Transfer learning method combining multi-dimensional similarity measure with adaptive weight adjustment

Abstract

The application discloses a multi-dimensional similarity measurement combined with adaptive weight adjustment migration learning method, relates to the technical field of model training, and comprises the following steps: for any mechanical equipment to be detected, at least two historical maintenance records of the mechanical equipment to be detected in different maintenance institutions are called, and all the historical maintenance records are integrated and cleaned to generate maintenance data of the mechanical equipment to be detected; feature extraction is performed on the maintenance data and different damage category sample data by using a pre-trained model to obtain a first feature vector and a second feature vector; similarity between the maintenance data and the different damage category sample data is calculated according to the first feature vector and the second feature vector; and the service life of the damage category corresponding to the damage category sample data with the highest similarity is taken as a detection result of the mechanical equipment to be detected. The application can effectively identify potential problems of the mechanical equipment to be detected, and improve the accuracy and efficiency of detection.

Description

Technical Field

[0001] This invention relates to the field of model training technology, and in particular to a transfer learning method that combines multi-dimensional similarity measurement with adaptive weight adjustment. Background Technology

[0002] In the field of mechanical equipment maintenance, the amount and characteristics of data for different types of damage vary significantly, and data acquisition costs are high. Traditional single-damage model training methods struggle to fully utilize limited data resources, resulting in limitations in model generalization ability and diagnostic accuracy. Existing transfer learning methods often lack a refined understanding of the relevance and importance of samples when processing source and target data, making it difficult to effectively utilize useful information in the source data. This affects the model's adaptability and diagnostic accuracy on the target data, further leading to the inability to accurately determine the damage category or condition of mechanical equipment based on the source data. Summary of the Invention

[0003] The technical problem to be solved by this invention is to address the shortcomings of existing technologies. Specifically, it provides a transfer learning method that combines multi-dimensional similarity measurement with adaptive weight adjustment, as detailed below:

[0004] 1) In a first aspect, the present invention provides a transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment, the specific technical solution of which is as follows:

[0005] For any piece of machinery to be tested, retrieve at least two historical maintenance records of the machinery to be tested from different maintenance institutions, and integrate and clean all historical maintenance records to generate maintenance data for the machinery to be tested.

[0006] The maintenance data and different damage category sample data are subjected to feature extraction by a pre-trained model to obtain a first feature vector corresponding to the maintenance data and a second feature vector corresponding to the different damage category sample data. Based on the first feature vector and the second feature vector, the similarity between the maintenance data and the different damage category sample data is calculated. The service life of the damage category corresponding to the damage category sample data with the highest similarity is taken as the detection result of the mechanical equipment to be tested.

[0007] The weights of the pre-trained model are determined through adaptive adjustment.

[0008] The beneficial effects of the transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment provided by this invention are as follows:

[0009] First, by integrating and cleaning historical maintenance records from different repair organizations, comprehensive and high-quality maintenance data can be generated, providing a reliable data foundation for subsequent analysis and inspection. Second, feature extraction using pre-trained models can fully utilize the rich feature representation capabilities learned by pre-trained models on large-scale data, thereby more accurately extracting features from maintenance data and damage category sample data. Third, determining the damage category by calculating the similarity between feature vectors can effectively identify potential problems in the mechanical equipment under inspection, improving the accuracy and efficiency of inspection. Finally, the adaptive adjustment of pre-trained model weights can further optimize model performance, making it better adaptable to specific maintenance data and damage categories, further improving the reliability of inspection results.

[0010] Based on the above solution, the present invention can be further improved as follows.

[0011] Furthermore, the cleaning process specifically includes:

[0012] Abnormal data is removed, and data of different formats under the same project is processed uniformly.

[0013] Furthermore, the similarity includes: feature similarity, label similarity, and time series similarity.

[0014] Furthermore, the calculation process for the feature similarity is as follows:

[0015]

[0016] Among them, S feat (x i ,y j ) represents feature similarity, f(·) represents the feature vector extracted by the convolutional neural network, and x i Let y represent the first eigenvector. j This represents the second eigenvector.

