Operation and maintenance decision optimization method and device, nonvolatile storage medium and electronic equipment

Abstract

The application discloses an operation and maintenance decision optimization method and device, a nonvolatile storage medium and an electronic device. The method comprises the following steps: obtaining a to-be-detected image shot by an operation and maintenance personnel, wherein the to-be-detected image comprises a target device, the target device is a device that needs to be inspected in an operation and maintenance decision, the operation and maintenance decision is generated by an operation and maintenance decision generation model, and the operation and maintenance decision records the target device and a first device fault corresponding to the target device; detecting the to-be-detected image to determine a second device fault existing in the target device; determining whether the first device fault and the second device fault are matched to obtain a fault matching result; and reporting decision failure information in the case that the determination result is not matched, wherein the decision failure information comprises the to-be-detected image, the operation and maintenance decision and the second device fault. The application solves the technical problems of low feedback efficiency and slow operation and maintenance decision updating and optimization caused by manual feedback in the existing operation and maintenance decision optimization process.

Description

Technical Field

[0001] This application relates to the field of operation and maintenance decision-making, and more specifically, to an operation and maintenance decision optimization method, apparatus, non-volatile storage medium, and electronic device. Background Technology

[0002] In the operation and maintenance (O&M) management of power distribution networks, traditional O&M decision-making processes rely on manual feedback to assess equipment status and the effectiveness of decisions. This method has significant limitations in terms of efficiency and accuracy. The inefficiency and inaccuracy of manual feedback make it difficult to identify and correct O&M decision-making errors in a timely manner, resulting in a slow O&M decision optimization process and hindering the formation of an effective closed loop. This severely impacts the quality of the O&M decision-making model and the overall performance of the O&M system.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This application provides an operation and maintenance decision optimization method, apparatus, non-volatile storage medium, and electronic device to at least solve the technical problems of low feedback efficiency and slow operation and maintenance decision update and optimization caused by the use of manual feedback in the existing operation and maintenance decision optimization process.

[0005] According to one aspect of the embodiments of this application, an operation and maintenance decision optimization method is provided, comprising: acquiring an image to be inspected taken by operation and maintenance personnel, wherein the image to be inspected includes a target device, the target device being a device that needs to be inspected as recorded in the operation and maintenance decision, the operation and maintenance decision being generated by an operation and maintenance decision generation model, the operation and maintenance decision recording including the target device and a first device fault corresponding to the target device, the first device fault being predicted by the operation and maintenance decision generation model; inspecting the image to be inspected to determine a second device fault existing in the target device; determining whether the first device fault and the second device fault match, and obtaining a fault matching result, wherein the fault matching result includes fault matching and fault mismatch; if the determination result is a mismatch, reporting decision error information, wherein the decision error information includes the image to be inspected, the operation and maintenance decision, and the second device fault, and the decision error information is used to optimize the operation and maintenance decision generation model.

[0006] Optionally, after reporting decision error information, the method further includes: adding the decision error information to the training dataset; and training the operation and maintenance decision generation model based on the updated training dataset.

[0007] Optionally, training the operation and maintenance decision generation model based on the updated training dataset includes: determining the decision error rate, wherein the decision error rate is the proportion of fault-mismatched operation and maintenance decisions to all operation and maintenance decisions; determining the training parameters of the operation and maintenance decision generation model based on the decision error rate, wherein the training parameters include the learning rate, and the larger the decision error rate, the larger the learning rate; and training the operation and maintenance decision generation model based on the updated training parameters.

[0008] Optionally, detecting the image to be detected and determining the second device fault in the image to be detected includes: extracting features from the image to be detected to obtain image features; determining the fault identification type of the target device in the image to be detected and the fault confidence level corresponding to the fault identification type based on the image features; and determining the fault identification type corresponding to the fault confidence level as the second device fault if the fault confidence level is greater than a preset confidence threshold.

[0009] Optionally, before acquiring the image to be inspected taken by the maintenance personnel, the method further includes: inputting historical equipment data into the maintenance decision generation model, wherein the historical equipment data includes at least one of the following: historical equipment maintenance records, historical equipment fault data; acquiring the target equipment output by the maintenance decision generation model, and the first equipment fault corresponding to the target equipment; generating a maintenance work order based on the target equipment and the first equipment fault, wherein the maintenance work order is used to instruct the maintenance personnel to perform inspections on the target equipment; and sending the maintenance work order to the maintenance personnel.

