A device fault maintenance intelligent recommendation method

Abstract

This invention relates to an intelligent recommendation method for equipment fault repair. The method includes: constructing an initial equipment repair measure recommendation model; collecting historical fault data of the target equipment and training the initial model using this data to obtain a target equipment repair measure recommendation model; acquiring user-inputted target equipment fault information and inputting this information into the target equipment repair measure recommendation model to output a recommended repair measure scheme. This method, by constructing an initial equipment repair measure recommendation model, training it using historical fault data, and then acquiring user-inputted target equipment fault information and inputting it into the target equipment repair measure recommendation model to output a recommended repair measure scheme, achieves intelligent analysis of equipment faults and recommends repair strategies, improving repair efficiency and decision-making accuracy.

Description

Technical Field

[0001] This invention relates to the field of equipment operation and maintenance technology, and in particular to an intelligent recommendation method for equipment fault repair. Background Technology

[0002] In modern industrial production, equipment maintenance is a crucial link in ensuring continuous and stable operation. However, current equipment failure maintenance still relies heavily on the personal experience of maintenance personnel and expert judgment, which has many limitations. First, failure records are usually in the form of work orders, reports, and other texts in natural language, with diverse descriptions and fragmented information, making them difficult for computers to directly understand and structure. This results in a large amount of valuable historical maintenance data not being effectively integrated and utilized. Second, the maintenance decision-making process is highly subjective. Different personnel often have different analyses and solutions for similar failures, making it difficult to standardize and solidify experience, resulting in poor knowledge reusability and decision consistency. Furthermore, with the increasing complexity of equipment and the diversification of operating conditions, failure modes are becoming increasingly complex. Relying solely on manual analysis is not only inefficient but also prone to misjudgment due to experience limitations or fatigue, increasing the risk of unplanned downtime and maintenance costs. Therefore, the existing maintenance model is no longer suitable for the development needs of intelligent operation and maintenance. Summary of the Invention

[0003] This invention provides an intelligent recommendation method for equipment fault repair, which overcomes at least one of the above-mentioned technical problems existing in the prior art.

[0004] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions: In a first aspect, the present invention provides an intelligent recommendation method for equipment fault repair, comprising: Construct an initial equipment maintenance recommendation model; Collect historical fault data of the target equipment, and use the historical fault data to train the initial equipment maintenance measure recommendation model to obtain the target equipment maintenance measure recommendation model; Obtain the target equipment fault information input by the user, input the target equipment fault information into the target equipment maintenance measure recommendation model, and output the target equipment maintenance measure recommendation scheme.

[0005] In one possible implementation of the first aspect, the construction of the initial equipment maintenance measure recommendation model includes: An initial equipment maintenance measure recommendation model based on the Transformer architecture is constructed. The initial equipment maintenance measure recommendation model includes an input layer, an encoding layer, a feature fusion layer, and an output layer. The encoding layer includes a text encoding channel and a structured data encoding channel. The text encoding channel is used to extract text features of the target device fault information, and the structured data encoding channel is used to extract structured features of the target device fault information. The feature fusion layer is used to fuse the text features and the structured features to generate comprehensive features of equipment faults; The output layer is used to output recommended maintenance measures based on the comprehensive characteristics of the equipment fault.

[0006] In one possible implementation of the first aspect, the feature fusion layer is an attention-based feature fusion layer.

[0007] In one possible implementation of the first aspect, the output layer includes a fault severity classification head and a maintenance measure matching score head; The fault severity classification head outputs the defect severity level through a first fully connected layer and a first Softmax function; The maintenance measure matching score head outputs the maintenance measure matching score through the second fully connected layer and the second Softmax function.

[0008] In one possible implementation of the first aspect, prior to training the initial equipment maintenance measure recommendation model using the historical fault data, the method further includes: The historical fault data is preprocessed to generate a standardized dataset for training the initial equipment maintenance measure recommendation model; The preprocessing includes: The duplicate data in the historical fault data is deduplicated; Complete or correct data with missing key fields; Correct abnormal or non-compliant data; Standardize the fault description text; Convert fault classification information into numerical labels; and, Enhancement processing is applied to minority class samples.

