information weight parameters corresponding to each model feature information. Based on the mapping relationship and the pre-determined random forest regression model, the average prediction parameters corresponding to the target feature information are determined. The model characteristic information includes equipment usage frequency characteristics, equipment workload characteristics, equipment operation habit characteristics, and equipment material characteristics.

[0076] In this optional embodiment, the equipment usage frequency characteristic information may include the cumulative number of times the target equipment has been used and the interval between the last three uses of the target equipment; the equipment workload characteristic information may include the longest single working time of the target equipment and the frequency of peak torque occurrence; the equipment operation habit characteristic information may include the user's operation tag information and the average speed fluctuation coefficient information; and the equipment material characteristic information may include the material fatigue coefficient information.

[0077] In this optional embodiment, the above-mentioned feature subset partitioning operation on the target feature information based on each model feature information and the information weight parameter corresponding to each model feature information to obtain at least one feature information subset may include: Based on each model feature information and the information weight parameter corresponding to each model feature information, determine the number of subset data corresponding to each feature subset and the data subset weight corresponding to each feature subset. Perform feature subset partitioning operation on the target feature information based on the number of subset data and the data subset weight to obtain at least one feature information subset. Feature subset partitioning involves assigning differentiated feature combinations to each decision tree in the random forest. The aim is to improve the generalization ability of the model by having multiple trees in the random forest regression model learn from different feature subsets, thereby avoiding overfitting.

[0078] In this optional embodiment, the mapping relationship between generating each model feature information and the corresponding information weight parameter based on each feature information subset may include: Based on each subset of feature information, generate the model feature information corresponding to that subset of feature information, and determine the corresponding information weight parameters according to the model feature information, thereby generating the mapping relationship between each model feature information and the information weight parameters corresponding to each model feature information. The mapping relationship can include the following: when the model characteristic information is the equipment workload characteristic information, the corresponding information weight parameter is 30%; when the model characteristic information is the equipment material characteristic information, the corresponding information weight parameter is 25%; when the model characteristic information is the equipment usage frequency characteristic information, the corresponding information weight parameter is 20%; when the model characteristic information is the equipment operation habit characteristic information, the corresponding information weight parameter is 25%; when the model characteristic information is the equipment workload characteristic information, the corresponding information weight parameter is 30%; and when the model characteristic information is the equipment operation habit characteristic information, the corresponding information weight parameter is 25%.

[0079] In this optional embodiment, the above-mentioned determination of the average prediction parameters corresponding to the target feature information based on the mapping relationship and the pre-determined random forest regression model may include: Based on the mapping relationship and the predetermined random forest regression model, based on the corresponding feature information subset and mapping relationship, the remaining lifetime is predicted by the node splitting and leaf node output of the random forest regression model. The prediction process is repeated by traversing all decision trees included in the random forest regression model to obtain the predicted value of each tree. Based on the predicted values ​​of all decision trees included in the random forest regression model, the average prediction parameter corresponding to the target feature information is determined. The average prediction parameter corresponding to the target feature information may include one or more of the predicted lifetime average and the predicted usage status average corresponding to the target feature information.

[0080] In this optional embodiment, for example, a random forest regression model (ensemble of 100 decision trees) may be used: each tree learns a mapping rule of "feature combination - remaining lifetime" based on different feature subsets, and finally outputs an average prediction value (unit: times).

[0081] As can be seen, implementing this optional embodiment can perform feature subset partitioning on the target feature information based on each model feature information and the corresponding information weight parameters, obtaining at least one feature information subset, and generating a mapping relationship between each model feature information and the corresponding information weight parameters. This is then combined with a random forest regression model to determine the average prediction parameters corresponding to the target feature information. By combining features from multiple dimensions and their corresponding information weight parameters, the determined mapping relationship is more aligned with the actual wear patterns of the target equipment. Furthermore, by combining information weight parameters and using weighted random sampling when partitioning feature subsets, high-weight features appear in more subsets, which improves the accuracy and reliability of subsequent mapping relationship generation and average prediction parameter determination. Weighted sampling ensures subset diversity, addressing the problem of overfitting of a single model leading to unstable predictions in existing technologies, thus improving the stability and reliability of the average prediction parameters. Additionally, based on the mapping relationship, it ensures that each tree in the random forest regression model prioritizes high-weight features during splitting, further enhancing the stability of the average prediction parameters. Finally, it can combine artificial intelligence to evaluate equipment status, improving the intelligence and efficiency of equipment status evaluation, as well as its accuracy and reliability.