[0017] Furthermore, the calculation process for the tag similarity is as follows:

[0018]

[0019] Among them, S label (x i ,y j ) represents tag similarity, g(·) represents the function that maps tag text to semantic space, and l i The label representing the maintenance data, l j Labels representing sample data corresponding to different damage categories.

[0020] Furthermore, the calculation process for the time series similarity is as follows:

[0021]

[0022] Among them, S time (x i ,y j ) represents time series similarity, DTW(T i ,T j ) represents the time series T i and T j The dynamic time-warped distance between them, T i T represents the time series corresponding to the maintenance data. j This represents the time series corresponding to sample data of different damage categories.

[0023] 2) Secondly, the present invention also provides a transfer learning system combining multi-dimensional similarity measurement and adaptive weight adjustment, the specific technical solution of which is as follows:

[0024] The processing module is used to: retrieve at least two historical maintenance records of any mechanical equipment to be tested from different maintenance institutions, and integrate and clean all historical maintenance records to generate maintenance data for the mechanical equipment to be tested;

[0025] The detection module is used to: extract features from the maintenance data and sample data of different damage categories using a pre-trained model to obtain a first feature vector corresponding to the maintenance data and a second feature vector corresponding to the sample data of different damage categories; calculate the similarity between the maintenance data and sample data of different damage categories based on the first feature vector and the second feature vector; and take the service life of the damage category corresponding to the damage category sample data with the highest similarity as the detection result of the mechanical equipment to be detected.

[0026] The weights of the pre-trained model are determined through adaptive adjustment.

[0027] 3) In a third aspect, the present invention also provides an electronic device, the electronic device including a processor coupled to a memory, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor to enable the electronic device to perform any of the above methods.

[0028] 4) In a fourth aspect, the present invention also provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to perform any of the above methods.

[0029] It should be noted that the beneficial effects of the technical solutions of the second to fourth aspects of the present invention and their corresponding possible implementations can be found in the above description of the technical effects of the first aspect and its corresponding possible implementations, and will not be repeated here. Attached Figure Description

[0030] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0031] Figure 1 This is a flowchart illustrating a transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to an embodiment of the present invention.

[0032] Figure 2 This is a structural framework diagram of an electronic device according to the present invention. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0034] like Figure 1 As shown in the figure, a transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to an embodiment of the present invention includes the following steps:

[0035] S1. For any mechanical equipment to be tested, retrieve at least two historical maintenance records of the mechanical equipment to be tested from different maintenance institutions, and integrate and clean all historical maintenance records to generate maintenance data of the mechanical equipment to be tested.

[0036] S2, the maintenance data and sample data of different damage categories are extracted by a pre-trained model to obtain the first feature vector corresponding to the maintenance data and the second feature vector corresponding to the sample data of different damage categories. Based on the first feature vector and the second feature vector, the similarity between the maintenance data and the sample data of different damage categories is calculated. The service life of the damage category corresponding to the damage category sample data with the highest similarity is taken as the detection result of the mechanical equipment to be tested.

[0037] The weights of the pre-trained model are determined through adaptive adjustment.

[0038] The beneficial effects of the transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment provided by this invention are as follows:

[0039] First, by integrating and cleaning historical maintenance records from different repair organizations, comprehensive and high-quality maintenance data can be generated, providing a reliable data foundation for subsequent analysis and inspection. Second, feature extraction using pre-trained models can fully utilize the rich feature representation capabilities learned by pre-trained models on large-scale data, thereby more accurately extracting features from maintenance data and damage category sample data. Third, determining the damage category by calculating the similarity between feature vectors can effectively identify potential problems in the mechanical equipment under inspection, improving the accuracy and efficiency of inspection. Finally, the adaptive adjustment of pre-trained model weights can further optimize model performance, making it better adaptable to specific maintenance data and damage categories, further improving the reliability of inspection results.

[0040] Different maintenance organizations refer to any organization that can inspect or repair mechanical equipment and generate maintenance records.

[0041] Mechanical equipment can be used for diagnostic purposes:

[0042] Imaging equipment: including X-ray machines, CT scanners, MRI (magnetic resonance imaging) equipment, ultrasound diagnostic instruments, etc., used to acquire images of the internal structure and function of the human body.

[0043] Endoscopic equipment, such as fiber optic endoscopes and electronic endoscopes, is used to examine internal organs of the human body.