[0009] In one possible implementation of the first aspect, training the initial equipment maintenance measure recommendation model using the historical fault data includes: The following composite loss function is used to guide the training of the initial equipment maintenance measure recommendation model: ; ; ; in, For classifying losses, The weights for each category c, One-hot encoding of the real label. To predict probabilities for the model, For the matching degree loss, For matching degree, For the true value of the matching degree, θ is the composite loss function; α and β are hyperparameters, λ is the regularization coefficient, and θ is all weight parameters of the model.

[0010] In one possible implementation of the first aspect, training the initial equipment maintenance measure recommendation model using the historical fault data further includes: The equipment diagnosis rules summarized by experts are transformed into supervised samples to assist in training the initial equipment maintenance measure recommendation model.

[0011] In one possible implementation of the first aspect, the step of inputting the target equipment fault information into the target equipment maintenance measure recommendation model and outputting a target equipment maintenance measure recommendation scheme includes: The fault information of the target equipment is input into multiple trained target equipment maintenance measure recommendation models. Each target equipment maintenance measure recommendation model outputs an initial target equipment maintenance measure recommendation scheme, which includes an initial target equipment fault category and an initial target equipment maintenance measure matching degree. Based on all the initial target equipment maintenance measure recommendations, generate the target equipment maintenance measure recommendation plan; Output a recommended maintenance plan for the target equipment.

[0012] In one possible implementation of the first aspect, generating the target equipment maintenance measure recommendation scheme based on all the initial target equipment maintenance measure recommendation schemes includes: Based on all the initial target equipment maintenance measure recommendations, a selected target equipment maintenance measure recommendation scheme is generated using a weighted average algorithm or a voting mechanism. The selected target equipment maintenance measure recommendation scheme includes the selected target equipment fault category and the selected target equipment maintenance measure matching degree. Determine whether the matching degree of the selected target equipment maintenance measures exceeds the preset maintenance measure matching degree threshold; If so, the recommended maintenance measures for the selected target equipment shall be used as the recommended maintenance measures for the target equipment. If not, then according to the fallback mechanism, the target equipment diagnosis rule corresponding to the target equipment fault information is matched from the preset expert rule base, and the target equipment maintenance case similar to the target equipment fault information is retrieved from the preset historical case base. Based on the target equipment diagnosis rule and the target equipment maintenance case, a recommended solution for the target equipment maintenance measures is generated.

[0013] In one possible implementation of the first aspect, after outputting the recommended maintenance measures for the target equipment, the method further includes: Obtain user feedback data on the recommended maintenance measures for the target equipment, and dynamically update the recommended maintenance measures model for the target equipment based on the feedback data.

[0014] Compared with the prior art, the present invention has at least the following beneficial effects: This invention provides an intelligent equipment fault repair recommendation method. It constructs an initial equipment repair measure recommendation model, then trains the initial model using historical fault data to obtain a target equipment repair measure recommendation model. Finally, it acquires user-inputted target equipment fault information, inputs this information into the target equipment repair measure recommendation model, and outputs a target equipment repair measure recommendation scheme. This method achieves intelligent analysis of equipment faults and recommends repair strategies, improving repair efficiency and decision-making accuracy.

[0015] Secondly, this invention provides an intelligent recommendation system for equipment fault repair, comprising: The model building module is used to build a recommendation model for initial equipment maintenance measures; The model training module is used to collect historical fault data of the target equipment, and use the historical fault data to train the initial equipment maintenance measure recommendation model to obtain the target equipment maintenance measure recommendation model. The intelligent recommendation module is used to obtain the target equipment fault information input by the user, input the target equipment fault information into the target equipment maintenance measure recommendation model, and output the target equipment maintenance measure recommendation scheme.

[0016] Thirdly, the present invention provides an electronic device comprising: at least one processor and at least one memory, wherein the memory stores computer-readable instructions; the computer-readable instructions are executed by one or more of the processors, causing the electronic device to implement the intelligent recommendation method for equipment fault repair as in any implementation of the first aspect.

[0017] Fourthly, the present invention provides a storage medium having a computer-executable program stored thereon, the computer-executable program being used to cause a computer to execute the intelligent recommendation method for equipment fault repair as in any implementation of the first aspect.