[0082] In yet another optional embodiment, the method further includes: Based on the equipment status assessment results corresponding to the target equipment, determine whether the equipment status assessment results are used to indicate that the target equipment does not meet the preset equipment operating conditions; When it is determined that the equipment status assessment result indicates that the target equipment does not meet the preset equipment operating conditions, real-time environmental information of the current environment where the target equipment is located is obtained; Based on the equipment status assessment results, generate equipment status early warning information for the target equipment, and generate target status early warning information for the target equipment based on the equipment status early warning information and real-time environmental information, and determine the corresponding early warning method parameters for the target equipment based on the real-time environmental information. Based on the target status warning information and warning method parameters, device warning parameters for the target device are generated and transmitted to the warning receiving terminal corresponding to the target device.

[0083] In this optional embodiment, it is further possible to terminate the process when it is determined that the equipment status evaluation result indicates that the target equipment meets the preset equipment operating conditions.

[0084] In this optional embodiment, the determination of whether the device status assessment result is used to indicate that the target device does not meet the preset device operating conditions may include: Determine the equipment operation risk value corresponding to the equipment status assessment result, and determine whether the equipment operation risk value is greater than or equal to the preset operation risk threshold corresponding to the equipment operation conditions; When it is determined that the equipment operation risk value is greater than or equal to the preset equipment operation risk threshold, the equipment status assessment result is used to indicate that the target equipment does not meet the preset equipment operation conditions; when it is determined that the equipment operation risk value is less than the preset equipment operation risk threshold, the equipment status assessment result is used to indicate that the target equipment meets the preset equipment operation conditions.

[0085] In this optional embodiment, optionally, for example, if it is determined that the equipment operation risk value is used to indicate that the remaining lifespan is less than the preset equipment operation conditions, then the equipment operation risk value is greater than or equal to the operation risk threshold corresponding to the preset equipment operation conditions, and the equipment status assessment result is used to indicate that the target equipment does not meet the preset equipment operation conditions; if it is determined that the equipment operation risk value is used to indicate that the loss registration is less than the preset equipment operation conditions, then the equipment operation risk value is greater than or equal to the operation risk threshold corresponding to the preset equipment operation conditions, and the equipment status assessment result is used to indicate that the target equipment does not meet the preset equipment operation conditions.

[0086] In this optional embodiment, the above-mentioned generation of device status warning information for the target device based on the device status assessment results may include: Based on the equipment status assessment results and the preset equipment operating conditions, determine the equipment operation assessment parameters corresponding to the target equipment. The equipment operation assessment parameters corresponding to the target equipment may include one or more of the following: remaining service life, wear level, and usage risk level. Based on the equipment operation evaluation parameters corresponding to the target equipment, generate equipment status early warning information for the target equipment.

[0087] In this optional embodiment, the real-time environmental information may optionally include one or more of the following: ambient temperature information, ambient humidity information, ambient vibration value information, and ambient noise value information.

[0088] In this optional embodiment, the process of generating target status warning information for the target device based on device status warning information and real-time environmental information, and determining the warning method parameters corresponding to the target device based on real-time environmental information, may include: Based on the equipment status warning information and real-time environmental information, determine the warning impact factor of real-time environmental information on the equipment status warning information, and based on the warning impact factor, perform an update operation on the equipment status warning information to generate the target status warning information of the target equipment. Based on the target status warning information and real-time environmental information of the target device, determine the warning method corresponding to the target device, and determine the warning method parameters corresponding to the target device based on the warning method; Among them, the early warning influencing factors may include one or more of the following: temperature influencing factors, humidity influencing factors, vibration influencing factors, etc.; the early warning methods may include one or more of the following: text early warning method, voice early warning method, video early warning method.

[0089] In this optional embodiment, the device warning parameters of the target device may optionally include one or more of the following: basic representation parameters, warning information parameters, sound warning parameters, light warning parameters, and video warning parameters; furthermore, the device warning parameters are a structured combination of target status warning information and warning method parameters, and are a complete instruction set for the warning receiving terminal to perform warning operations.

[0090] In this optional embodiment, the device warning parameters may be transmitted to the warning receiving terminal corresponding to the target device, so that the user corresponding to the warning receiving terminal can receive the device warning information.

[0091] In this optional embodiment, the target component of the target device may optionally include a root canal motor, wherein the cloud sends the instrument's rated parameters (such as recommended rotation speed of 300-500 rpm and maximum torque of 2.5 N•cm) to the root canal motor, and the motor automatically retrieves the matching working program; if the doctor manually adjusts the parameters beyond the safe range, the motor triggers an audible and visual alarm and restricts its operation.