[0044] In vitro diagnostic equipment: including laboratory testing equipment and point-of-care testing equipment, used to test samples such as blood and urine.

[0045] Treatment equipment:

[0046] Surgical instruments: including general surgical instruments, optical surgical instruments (such as laser therapy machines), and auxiliary surgical instruments (such as anesthesia machines, ventilators, and extracorporeal circulation equipment).

[0047] Radiotherapy equipment: such as deep X-ray therapy machine, cobalt-60 therapy machine, accelerator, gamma knife, etc.

[0048] Interventional treatment equipment: such as angiography machines, interventional catheters, etc.

[0049] Rehabilitation therapy equipment: such as rehabilitation robots, physiotherapy instruments, functional training equipment, etc.

[0050] Patient monitoring equipment

[0051] Vital signs monitoring equipment: such as electrocardiogram (ECG) machines, electroencephalogram (EEG) machines, electromyogram (EMG) machines, multi-parameter monitors, etc.

[0052] Respiratory support equipment: such as ventilators, oxygen concentrators, oxygen cylinders, etc.

[0053] The process of integrating and cleaning all historical maintenance records to generate maintenance data for the mechanical equipment under inspection is as follows:

[0054] The integration process is as follows:

[0055] Based on the unique serial number of the mechanical equipment to be tested, all data related to that unique serial number are integrated and processed.

[0056] The cleaning process is as follows:

[0057] Remove duplicate data: Check if there are duplicate records in the dataset consisting of unique labels and their corresponding data, and delete them.

[0058] The criteria for determining duplicate records are: maintenance records or inspection records of the same mechanical equipment at different maintenance institutions within the same time period.

[0059] The same time period refers to a preset period, such as 1 hour or 6 hours.

[0060] Handling missing data: For missing data, fill in the missing data.

[0061] The specific filling process is as follows:

[0062] The missing data category is determined. If it is time-related or maintenance cost-related data, the time series of the dataset is used to determine the start and end times of the target time range for the missing data, based on the time corresponding to the column preceding and following the column containing the missing data. In the historical database, the top n sets of relevant historical data with the highest similarity to the type and model of the machinery to be inspected are retrieved. The maintenance cycle corresponding to the maintenance type of the missing data is determined from these n sets of relevant historical data. Based on this maintenance cycle, the first maintenance record under this maintenance type is determined, and the target time overlapping with the maintenance cycle within the target time range is identified. This target time is then used as the filling data for filling. The maintenance cost is determined based on the filling results of the time-related data.

[0063] When the missing data is classified as damaged area data or component replacement data, feature extraction is performed on the current image information of the mechanical equipment to be inspected to determine the repair area, and the damaged area is determined by the repair area. The component replacement information is obtained by comparing the current image information with the original image information of the mechanical equipment to be inspected.

[0064] The process of retrieving the top n sets of relevant historical data that are most similar to the type and model of the mechanical equipment to be tested is as follows:

[0065] Determine whether there are historical mechanical devices in the historical database that are the same or similar to the mechanical device to be tested. Similarity refers to the same function and similar size, or the top m historical mechanical devices that have the most identical parts to the mechanical device to be tested.

[0066] Search for the first historical mechanical equipment with the same or similar equipment model as the mechanical equipment to be tested. Similarity means that there is a generational relationship between the equipment model and the mechanical equipment to be tested (such as the first generation, second generation and third generation of a certain machine).

[0067] Based on the component corresponding to the missing data and the names of other first components adjacent to it, the component and the first components are labeled, and the weights are adjusted to obtain the first weights corresponding to the component and the first components. Based on the first weights and the second weights (preset weights) corresponding to different components, the scores corresponding to all first historical mechanical equipment are calculated, that is, the similarity scores. The similarity calculation results are arranged in descending order, and the relevant historical data corresponding to the first historical mechanical equipment with the first n similarities are output.

[0068] The process of determining the maintenance cycle corresponding to the missing data type from the n sets of relevant historical data is as follows:

[0069] Determine the first maintenance cycle for the maintenance type corresponding to the first historical mechanical equipment, and sort all the first maintenance cycles according to the descending order results above. Calculate the target maintenance cycle corresponding to the maintenance type based on the second weight corresponding to the descending order results and the first maintenance cycle. The target maintenance cycle is determined as the maintenance cycle for the maintenance type corresponding to the missing data in the n sets of relevant historical data.