[0018] Understandably, the beneficial effects achieved by the system of the second aspect, the electronic device of the third aspect, and the storage medium of the fourth aspect provided above can be referred to in light of the beneficial effects of the first aspect and any of its possible design embodiments, which will not be repeated here. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention; Figure 2 A flowchart of an intelligent recommendation method for equipment fault repair provided in an embodiment of the present invention; Figure 3 This is a structural block diagram of an intelligent recommendation system for equipment fault repair provided in an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be described below with reference to the accompanying drawings. In the description of the present invention, unless otherwise stated, " / " indicates that the objects before and after are in an "or" relationship. For example, A / B can represent A or B. The "or" in the present invention is merely a description of the relationship between the related objects, indicating that three relationships can exist. For example, A or B can represent: A alone, A and B simultaneously, and B alone. A and B can be singular or plural. Furthermore, in the description of the present invention, unless otherwise stated, "multiple" refers to two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items.

[0022] Furthermore, in the embodiments of the present invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of the present invention should not be construed as superior or more advantageous than other embodiments or designs. Rather, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner to facilitate understanding.

[0023] In modern industrial production, equipment maintenance is a crucial link in ensuring continuous and stable operation. However, current equipment failure maintenance still relies heavily on the personal experience of maintenance personnel and expert judgment, which has many limitations. First, defect records are usually in the form of natural language in work orders, reports, and other texts, with diverse descriptions and fragmented information, making them difficult for computers to directly understand and structure. This results in a large amount of valuable historical maintenance data not being effectively integrated and utilized. Second, the maintenance decision-making process is highly subjective. Different personnel often have different analyses and solutions for similar defects, making it difficult to standardize and solidify experience, resulting in poor knowledge reusability and decision consistency. Furthermore, with the increasing complexity of equipment and the diversification of operating conditions, defect patterns are becoming increasingly complex. Relying solely on manual analysis is not only inefficient but also prone to misjudgment due to experience limitations or fatigue, increasing the risk of unplanned downtime and maintenance costs. Therefore, the existing maintenance model is no longer suitable for the development needs of intelligent operation and maintenance.

[0024] In view of this, on the one hand, embodiments of the present invention provide an intelligent recommendation method for equipment fault repair, comprising: constructing an initial equipment repair measure recommendation model; collecting historical fault data of the target equipment, using the historical fault data to train the initial equipment repair measure recommendation model to obtain a target equipment repair measure recommendation model; obtaining target equipment fault information input by the user, inputting the target equipment fault information into the target equipment repair measure recommendation model, and outputting a target equipment repair measure recommendation scheme.

[0025] This invention provides an intelligent equipment fault repair recommendation method. It constructs an initial equipment repair measure recommendation model, then trains the initial model using historical fault data to obtain a target equipment repair measure recommendation model. Finally, it acquires user-inputted target equipment fault information, inputs this information into the target equipment repair measure recommendation model, and outputs a target equipment repair measure recommendation scheme. This method achieves intelligent analysis of equipment faults and recommends repair strategies, improving repair efficiency and decision-making accuracy.

[0026] In some embodiments, the intelligent recommendation method for equipment fault repair provided by the present invention can be executed by any electronic device 20 with data processing capabilities, such as a general-purpose computer, personal computer, laptop computer, switch, or tablet computer, etc. The specific implementation of the electronic device 20 is not limited here.

[0027] Figure 1 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention is shown. The electronic device 20 includes a processor 210, a memory 220, and a communication interface 230.

[0028] Processor 210 may include one or more processing cores. Processor 210 connects to various parts within electronic device 200 using various interfaces and lines, and performs various functions and processes data of electronic device 200 by running or executing instructions, programs, code sets, or instruction sets stored in memory 220, and by calling data stored in memory 220. Optionally, processor 210 may be implemented using at least one of the following hardware forms: Central Processing Unit (CPU), Graphics Processing Unit (GPU), Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA).

[0029] The memory 220 may include random access memory (RAM) or read-only memory (ROM). Optionally, the memory 220 may include a non-transitory computer-readable storage medium. The memory 220 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 220 may include a program storage area. This program storage area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described above, etc.