[0092] As can be seen, implementing this optional embodiment can determine whether the equipment status assessment result indicates that the target equipment does not meet the preset equipment operating conditions. If so, it acquires real-time environmental information and generates equipment status warning information and target status warning information, thereby determining the warning method parameters. Based on the target status warning information and the warning method parameters, it generates equipment warning parameters for the target equipment and transmits the equipment warning parameters to the warning receiving terminal corresponding to the target equipment. It can generate equipment status warning information based on the equipment status assessment result, which is beneficial to improving the accuracy and reliability of generating equipment status warning information, as well as improving the intelligence and efficiency of generating equipment status warning information. Furthermore, it can combine real-time environmental information to assess the equipment. The influence of the status allows for the comprehensive generation of corresponding early warning information by combining multiple factors, which improves the comprehensiveness, accuracy, and reliability of the generated equipment early warning information. It also enables the determination of corresponding early warning methods by combining real-time environmental information, which improves the matching degree between the determined early warning method and the current environment, and enhances the convenience for users to view early warning information. This, in turn, improves the safety and reliability of equipment operation and user use of the target equipment. Furthermore, it allows for the evaluation of equipment status by combining artificial intelligence, which improves the intelligence and efficiency of equipment status evaluation, as well as the accuracy and reliability of the evaluation.

[0093] In another optional embodiment, information preprocessing is performed on the target feature information to obtain model characteristic information corresponding to the target feature information, and information weight parameters corresponding to each model characteristic information, including: Perform information cleaning operation on the target feature information to obtain the feature information cleaning result, and perform information classification operation on the feature information cleaning result to obtain at least one information category, wherein the information category includes numerical information category and label information category; Based on the information category, the target information category is determined, and the model characteristic information that matches the target information category is determined from all the feature information cleaning results included in the target information category. Based on the pre-determined convolutional neural network, the local correlation information corresponding to the model characteristic information is extracted. Based on the local correlation information and the model characteristic information, the feature temporal transformation correlation relationship is determined. Based on the correlation of feature temporal transformation, determine the information weight parameters corresponding to each model characteristic information; Among them, the feature temporal transformation correlation includes feature dynamic correlation, temporal attenuation coefficient correlation, and material compatibility correlation.

[0094] In this optional embodiment, the information categories may optionally include usage frequency characteristic information categories, workload characteristic information categories, operating habit characteristic information categories, and material characteristic information categories. The frequency characteristic information category includes the cumulative number of uses and the interval between the last three uses. The workload characteristic information category includes the longest single working time and the frequency of torque peak occurrence. The operating habit characteristic information category includes the user's operating speed and the average speed fluctuation coefficient. The material characteristic information category includes the material fatigue coefficient, which may be pre-stored based on the material model and increases linearly with the number of uses.

[0095] In this optional embodiment, the information cleaning operation may include removing outliers, missing values, and duplicate values ​​from the target feature information to ensure data validity.

[0096] In this optional embodiment, the numerical information category may optionally include equipment usage frequency characteristics: cumulative number of uses (integer), interval between the last 3 uses (minutes, consecutive values); equipment workload characteristics: longest single working time (seconds, consecutive values); frequency of torque peak occurrence (times / hour, consecutive values); and equipment material characteristic characteristics: material fatigue coefficient (decimal, consecutive values). The label information category may include equipment operation habit characteristics, including the user's operation label (high speed / stable); equipment material characteristic characteristics: material type (NiTi alloy / stainless steel); and equipment risk label (extension features of the disclosure document): whether the torque exceeds the threshold.

[0097] In this optional embodiment, optionally, the process of determining model characteristic information matching the target information category from all feature information cleaning results included in the target information category, extracting local correlation information corresponding to the model characteristic information based on a pre-determined convolutional neural network, and determining the feature temporal transformation correlation based on the local correlation information and the model characteristic information may include: In the cleaned results of all feature information included in the target information category, model characteristic information that matches the target information category is identified, and a corresponding model feature structured dataset is output based on the model characteristic information. The dependencies between indicators in the model characteristic information are extracted based on a pre-determined convolutional neural network, and local correlation information is generated based on the dependencies between indicators. Information association operations are performed based on local association information and model characteristic information to obtain association parameters, and the association relationship of feature temporal transformation is determined based on the association parameters; wherein, the association parameters include one or more of the following: association feature pairs, association strength values, and association types.

[0098] In this optional embodiment, the characteristic time-series transformation correlation may include the dynamic dependency of model characteristic information indicators as the number of uses increases; the time-series decay coefficient relationship may include the decay law of the influence of model characteristic information indicators over time (usage interval), for example, the longer the interval between the last 3 uses, the smaller the impact of the torque peak of the previous use on the current lifespan; the material compatibility correlation may include the compatibility dependency between model characteristic information indicators and material type, for example, NiTi alloy instruments are more sensitive to torque peaks, and stainless steel instruments are more sensitive to working time.

[0099] In this optional embodiment, the determination of the information weight parameter corresponding to each model characteristic information based on the feature temporal transformation correlation may include: Based on the correlation relationship of feature time series transformation, determine the correlation strength score corresponding to each model feature information, and based on the correlation strength score, determine the information weight coefficient corresponding to each model feature information; For each type of model characteristic information, a correlation strength score is calculated based on three types of correlation relationships to quantify the impact of that type of feature on lifespan. The higher the score, the higher the weight.