[0070] Remove abnormal data: Based on the equipment's normal operating range and maintenance experience, identify and remove abnormal data. For example, remove data that exceeds the equipment's normal operating parameter range or data with obviously unreasonable maintenance costs.

[0071] Data consistency check: Check whether the data is logically consistent. For example, the repair time should be later than the time of the failure, and the repair cost should match the repair content.

[0072] Data validation and quality assessment

[0073] Verify data accuracy: Verify the accuracy of the data by comparing it with actual maintenance records, equipment operating status, etc.

[0074] Assess data quality: Conduct a quality assessment of the cleaned data to ensure its integrity, accuracy, and consistency.

[0075] The test result refers to the service life of the mechanical equipment under test.

[0076] Furthermore, the cleaning process specifically includes:

[0077] Abnormal data is removed, and data of different formats under the same project is processed uniformly.

[0078] Furthermore, the similarity includes: feature similarity, label similarity, and time series similarity.

[0079] Furthermore, the calculation process for the feature similarity is as follows:

[0080]

[0081] Among them, S feat (x i ,y j ) represents feature similarity, f(·) represents the feature vector extracted by the convolutional neural network, and x i Let y represent the first eigenvector. j This represents the second eigenvector.

[0082] Furthermore, the calculation process for the tag similarity is as follows:

[0083]

[0084] Among them, S label (x i ,y j ) represents tag similarity, g(·) represents the function that maps tag text to semantic space, and l i The label representing the maintenance data, l j Labels representing sample data corresponding to different damage categories.

[0085] Furthermore, the calculation process for the time series similarity is as follows:

[0086]

[0087] Among them, S time (x i ,y j ) represents time series similarity, DTW(T i ,T j ) represents the time series T i and T j The dynamic time-warped distance between them, T i T represents the time series corresponding to the maintenance data. j This represents the time series corresponding to sample data of different damage categories.

[0088] Example 1, Part 1: Data Cleaning

[0089] 1. Remove duplicate records: When integrating data from multiple repair shops, duplicate records may occur because different repair shops may have recorded the same basic information about the machinery and equipment, such as unique identifiers, repair details, inspection items, and costs. Data cleaning can identify and remove these duplicate records, ensuring the uniqueness and accuracy of the data.

[0090] 2. Correcting Erroneous Data: Some data may contain incorrectly entered inspection indicator values. By setting reasonable threshold ranges for these values, data exceeding the range can be marked and corrected, thereby improving data quality.

[0091] 3. Standardize Data Formats: Different data may have different record formats, such as date formats like "YYYY-MM-DD" or "MM / DD / YYYY". During data migration, these date formats need to be standardized into a single format to facilitate subsequent data processing and analysis.

[0092] 4. Numerical normalization: For numerical data with different dimensions, in order to eliminate the influence of dimensions, the Min-Max normalization method is used to uniformly map these values ​​to the [0,1] interval, so that different data are numerically comparable.

[0093] II. Multi-dimensional similarity calculation

[0094] It is responsible for calculating the similarity between source data and target data samples across multiple dimensions, including features, labels, and time series.

[0095] The specific steps are as follows:

[0096] 1. For feature similarity calculation, a pre-trained convolutional neural network is used to extract the feature vectors of the samples, and then cosine similarity is calculated. For example, for source data sample x... i and target damage category sample y j Its feature similarity S feat (x i ,y j This can be represented as:

[0097]

[0098] Where f(·) represents the feature vector extracted by the convolutional neural network.

[0099] 2. For label similarity calculation, the sample label text is converted into semantic vectors, and the cosine similarity method is used. Let the source data sample x... i The tag is l i Target damage category sample yj The label is y j Its tag similarity S label (x i ,y j This can be represented as:

[0100]

[0101] Here, g(·) represents the function that maps the labeled text to the semantic space, using word vectors trained using the Word2Vec model.