[0030] Communication interface 230 is used to communicate with other devices, equipment or communication networks, such as data storage devices, image processing devices or Ethernet, wireless access network (RAN), wireless local area network (WLAN), etc.

[0031] In terms of physical implementation, the aforementioned devices (such as processor 210, memory 220, and communication interface 230) can each be devices within the same device (such as a laptop computer). Alternatively, at least two of these devices can be located within the same device, i.e., as different devices within the same device, similar to the deployment of devices or components in a distributed system.

[0032] It is understood that the structure illustrated in this embodiment does not constitute a specific limitation on the electronic device 20. In other embodiments of the present invention, the electronic device 20 may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0033] The following description, in conjunction with the accompanying drawings, illustrates an intelligent recommendation method for equipment fault repair provided by an embodiment of the present invention.

[0034] like Figure 2 As shown, this embodiment of the invention provides an intelligent recommendation method for equipment fault repair, which may include, but is not limited to: S1: Construct an initial equipment maintenance measure recommendation model.

[0035] In one feasible implementation, the construction of the initial equipment maintenance measure recommendation model described in this embodiment of the invention may include, but is not limited to: An initial equipment maintenance measure recommendation model based on the Transformer architecture is constructed. The initial equipment maintenance measure recommendation model includes an input layer, an encoding layer, a feature fusion layer, and an output layer. The encoding layer includes a text encoding channel and a structured data encoding channel. The text encoding channel is used to extract text features of the target device fault information, and the structured data encoding channel is used to extract structured features of the target device fault information. The feature fusion layer is used to fuse the text features and the structured features to generate comprehensive features of equipment faults; The output layer is used to output recommended maintenance measures based on the comprehensive characteristics of the equipment fault.

[0036] The initial equipment maintenance measure recommendation model in this embodiment of the invention uses a domain-adaptive pre-trained language model as its backbone, and extends the structured data interface and dual-task output head.

[0037] In specific implementation, the text encoding channel in this embodiment of the invention uses a domain-adjusted BERT model (such as BERT-Base) as the encoder. After word segmentation and the addition of special markers (such as [CLS], [SEP], etc.), the input text is fed into the BERT Transformer layer. The Transformer layer extracts the final hidden state of the markers (e.g., a 768-dimensional vector) as the semantic summary representation of the entire text sequence, i.e., the text features, denoted as... In this embodiment of the invention, the structured data encoding channel maps discrete structured features such as "equipment type," "professional field," and "defect classification (label)" into dense vectors using an embedding layer. For continuous features such as "equipment service life," the data is standardized and either directly used or encoded through a small fully connected network. Then, the embedding vectors of all such structured features are concatenated into a comprehensive structured feature vector, i.e., the structured feature, denoted as […]. .

[0038] In one feasible implementation, the feature fusion layer in this embodiment of the invention may be an attention-based feature fusion layer.

[0039] In the specific implementation process, the feature fusion layer uses As a query, Using structured features as both Key and Value, we compute importance weights of the structured features relative to the semantics of the text, and generate a context-aware fusion representation. ,Then and By splicing or weighted summation, the comprehensive fault features of the equipment are generated. .

[0040] In one feasible implementation, the output layer in this embodiment of the invention includes a fault severity classification head and a maintenance measure matching score head; The fault severity classification head outputs the defect severity level through a first fully connected layer and a first Softmax function; The maintenance measure matching score head outputs the maintenance measure matching score through the second fully connected layer and the second Softmax function.

[0041] In the specific implementation process, the initial equipment maintenance measure recommendation model in this embodiment of the invention simultaneously performs two key tasks: the first task is to classify the severity of the fault. The first fully connected layer and the first Softmax function output a probability distribution corresponding to three levels of defects: "minor," "moderate," and "severe." This task aims to prioritize faults. Task two is to score the matching degree of maintenance measures. The second fully connected layer and the second Softmax function output a scalar score between 0 and 1. This scalar score represents the degree of match between the current input defect and the "best repair action" in the model's implicit knowledge.