[0100] As can be seen, implementing this optional embodiment can perform information cleaning operations on target feature information to obtain feature information cleaning results, perform information classification operations to obtain at least one information category and determine the target information category, thereby determining model characteristic information and extracting local correlation information by combining convolutional neural networks. Based on the local correlation information and model characteristic information, the feature temporal transformation correlation relationship is determined, and the corresponding information weight coefficient is determined based on the feature temporal transformation correlation relationship. It can remove invalid data through information cleaning operations, ensure data authenticity, and avoid subsequent model prediction deviations due to abnormal data. This is beneficial to improving the accuracy and reliability of determining information weight coefficients, thereby improving the accuracy and reliability of subsequent determination of target evaluation models and the determination of target device status evaluation results based on target evaluation models. It can also classify the cleaning results to avoid data type mismatch. The resulting model errors further improve the accuracy and reliability of subsequent target evaluation model determination and target device status assessment results based on the target evaluation model. It also allows for the extraction of local correlations using convolutional neural networks, thereby improving the accuracy and reliability of determining feature temporal transformation correlations. This, in turn, enhances the accuracy and reliability of subsequent target evaluation model determination and target device status assessment results based on the target evaluation model. Furthermore, based on feature temporal transformation correlations, the model better reflects actual loss patterns, facilitating real-time and intelligent assessment of target device status accuracy and reliability. Finally, combining artificial intelligence with device status assessment improves the intelligence and efficiency of device status assessment, as well as its accuracy and reliability.

[0101] Example 3 Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of an artificial intelligence-based device for evaluating equipment status, as disclosed in an embodiment of the present invention. Figure 3 As shown, the AI-based device status assessment device may include: The acquisition module 301 is used to acquire target feature information of the target device; The processing module 302 is used to perform information preprocessing operations on the target feature information to obtain the model characteristic information corresponding to the target feature information, and the information weight parameters corresponding to each model characteristic information. The determination module 303 is used to determine the average prediction parameter corresponding to the target feature information based on each model feature information and the information weight parameter corresponding to each model feature information, combined with the pre-determined random regression model. Training module 304 is used to perform training operations on the initial evaluation model based on the average prediction parameters to obtain the model training results; The generation module 305 is used to generate updated feature information based on the model training results; Iteration module 306 is used to perform model iteration operations on the initial evaluation model based on updated feature information to obtain the target evaluation model; The acquisition module 301 is also used to acquire the device operation information of the target device; The input module 307 is used to input equipment operation information into the target evaluation model to obtain the equipment status evaluation result corresponding to the target equipment.

[0102] It is evident that implementation Figure 3 The described device can acquire target feature information of a target device and perform information preprocessing operations to obtain model characteristic information and information weight parameters of the target feature information. It then combines this with a stochastic regression model to determine the average prediction parameters of the target feature information. Based on the average prediction parameters, it performs training operations on an initial evaluation model to obtain model training results. Based on the model training results, it generates updated feature information and performs model iteration operations on the initial evaluation model based on the updated feature information to obtain a target evaluation model. Finally, it acquires device operating information and inputs it into the target evaluation model to obtain device status evaluation results. This device can comprehensively determine model feature information and corresponding information weight parameters based on multiple indicator dimensions of the target feature information, which is beneficial for improving the subsequent determination of the average prediction parameters of the target feature information and the generation of the target evaluation model. This improves the accuracy and reliability of equipment condition assessment results obtained from target assessments, as well as the intelligence and efficiency of such assessments. Each model characteristic corresponds to a specific information weight parameter, and the model corrects prediction results through high-weight features, making the average prediction parameters closer to actual losses. This enhances the accuracy and reliability of equipment condition assessment results. Furthermore, combining this with a stochastic regression model improves prediction accuracy and reliability, thereby increasing the precision of equipment condition assessments. Finally, combining this with artificial intelligence for equipment condition assessment further enhances the intelligence and efficiency of equipment condition assessments, and ultimately improves their accuracy and reliability.

[0103] In an optional embodiment, such as Figure 4 As shown, the device also includes: The judgment module 308 is used to determine whether the target evaluation model meets the preset model evaluation conditions based on the equipment status evaluation results and the predetermined cross-validation optimization model after the input module 307 inputs the equipment operation information into the target evaluation model and obtains the equipment status evaluation results corresponding to the target equipment. The determination module 303 is also used to determine the model update parameters corresponding to the target evaluation model when the judgment module 308 determines that the target evaluation model does not meet the preset model evaluation conditions. The update module 309 is also used to perform a model update operation on the target evaluation model based on the model update parameters to obtain an updated evaluation model.