[0102] 3. For time series similarity calculation, the dynamic time warping algorithm is used. Let the source data sample x... i The time series is T i ={t i1 ,t i2 ,..t in}, Target damage category sample y j The time series is T j ={t j1 ,t j2 ,..t jn}, their time series similarity S time (x i ,y j This can be represented as:

[0103]

[0104] Among them, DTW(T i ,T j ) represents the time series T i and T j The dynamic time-normalized distance between them.

[0105] III. Adaptive Weight Adjustment

[0106] Based on the multi-dimensional similarity calculation results and model training feedback, the weights of the source data and the target damaged category samples are dynamically adjusted. The specific algorithm is as follows:

[0107] 1. Initialize sample weights: Based on prior knowledge of the source data and the target damage category, assign weights to each source data sample x. i and target damage category sample y j Set initial weights and For example, source data samples that are highly similar to the characteristics of the target damage category are given relatively high initial weights.

[0108] 2. During model training, the contribution of each sample to model training is evaluated by monitoring metrics such as the model's loss function value L and accuracy A. For the k-th iteration, the update amount of the sample weights is calculated. and

[0109]

[0110] Where η is the learning rate. and These are the gradients of the loss function with respect to the weights of the source data samples and the target damaged category samples, respectively.

[0111] 3. Combining reinforcement learning mechanisms, the sample weights are adjusted based on changes in the model's diagnostic accuracy across target damage categories. The reward function R is defined as:

[0112] R = A current -A previous

[0113] Among them, A current and A previous These represent the diagnostic accuracy of the current iteration and the previous iteration on the target damage category, respectively.

[0114] Update the sample weights based on the reward signal R:

[0115]

[0116] Where α is the learning rate of reinforcement learning.

[0117] The above method can quickly identify the final target damage category sample and output the service life corresponding to the damage category sample as the detection result.

[0118] In the above embodiments, although the steps are numbered S1, S2, etc., they are only specific embodiments given by the present invention. Those skilled in the art can adjust the execution order of S1, S2, etc. according to the actual situation, which is also within the protection scope of the present invention. It can be understood that in some embodiments, some or all of the above embodiments may be included.

[0119] This invention also provides a transfer learning system that combines multi-dimensional similarity measurement with adaptive weight adjustment, the specific technical solution of which is as follows:

[0120] The processing module is used to: retrieve at least two historical maintenance records of any mechanical equipment to be tested from different maintenance institutions, and integrate and clean all historical maintenance records to generate maintenance data for the mechanical equipment to be tested.

[0121] The detection module is used to: extract features from the maintenance data and sample data of different damage categories using a pre-trained model to obtain a first feature vector corresponding to the maintenance data and a second feature vector corresponding to the sample data of different damage categories; calculate the similarity between the maintenance data and sample data of different damage categories based on the first feature vector and the second feature vector; and take the service life of the damage category corresponding to the damage category sample data with the highest similarity as the detection result of the mechanical equipment to be detected.

[0122] The weights of the pre-trained model are determined through adaptive adjustment.

[0123] It should be noted that the beneficial effects of the transfer learning system combining multi-dimensional similarity measurement and adaptive weight adjustment provided in the above embodiments are the same as those of the transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment, and will not be repeated here. Furthermore, the system provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the system can be divided into different functional modules according to the actual situation to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process is detailed in the method embodiments, and will not be repeated here.

[0124] like Figure 2 As shown, an electronic device 300 according to an embodiment of the present invention includes a processor 320 coupled to a memory 310. The memory 310 stores at least one computer program 330, which is loaded and executed by the processor 320 to enable the electronic device 300 to implement any of the above-mentioned methods. Specifically:

[0125] The electronic device 300 can vary considerably due to differences in configuration or performance. It may include one or more processors 320 (Central Processing Units, CPUs) and one or more memories 310. The memories 310 store at least one computer program 330, which is loaded and executed by the processors 320 to enable the electronic device 300 to implement the multi-dimensional similarity metric combined with adaptive weight adjustment transfer learning method provided in the above embodiments. Of course, the electronic device 300 may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. It may also include other components for implementing device functions, which will not be elaborated upon here.

[0126] An embodiment of the present invention provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to implement any of the above-described methods.

[0127] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0128] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform any of the methods described above.

[0129] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0130] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this disclosure can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product in one or more computer-readable media containing computer-readable program code.