[0042] S2: Collect historical fault data of the target equipment, and use the historical fault data to train the initial equipment maintenance measure recommendation model to obtain the target equipment maintenance measure recommendation model.

[0043] In specific implementation processes, embodiments of the present invention may, but are not limited to, store equipment operation and fault-related data through an equipment management system, wherein the equipment operation and fault-related data may, but are not limited to, include: Basic equipment information, including equipment type, equipment model, installation date, and operating status; Information on the professional field to which the equipment belongs; Fault phenomenon information is used to characterize abnormal operating conditions of equipment; Fault classification information is used to indicate the severity of the fault; Fault description information, wherein the fault description is natural language text; Historical processing strategy information, including measures such as repair, adjustment or replacement.

[0044] In specific implementation, embodiments of the present invention can connect to the equipment management system via an interface to collect historical fault data of the target equipment. During the data collection process, the data can be associated and stored according to the equipment dimension and the fault event dimension, and the data can be synchronized with the equipment management system periodically or in real time to ensure the timeliness and integrity of the data.

[0045] In one feasible implementation, before training the initial equipment maintenance measure recommendation model using the historical fault data, the embodiments of the present invention may include, but are not limited to: The historical fault data is preprocessed to generate a standardized dataset for training the initial equipment maintenance measure recommendation model; The preprocessing includes: The duplicate data in the historical fault data is deduplicated; Complete or correct data with missing key fields; Correct abnormal or non-compliant data; Standardize the fault description text; Convert fault classification information into numerical labels; and, Enhancement processing is applied to minority class samples.

[0046] In this embodiment of the invention, after obtaining a standardized dataset that can be used for training the initial equipment maintenance measure recommendation model through the above preprocessing, the standardized dataset can be divided into a training sample set, a test sample set, and a validation sample set according to a certain ratio. Then, these samples are sequentially input into the initial equipment maintenance measure recommendation model for training, testing, and validation, thereby obtaining the target equipment maintenance measure recommendation model.

[0047] In one feasible implementation, the training of the initial equipment maintenance measure recommendation model using the historical fault data in this embodiment of the invention may include, but is not limited to: The following composite loss function is used to guide the training of the initial equipment maintenance measure recommendation model: ; ; ; in, For classifying losses, The weights for each category c, One-hot encoding of the real label. To predict probabilities for the model, For the matching degree loss, For matching degree, For the true value of the matching degree, θ is the composite loss function; α and β are hyperparameters, λ is the regularization coefficient, and θ is all weight parameters of the model.

[0048] In the specific implementation process, due to the imbalance in the severity samples of various faults in the historical fault data, the classification loss in this embodiment of the invention... We employ weighted cross-entropy loss, assigning a weight to each fault category c. The weights are inversely proportional to the class frequency, making the model pay more attention to the minority class.

[0049] In the specific implementation process, the matching degree loss in the embodiments of the present invention Using mean squared error loss, the ground truth of the matching degree in the training sample set The evaluation is based on historical maintenance records. For example, if the equipment's trouble-free operating time exceeds a threshold after a maintenance, it is considered a successful match. Set to 1 (or a high value), otherwise set to 0 (or a low value).

[0050] In specific implementation, the composite loss function in the embodiments of the present invention Classification loss and matching degree loss The weighted sum is added, and an L2 regularization term is added to prevent overfitting.

[0051] In one feasible implementation, the method of training the initial equipment maintenance measure recommendation model using the historical fault data in this embodiment of the invention may, but is not limited to, include: The equipment diagnosis rules summarized by experts are transformed into supervised samples to assist in training the initial equipment maintenance measure recommendation model.

[0052] This invention transforms expert-summarized equipment diagnostic rules into supervised samples for auxiliary training of the model. This allows the model to learn from historical data patterns while inheriting expert diagnostic logic, thereby further improving recommendation accuracy.

[0053] S3: Obtain the target equipment fault information input by the user, input the target equipment fault information into the target equipment maintenance measure recommendation model, and output the target equipment maintenance measure recommendation scheme.