[0104] It is evident that implementation Figure 4 The described device can determine whether a target evaluation model meets preset model evaluation conditions based on equipment evaluation results and cross-validation optimization models. If not, it determines the corresponding model update parameters and performs a model update operation on the target evaluation model to obtain an updated evaluation model. It can perform intelligent evaluation operations on equipment evaluation results through cross-validation optimization models, which helps improve the accuracy and reliability of updating the target evaluation model. Furthermore, it can perform update operations on the target evaluation model based on the determined model update parameters to improve the intelligence and efficiency of model adjustment and updating. When the model does not meet the evaluation conditions, it can specifically correct accuracy defects through model update parameters, further improving the intelligence and efficiency of model adjustment and updating. It can also dynamically update the target evaluation model based on equipment evaluation results to obtain an updated evaluation model, which helps improve the model adaptability and real-time performance of the target evaluation model. It can achieve intelligent and real-time equipment status evaluation, and further, it can combine artificial intelligence to evaluate equipment status, which helps improve the intelligence and efficiency of equipment status evaluation, as well as the accuracy and reliability of equipment status evaluation.

[0105] In another alternative embodiment, such as Figure 4 As shown, the specific methods by which the generation module 305 generates updated feature information based on the model training results include: Based on the model training results, the model training results are uploaded to the predetermined target cloud to obtain the model feedback information from the target cloud; Based on model feedback information, the parameters to be adjusted in the initial evaluation model and the parameter adjustment factor corresponding to each parameter to be adjusted are determined. Updated feature information is then generated based on each parameter to be adjusted and the parameter adjustment factor corresponding to each parameter to be adjusted.

[0106] It is evident that implementation Figure 4The described device can upload model training results to the target cloud to obtain model feedback information from the target cloud. Based on the model feedback information, it determines the parameters of the initial evaluation model to be tuned and the corresponding parameter adjustment factors for each parameter. It generates updated feature information based on each parameter and its corresponding adjustment factor. It can combine the determination of the parameters of the model to be tuned and their corresponding adjustment factors with the target cloud, and can transfer resource-intensive operations such as feature importance analysis and scene error distribution to the cloud, thereby freeing up local resources and memory. This is beneficial to improving the intelligence and efficiency of model optimization. Furthermore, by uniformly processing multi-terminal data through the cloud, it ensures that the model adjustment standards of all devices are consistent, which is beneficial to improving the consistency and objectivity of subsequent performance evaluation of the target device. It can also perform smooth optimization of the model through parameter adjustment factors, avoiding over-adjustment that could lead to model instability, ensuring that the model is adapted to diverse clinical scenarios, which is beneficial to improving the adaptability and flexibility of the model. It can achieve intelligent and real-time device status evaluation, and further, it can combine artificial intelligence to evaluate device status, which is beneficial to improving the intelligence and efficiency of device status evaluation, as well as the accuracy and reliability of device status evaluation.

[0107] In yet another alternative embodiment, such as Figure 4 As shown, the acquisition module 301 is also used to acquire the rated operating parameters corresponding to the target device; The determination module 303 is also used to determine the operating procedure corresponding to the target device based on the equipment status assessment results and rated operating parameters corresponding to the target device; The acquisition module 301 is also used to acquire real-time operating information corresponding to the target device; The judgment module 308 is also used to determine whether the real-time running information matches the running program; The determining module 303 is also used to determine the safe operating range parameters of the target device based on the real-time operating information and the rated operating parameters when the judging module 308 determines that the real-time operating information does not match the operating program. The generation module 305 is also used to generate operation warning information and operation warning control parameters corresponding to the target device based on the safe operation range parameters; The control module 310 is used to transmit the operation warning information to the target terminal corresponding to the target device, and control the target device to perform the warning control operation that matches the warning control parameters.

[0108] It is evident that implementation Figure 4The described device can acquire the rated operating parameters of the target equipment and determine the corresponding operating procedure based on the equipment status assessment results and the rated operating parameters. It acquires the real-time operating information of the target equipment to determine whether it matches the operating procedure. If they do not match, it determines the safe operating range parameters based on the real-time operating information and the rated operating parameters. Based on the safe operating range parameters, it generates operating warning information and operating warning control information for the target equipment and transmits the operating warning information to the target terminal corresponding to the target equipment. It also controls the target equipment to execute warning control operations that match the warning control parameters. This device can comprehensively determine the corresponding operating procedure based on the equipment status assessment results and the rated operating parameters, which helps improve the accuracy and reliability of determining the corresponding operating procedure for the target component, and also helps improve the efficiency of determining the operating procedure. The accuracy and reliability of the system are enhanced by the ability to collect real-time data on the speed, torque, and working time of the root canal motor to determine the compatibility with the operating program. This improves the precision of judging whether real-time operating information matches the operating program. Furthermore, it allows for the comprehensive determination of safe operating range parameters based on real-time operating information and rated operating parameters, thus improving the accuracy and reliability of determining these parameters, as well as their intelligence and efficiency. This, in turn, enhances the accuracy and reliability of generating operating warning information and operating warning control parameters, thereby improving the safety and reliability of the target equipment. Finally, it enables the use of artificial intelligence to assess equipment status, improving the intelligence and efficiency of equipment status assessment, as well as its accuracy and reliability.