[0131] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0132] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment, characterized in that, include: For any piece of machinery to be tested, retrieve at least two historical maintenance records of the machinery to be tested from different maintenance institutions, and integrate and clean all historical maintenance records to generate maintenance data for the machinery to be tested. The maintenance data and different damage category sample data are subjected to feature extraction by a pre-trained model to obtain a first feature vector corresponding to the maintenance data and a second feature vector corresponding to the different damage category sample data. Based on the first feature vector and the second feature vector, the similarity between the maintenance data and the different damage category sample data is calculated. The service life of the damage category corresponding to the damage category sample data with the highest similarity is taken as the detection result of the mechanical equipment to be tested. The weights of the pre-trained model are determined through an adaptive adjustment method. The data cleaning process includes: handling missing data: filling in missing data; The specific filling process is as follows: Determine the data category of the missing data. If it is time-related data or maintenance cost data, combine the time series of the dataset and use the time corresponding to the column before and after the column containing the missing data as the start and end times of the target time range for the missing data. In the historical database, retrieve the top n sets of relevant historical data with the highest similarity to the type and model of the mechanical equipment to be tested. In these n sets of relevant historical data, determine the maintenance cycle of the maintenance type corresponding to the missing data. Based on the maintenance cycle, determine the first maintenance record under this maintenance type, and determine the target time that overlaps with the maintenance cycle within the target time range. Use this target time as the filling data and perform filling processing accordingly. Determine the maintenance cost based on the filling results of the time-related data. When the missing data is classified as damaged area data or component replacement data, feature extraction is performed on the current image information of the mechanical equipment to be inspected to determine the repair area, and the damaged area is determined by the repair area. The component replacement information is obtained by comparing the current image information with the original image information of the mechanical equipment to be inspected.

2. The transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to claim 1, characterized in that, The cleaning process is specifically as follows: Abnormal data is removed, and data of different formats under the same project is processed uniformly.

3. The transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to claim 1, characterized in that, The similarity includes: feature similarity, label similarity, and time series similarity.

4. The transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to claim 3, characterized in that, The specific process for calculating the feature similarity is as follows: in, Indicates feature similarity. This represents the feature vector extracted by the convolutional neural network. Represents the first eigenvector. This represents the second eigenvector.

5. The transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to claim 3, characterized in that, The specific process for calculating the tag similarity is as follows: in, Indicates label similarity. This represents a function that maps label text to a semantic space. The tags that represent the maintenance data Labels representing sample data corresponding to different damage categories.

6. The transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment according to claim 3, characterized in that, The calculation process for the time series similarity is as follows: in, Indicates time series similarity. Representing time series and The dynamic time-normalized distance between them This represents the time series corresponding to the maintenance data. This represents the time series corresponding to sample data of different damage categories.

7. A transfer learning system combining multi-dimensional similarity measurement and adaptive weight adjustment, employing the transfer learning method combining multi-dimensional similarity measurement and adaptive weight adjustment as described in claim 1, characterized in that... include: The processing module is used to: retrieve at least two historical maintenance records of any mechanical equipment to be tested from different maintenance institutions, and integrate and clean all historical maintenance records to generate maintenance data for the mechanical equipment to be tested. The detection module is used to: extract features from the maintenance data and sample data of different damage categories using a pre-trained model to obtain a first feature vector corresponding to the maintenance data and a second feature vector corresponding to the sample data of different damage categories; calculate the similarity between the maintenance data and sample data of different damage categories based on the first feature vector and the second feature vector; and take the service life of the damage category corresponding to the damage category sample data with the highest similarity as the detection result of the mechanical equipment to be detected. The weights of the pre-trained model are determined through adaptive adjustment.

8. The transfer learning system combining multi-dimensional similarity measurement and adaptive weight adjustment according to claim 7, characterized in that, The cleaning process is specifically as follows: Abnormal data is removed, and data of different formats under the same project is processed uniformly.

9. An electronic device, characterized in that, The electronic device includes a processor coupled to a memory storing at least one computer program, which is loaded and executed by the processor to enable the electronic device to perform the method as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to enable the computer to perform the method as described in any one of claims 1 to 6.

Images