[0054] In one feasible implementation, the step of inputting the target equipment fault information into the target equipment maintenance measure recommendation model and outputting a target equipment maintenance measure recommendation scheme in this embodiment of the invention may include, but is not limited to: The fault information of the target equipment is input into multiple trained target equipment maintenance measure recommendation models. Each target equipment maintenance measure recommendation model outputs an initial target equipment maintenance measure recommendation scheme, which includes an initial target equipment fault category and an initial target equipment maintenance measure matching degree. Based on all the initial target equipment maintenance measure recommendations, generate the target equipment maintenance measure recommendation plan; Output a recommended maintenance plan for the target equipment.

[0055] This invention improves the stability and accuracy of the recommendation results by inputting the target equipment fault information into multiple trained target equipment maintenance measure recommendation models for processing, and then generating a target equipment maintenance measure recommendation scheme based on all the output initial target equipment maintenance measure recommendation schemes.

[0056] It should be noted that, in this embodiment of the invention, only one target equipment maintenance measure recommendation model may be trained. When there is only one target equipment maintenance measure recommendation model, the initial target equipment maintenance measure recommendation scheme output by the target equipment maintenance measure recommendation model is the target equipment maintenance measure recommendation scheme.

[0057] In its specific implementation, after acquiring the target device fault information input by the user, the input layer of this invention performs standardization processing on the target device fault information. The target device fault information includes at least the device type, a description of the fault phenomenon, and a fault text description. The standardization processing may include, but is not limited to, natural language preprocessing of the fault text description, such as replacing non-standard terms with preset standard terms, removing redundant words that do not contribute to defect identification, standardizing vocabulary, and extracting key entities and their relationship information from the defect text, thus giving it a unified and standardized text information expression form; and organizing structured information, such as organizing structured information like device type, device operating status, and historical fault frequency, etc., without limitation.

[0058] In one feasible implementation, the step of generating the target equipment maintenance measure recommendation scheme based on all the initial target equipment maintenance measure recommendation schemes in this embodiment of the invention may include, but is not limited to: Based on all the initial target equipment maintenance measure recommendations, a selected target equipment maintenance measure recommendation scheme is generated using a weighted average algorithm or a voting mechanism. The selected target equipment maintenance measure recommendation scheme includes the selected target equipment fault category and the selected target equipment maintenance measure matching degree. Determine whether the matching degree of the selected target equipment maintenance measures exceeds the preset maintenance measure matching degree threshold; If so, the recommended maintenance measures for the selected target equipment shall be used as the recommended maintenance measures for the target equipment. If not, then according to the fallback mechanism, the target equipment diagnosis rule corresponding to the target equipment fault information is matched from the preset expert rule base, and the target equipment maintenance case similar to the target equipment fault information is retrieved from the preset historical case base. Based on the target equipment diagnosis rule and the target equipment maintenance case, a recommended solution for the target equipment maintenance measures is generated.

[0059] In the specific implementation process, the expert rule base in this embodiment stores equipment diagnosis rules summarized by experts, and the historical case base stores equipment maintenance cases. During the model training phase, the equipment diagnosis rules in the expert rule base are used to assist training. During the recommendation phase, when the matching degree of the selected target equipment maintenance measures is lower than the preset maintenance measure matching degree threshold, the expert rule base is used to match the target equipment diagnosis rule corresponding to the target equipment defect, and the preset historical case base is used to retrieve target equipment maintenance cases similar to the target equipment fault information. Based on the target equipment diagnosis rules and the target equipment maintenance cases, a target equipment maintenance measure recommendation scheme is generated to replace the selected target equipment maintenance measure recommendation scheme.

[0060] The embodiments of the present invention employ a fallback mechanism to ensure that executable maintenance suggestions can still be output even when model inference is uncertain, thereby reducing the uncertainty risk caused by low matching degree and improving the reliability and applicability of the recommendation results.

[0061] In the specific implementation process, the embodiments of the present invention output the generated target equipment maintenance measure recommendation scheme to the equipment management system. The recommendation result may include, but is not limited to, maintenance measure description, recommendation confidence information and recommendation source identifier, etc., which are not limited here.

[0062] In one feasible implementation, after outputting the recommended maintenance measures for the target equipment in this embodiment of the invention, it may include, but is not limited to: Obtain user feedback data on the recommended maintenance measures for the target equipment, and dynamically update the recommended maintenance measures model for the target equipment based on the feedback data.