[0109] In yet another alternative embodiment, such as Figure 4 As shown, the specific method by which the determining module 303 determines the average prediction parameter corresponding to the target feature information based on each model feature information and the information weight parameter corresponding to each model feature information, combined with the pre-determined random regression model, includes: Based on each model feature information and the information weight parameter corresponding to each model feature information, a feature subset partitioning operation is performed on the target feature information to obtain at least one feature information subset. Based on each subset of feature information, a mapping relationship is generated between each model feature information and the information weight parameters corresponding to each model feature information. Based on the mapping relationship and the pre-determined random forest regression model, the average prediction parameters corresponding to the target feature information are determined. The model characteristic information includes equipment usage frequency characteristics, equipment workload characteristics, equipment operation habit characteristics, and equipment material characteristics.

[0110] It is evident that implementation Figure 4The described device can perform feature subset partitioning on target feature information based on each model characteristic information and its corresponding information weight parameters, obtaining at least one feature information subset. It then generates a mapping relationship between each model characteristic information and its corresponding information weight parameters, and combines this with a random forest regression model to determine the average prediction parameters corresponding to the target feature information. By combining features from multiple dimensions and their corresponding information weight parameters, the determined mapping relationship better reflects the actual wear patterns of the target equipment. Furthermore, by combining information weight parameters and employing weighted random sampling when partitioning feature subsets, it ensures that high-weight features appear in more subsets, improving the accuracy and reliability of subsequent mapping relationship generation and average prediction parameter determination. The weighted sampling ensures subset diversity, addressing the problem of overfitting of a single model leading to unstable predictions in existing technologies, thus improving the stability and reliability of the average prediction parameters. Additionally, based on the mapping relationship, it ensures that each tree in the random forest regression model prioritizes high-weight features during splitting, further enhancing the stability of the average prediction parameters. Finally, it can combine artificial intelligence to evaluate equipment status, improving the intelligence and efficiency of equipment status evaluation, as well as its accuracy and reliability.

[0111] In yet another alternative embodiment, such as Figure 4 As shown, the judgment module 308 is also used to determine whether the equipment status assessment result is used to indicate that the target device does not meet the preset equipment operating conditions based on the equipment status assessment result corresponding to the target device. The acquisition module 301 is also used to acquire real-time environmental information of the current environment where the target device is located when the judgment module 308 determines that the device status evaluation result indicates that the target device does not meet the preset device operating conditions. The generation module 305 is also used to generate equipment status warning information for the target device based on the equipment status assessment results, and to generate target status warning information for the target device based on the equipment status warning information and real-time environmental information. The determination module 303 is also used to determine the early warning mode parameters corresponding to the target device based on real-time environmental information; The generation module 305 is also used to generate device warning parameters for the target device based on the target status warning information and warning method parameters, and transmit the device warning parameters to the warning receiving terminal corresponding to the target device.

[0112] It is evident that implementation Figure 4The described device can determine whether the equipment status assessment result indicates that the target equipment does not meet the preset equipment operating conditions. If so, it acquires real-time environmental information and generates equipment status warning information and target status warning information. It then determines the warning method parameters, generates equipment warning parameters for the target equipment based on the target status warning information and warning method parameters, and transmits these parameters to the corresponding warning receiving terminal of the target equipment. This device can generate equipment status warning information based on the equipment status assessment result, which improves the accuracy and reliability of the generated equipment status warning information, as well as its intelligence and efficiency. Furthermore, it can combine real-time environmental information to assess the impact on the equipment status, allowing for the comprehensive generation of corresponding warning information by incorporating multiple factors, thus improving the comprehensiveness, accuracy, and reliability of the generated equipment warning information. It can also combine real-time environmental information to determine the corresponding warning method, improving the matching degree between the determined warning method and the current environment, and enhancing the convenience for users to view the warning information. This improves the safety and reliability of equipment operation and user operation of the target equipment. Finally, it can combine artificial intelligence to assess equipment status, improving the intelligence and efficiency of equipment status assessment, as well as its accuracy and reliability.

[0113] In yet another alternative embodiment, such as Figure 4 As shown, the processing module 302 performs information preprocessing operations on the target feature information to obtain the model characteristic information corresponding to the target feature information, and the specific method for obtaining the information weight parameters corresponding to each model characteristic information includes: Perform information cleaning operation on the target feature information to obtain the feature information cleaning result, and perform information classification operation on the feature information cleaning result to obtain at least one information category, wherein the information category includes numerical information category and label information category; Based on the information category, the target information category is determined, and the model characteristic information that matches the target information category is determined from all the feature information cleaning results included in the target information category. Based on the pre-determined convolutional neural network, the local correlation information corresponding to the model characteristic information is extracted. Based on the local correlation information and the model characteristic information, the feature temporal transformation correlation relationship is determined. Based on the correlation of feature temporal transformation, determine the information weight parameters corresponding to each model characteristic information; Among them, the feature temporal transformation correlation includes feature dynamic correlation, temporal attenuation coefficient correlation, and material compatibility correlation.

[0114] It is evident that implementation Figure 4The described apparatus can perform information cleaning operations on target feature information to obtain the cleaned feature information results, and perform information classification operations to obtain at least one information category and determine the target information category, thereby determining model characteristic information and extracting local correlation information by combining convolutional neural networks. Based on the local correlation information and model characteristic information, it determines the feature temporal transformation correlation relationship, and determines the corresponding information weight coefficients based on the feature temporal transformation correlation relationship. The information cleaning operation can remove invalid data, ensuring data authenticity and avoiding subsequent model prediction deviations due to abnormal data. This improves the accuracy and reliability of determining information weight coefficients, thereby improving the accuracy and reliability of subsequent target evaluation model determination and the determination of target device status evaluation results based on the target evaluation model. Furthermore, it can classify the cleaned results to avoid errors caused by data type mismatches. The model error reporting further improves the accuracy and reliability of subsequent target evaluation model determination and target equipment status assessment results based on the target evaluation model. It can also combine convolutional neural networks to extract local correlations, thereby improving the accuracy and reliability of determining feature time-series transformation correlations. This, in turn, improves the accuracy and reliability of subsequent target evaluation model determination and target equipment status assessment results based on the target evaluation model. Furthermore, based on feature time-series transformation correlations, the model can better fit actual loss patterns, which is conducive to further achieving real-time and intelligent accuracy and reliability of target equipment status assessment. In addition, it can combine artificial intelligence to assess equipment status, which is conducive to improving the intelligence and efficiency of equipment status assessment, as well as improving the accuracy and reliability of equipment status assessment.

[0115] Example 4 Please see Figure 5 , Figure 5 This is a schematic diagram of another device status assessment apparatus based on artificial intelligence disclosed in an embodiment of the present invention. Figure 5 As shown, the AI-based device status assessment device may include: Memory 401 storing executable program code; Processor 402 coupled to memory 401; The processor 402 calls the executable program code stored in the memory 401 to execute some or all of the steps in any of the artificial intelligence-based device status assessment methods in Embodiment 1 of the present invention.

[0116] Example 5 This invention discloses a computer storage medium storing computer instructions. When these computer instructions are invoked, they are used to execute some or all of the steps in any of the artificial intelligence-based device status assessment methods disclosed in Embodiment 1 of this invention.

[0117] Finally, it should be noted that the above embodiments are merely preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A device condition assessment method based on artificial intelligence, characterized in that, The method includes: Obtain target feature information of the target device, perform information preprocessing operation on the target feature information to obtain model characteristic information corresponding to the target feature information, and information weight parameters corresponding to each model characteristic information; Based on each of the model characteristic information and the information weight parameters corresponding to each of the model characteristic information, the average prediction parameters corresponding to the target feature information are determined in combination with a pre-determined random regression model. Based on the average prediction parameters, a training operation is performed on the initial evaluation model to obtain the model training result. Updated feature information is generated based on the model training result. Based on the updated feature information, a model iteration operation is performed on the initial evaluation model to obtain the target evaluation model. Obtain the device operation information of the target device, input the device operation information into the target evaluation model, and obtain the device status evaluation result corresponding to the target device.

2. The equipment status assessment method based on artificial intelligence according to claim 1, characterized in that, After inputting the equipment operation information into the target evaluation model to obtain the equipment status evaluation result corresponding to the target equipment, the method further includes: Based on the equipment status assessment results and the pre-determined cross-validation optimization model, determine whether the target assessment model meets the preset model assessment conditions. When it is determined that the target evaluation model does not meet the preset model evaluation conditions, the model update parameters corresponding to the target evaluation model are determined, and a model update operation is performed on the target evaluation model based on the model update parameters to obtain an updated evaluation model.