[0063] In specific implementation, the feedback data in this embodiment of the invention may include, but is not limited to, collecting user feedback on the recommendation results and actual repair results, recording recommendation adoption status, repair effects, and related operational indicators, etc., and is not limited here. This embodiment of the invention can continuously collect the feedback data and form an incremental dataset. When a preset time period is reached or when the performance of the target equipment repair measure recommendation model declines to a threshold, the model retraining or incremental fine-tuning process is automatically triggered; at the same time, combined with an active learning strategy, high-value samples are prioritized for model updates, thereby achieving continuous improvement in model performance.

[0064] The intelligent equipment fault repair recommendation method provided in this embodiment of the invention introduces natural language semantic understanding, multi-model integrated reasoning and fallback recommendation mechanism, which realizes intelligent, stable and engineering support for equipment defect repair decisions. It solves the problems of difficult structured equipment defect description and insufficient reliability of repair recommendations that rely on human experience in the prior art, and improves repair efficiency and decision accuracy.

[0065] Based on the intelligent recommendation method for equipment fault repair provided in the first aspect, embodiments of the present invention provide an intelligent recommendation system for equipment fault repair, such as... Figure 3 As shown, the intelligent equipment fault repair recommendation system includes: Model building module 110 is used to build an initial equipment maintenance measure recommendation model; The model training module 120 is used to collect historical fault data of the target equipment, and use the historical fault data to train the initial equipment maintenance measure recommendation model to obtain the target equipment maintenance measure recommendation model. The intelligent recommendation module 130 is used to obtain the target equipment fault information input by the user, input the target equipment fault information into the target equipment maintenance measure recommendation model, and output the target equipment maintenance measure recommendation scheme.

[0066] Based on the intelligent equipment fault repair recommendation method provided in the first aspect, this embodiment of the invention also provides a storage medium storing a computer-executable program. The computer-executable program is used to cause a computer to execute the intelligent equipment fault repair recommendation method as described in any implementation of the first aspect. Explanations of the relevant content and descriptions of the beneficial effects of any of the computer-readable storage media provided above can be found in the corresponding embodiments described above, and will not be repeated here.

[0067] Those skilled in the art will understand that the program for implementing all or part of the steps of the above embodiments, which can be executed by a program instructing related hardware, can be stored in a computer-readable storage medium. The storage medium mentioned above can be a read-only memory, a random access memory, etc. The processing unit or processor mentioned above can be a central processing unit, a general-purpose processor, an application-specific integrated circuit (ASIC), a microprocessor (DSP), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof.

[0068] This invention also provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform any of the methods described in the above embodiments. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., SSD), etc.

[0069] It should be noted that the devices for storing computer instructions or computer programs provided in the embodiments of the present invention, such as, but not limited to, the aforementioned memory, computer-readable storage medium, and communication chip, are all non-transitory. Those skilled in the art should recognize that the functions described in the embodiments of the present invention in one or more of the above examples can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable storage medium or transmitted as one or more instructions or code on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of computer programs from one place to another. Storage media can be any available medium accessible to general-purpose or special-purpose computers.

[0070] 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 method for intelligent recommendation of equipment fault repair, characterized in that, include: Construct an initial equipment maintenance recommendation model; Collect historical fault data of the target equipment, and use the historical fault data to train the initial equipment maintenance measure recommendation model to obtain the target equipment maintenance measure recommendation model; Obtain the target equipment fault information input by the user, input the target equipment fault information into the target equipment maintenance measure recommendation model, and output the target equipment maintenance measure recommendation scheme.

2. The intelligent recommendation method for equipment fault repair according to claim 1, characterized in that, The construction of the initial equipment maintenance measure recommendation model includes: An initial equipment maintenance measure recommendation model based on the Transformer architecture is constructed. The initial equipment maintenance measure recommendation model includes an input layer, an encoding layer, a feature fusion layer, and an output layer. The encoding layer includes a text encoding channel and a structured data encoding channel. The text encoding channel is used to extract text features of the target device fault information, and the structured data encoding channel is used to extract structured features of the target device fault information. The feature fusion layer is used to fuse the text features and the structured features to generate comprehensive features of equipment faults; The output layer is used to output recommended maintenance measures based on the comprehensive characteristics of the equipment fault.