3. The equipment status assessment method based on artificial intelligence according to claim 1, characterized in that, The generation of updated feature information based on the model training results includes: Based on the model training results, the model training results are uploaded to a predetermined target cloud to obtain model feedback information from the target cloud. Based on the model feedback information, the model parameters to be adjusted in the initial evaluation model and the parameter adjustment factor corresponding to each model parameter to be adjusted are determined. Updated feature information is generated according to each model parameter to be adjusted and the parameter adjustment factor corresponding to each model parameter to be adjusted.

4. The equipment status assessment method based on artificial intelligence according to claim 2, characterized in that, The method further includes: Obtain the rated operating parameters corresponding to the target device, and determine the operating procedure corresponding to the target device based on the device status assessment result and the rated operating parameters. Obtain the real-time operating information corresponding to the target device, and determine whether the real-time operating information matches the operating program; When it is determined that the real-time operation information does not match the operation program, the safe operation range parameters of the target device are determined based on the real-time operation information and the rated operation parameters. Based on the safe operation range parameters, operation warning information and operation warning control parameters corresponding to the target device are generated. The operation warning information is transmitted to the target terminal corresponding to the target device, and the target device is controlled to perform a warning control operation that matches the warning control parameters.

5. The equipment status assessment method based on artificial intelligence according to claim 1, characterized in that, The step of determining the average prediction parameter corresponding to the target feature information based on each model feature information and the information weight parameter corresponding to each model feature information, combined with a pre-determined random regression model, includes: Based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, a feature subset partitioning operation is performed on the target feature information to obtain at least one feature information subset. Based on each subset of the feature information, a mapping relationship is generated between each model feature information and the information weight parameter corresponding to each model feature information. Based on the mapping relationship and the pre-determined random forest regression model, the average prediction parameter corresponding to the target feature information is determined. The model characteristic information includes equipment usage frequency characteristics, equipment workload characteristics, equipment operation habit characteristics, and equipment material characteristics.

6. The equipment status assessment method based on artificial intelligence according to claim 2, characterized in that, The method further includes: Based on the equipment status assessment result corresponding to the target equipment, determine whether the equipment status assessment result is used to indicate that the target equipment does not meet the preset equipment operating conditions; When it is determined that the device status assessment result indicates that the target device does not meet the preset device operating conditions, real-time environmental information of the current environment where the target device is located is obtained; Based on the equipment status assessment results, generate equipment status early warning information for the target equipment, and generate target status early warning information for the target equipment based on the equipment status early warning information and the real-time environmental information, and determine the early warning method parameters corresponding to the target equipment based on the real-time environmental information. Based on the target status warning information and the warning method parameters, device warning parameters for the target device are generated, and the device warning parameters are transmitted to the warning receiving terminal corresponding to the target device.

7. The equipment status assessment method based on artificial intelligence according to claim 1, characterized in that, The step of performing information preprocessing on the target feature information to obtain model characteristic information corresponding to the target feature information and information weight parameters corresponding to each model characteristic information includes: An information cleaning operation is performed on the target feature information to obtain a feature information cleaning result, and an information classification operation is performed on the feature information cleaning result to obtain at least one information category, wherein the information category includes a numerical information category and a label information category; Based on the information category, a target information category is determined, and model characteristic information matching the target information category is determined from all the feature information cleaning results included in the target information category. Local correlation information corresponding to the model characteristic information is extracted based on a pre-determined convolutional neural network. Based on the local correlation information and the model characteristic information, the feature temporal transformation correlation relationship is determined. Based on the correlation of the feature time-series transformation, determine the information weight parameter corresponding to each of the model characteristic information; The feature temporal transformation correlation relationship includes feature dynamic correlation relationship, temporal attenuation coefficient relationship, and material compatibility correlation relationship.

8. An artificial intelligence-based equipment status assessment device, characterized in that, The device includes: The acquisition module is used to acquire target feature information of the target device; The processing module is used to perform information preprocessing operations on the target feature information to obtain model characteristic information corresponding to the target feature information, and information weight parameters corresponding to each model characteristic information; The determination module is used to determine the average prediction parameter corresponding to the target feature information based on each of the model feature information and the information weight parameter corresponding to each of the model feature information, combined with a pre-determined random regression model. The training module is used to perform training operations on the initial evaluation model based on the average prediction parameters to obtain the model training results; The generation module is used to generate updated feature information based on the model training results; An iteration module is used to perform model iteration operations on the initial evaluation model based on the updated feature information to obtain the target evaluation model; The acquisition module is also used to acquire the device operation information of the target device; The input module is used to input the device operation information into the target evaluation model to obtain the device status evaluation result corresponding to the target device.

9. An artificial intelligence-based equipment status assessment device, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the artificial intelligence-based device status assessment method as described in any one of claims 1-7.

10. A computer storage medium, characterized in that, The computer storage medium stores computer instructions, which, when invoked, are used to execute the artificial intelligence-based device status assessment method as described in any one of claims 1-7.

Images