3. The intelligent recommendation method for equipment fault repair according to claim 2, characterized in that, The feature fusion layer is an attention-based feature fusion layer.

4. The intelligent recommendation method for equipment fault repair according to claim 2, characterized in that, The output layer includes a fault severity classification header and a maintenance measure matching score header; The fault severity classification head outputs the defect severity level through a first fully connected layer and a first Softmax function; The maintenance measure matching score head outputs the maintenance measure matching score through the second fully connected layer and the second Softmax function.

5. The intelligent recommendation method for equipment fault repair according to claim 2, characterized in that, Before training the initial equipment maintenance measure recommendation model using the historical fault data, the method further includes: The historical fault data is preprocessed to generate a standardized dataset for training the initial equipment maintenance measure recommendation model; The preprocessing includes: The duplicate data in the historical fault data is deduplicated; Complete or correct data with missing key fields; Correct abnormal or non-compliant data; Standardize the fault description text; Convert fault classification information into numerical labels; and, Enhancement processing is applied to minority class samples.

6. The intelligent recommendation method for equipment fault repair according to claim 2, characterized in that, The step of training the initial equipment maintenance measure recommendation model using the historical fault data includes: The following composite loss function is used to guide the training of the initial equipment maintenance measure recommendation model: ； ； ； in, For classifying losses, The weights for each category c, One-hot encoding of the real label. To predict probabilities for the model, For the matching degree loss, For matching degree, For the true value of the matching degree, θ is the composite loss function; α and β are hyperparameters, λ is the regularization coefficient, and θ is all weight parameters of the model.

7. The intelligent recommendation method for equipment fault repair according to claim 6, characterized in that, The step of training the initial equipment maintenance measure recommendation model using the historical fault data further includes: The equipment diagnosis rules summarized by experts are transformed into supervised samples to assist in training the initial equipment maintenance measure recommendation model.

8. The intelligent recommendation method for equipment fault repair according to claim 2, characterized in that, The step of inputting the target equipment fault information into the target equipment maintenance measure recommendation model and outputting a target equipment maintenance measure recommendation scheme includes: The fault information of the target equipment is input into multiple trained target equipment maintenance measure recommendation models. Each target equipment maintenance measure recommendation model outputs an initial target equipment maintenance measure recommendation scheme, which includes an initial target equipment fault category and an initial target equipment maintenance measure matching degree. Based on all the initial target equipment maintenance measure recommendations, generate the target equipment maintenance measure recommendation plan; Output a recommended maintenance plan for the target equipment.

9. The intelligent recommendation method for equipment fault repair according to claim 8, characterized in that, The step of generating the target equipment maintenance measure recommendation scheme based on all the initial target equipment maintenance measure recommendation schemes includes: Based on all the initial target equipment maintenance measure recommendations, a selected target equipment maintenance measure recommendation scheme is generated using a weighted average algorithm or a voting mechanism. The selected target equipment maintenance measure recommendation scheme includes the selected target equipment fault category and the selected target equipment maintenance measure matching degree. Determine whether the matching degree of the selected target equipment maintenance measures exceeds the preset maintenance measure matching degree threshold; If so, the recommended maintenance measures for the selected target equipment shall be used as the recommended maintenance measures for the target equipment. If not, then according to the fallback mechanism, the target equipment diagnosis rule corresponding to the target equipment fault information is matched from the preset expert rule base, and the target equipment maintenance case similar to the target equipment fault information is retrieved from the preset historical case base. Based on the target equipment diagnosis rule and the target equipment maintenance case, a recommended solution for the target equipment maintenance measures is generated.

10. The intelligent recommendation method for equipment fault repair according to claim 9, characterized in that, After outputting the recommended maintenance measures for the target equipment, the method further includes: Obtain user feedback data on the recommended maintenance measures for the target equipment, and dynamically update the recommended maintenance measures model for the target equipment based on the feedback data.

Images