[0010] Optionally, before reporting decision error information, the method further includes: obtaining feedback information from maintenance personnel, wherein the feedback information includes text information and voice information; extracting semantic information from the feedback information, wherein the semantic information includes at least one of the following: evaluation information on maintenance decisions, evaluation information on second equipment failures, and evaluation information on failure matching results; and adding the semantic information to the decision error information.

[0011] Optionally, after determining that the target device has a second device fault, the method further includes: recording the second device fault in the operation and maintenance record, wherein the operation and maintenance record is used to generate operation and maintenance decisions.

[0012] According to another aspect of the embodiments of this application, an operation and maintenance decision optimization device is also provided, comprising: an acquisition module, configured to acquire an image to be inspected taken by operation and maintenance personnel, wherein the image to be inspected includes a target device, the target device being a device that needs to be inspected as recorded in the operation and maintenance decision, the operation and maintenance decision being generated by an operation and maintenance decision generation model, the operation and maintenance decision recording including the target device and a first device fault corresponding to the target device, the first device fault being predicted by the operation and maintenance decision generation model; a detection module, configured to detect the image to be inspected and determine a second device fault existing in the target device; a judgment module, configured to judge whether the first device fault and the second device fault match, and obtain a fault matching result, wherein the fault matching result includes fault matching and fault mismatch; and a reporting module, configured to report decision error information when the judgment result is mismatch, wherein the decision error information includes the image to be inspected, the operation and maintenance decision, and the second device fault, and the decision error information is used to optimize the operation and maintenance decision generation model.

[0013] According to another aspect of the embodiments of this application, a non-volatile storage medium is also provided, wherein a program is stored in the non-volatile storage medium, and the program controls the device where the non-volatile storage medium is located to execute an operation and maintenance decision optimization method when it runs.

[0014] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory and a processor, wherein the processor is configured to run a program stored in the memory, wherein the program executes an operation and maintenance decision optimization method during runtime.

[0015] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements an operation and maintenance decision optimization method.

[0016] In this embodiment, an image to be inspected is acquired by maintenance personnel. The image includes a target device, which is a device recorded in the maintenance decision that needs to be inspected. The maintenance decision is generated by an maintenance decision generation model. The maintenance decision record includes the target device and a first device fault corresponding to the target device. The first device fault is predicted by the maintenance decision generation model. The image to be inspected is then inspected to determine a second device fault existing in the target device. It is then determined whether the first device fault and the second device fault match, and a fault matching result is obtained. The fault matching result includes fault matching and fault mismatch. If the result is a mismatch, decision error information is reported. The decision error information includes the image to be inspected, the maintenance decision, and the second device fault. The decision error information is used to optimize the maintenance decision generation model. By automatically detecting device faults through image recognition and deep learning technologies and comparing the prediction results, the purpose of real-time identification and reporting of decision errors is achieved. This achieves the technical effect of rapidly optimizing the maintenance decision generation model, thereby solving the technical problems of low feedback efficiency and slow maintenance decision update optimization caused by manual feedback in the existing maintenance decision optimization process. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0018] Figure 1 This is a schematic diagram of the structure of a computer terminal according to an embodiment of this application;

[0019] Figure 2 This is a flowchart illustrating an operation and maintenance decision optimization method provided according to an embodiment of this application;

[0020] Figure 3 This is a schematic diagram of an operation and maintenance decision optimization device provided according to an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] To better understand the embodiments of this application, the technical terms involved in the embodiments of this application are explained below:

[0024] In related technologies, with the surge in the number of power distribution network devices and the increasing complexity of operation and maintenance, traditional manual feedback methods can no longer meet the demands of modern intelligent operation and maintenance systems for real-time performance, accuracy, and intelligent optimization. How to quickly and accurately feed on-site inspection results back to the decision-making system to form an immediate decision-making optimization loop has become a pressing technical challenge.

[0025] To address this issue, relevant solutions are provided in the embodiments of this application, which are described in detail below.

[0026] According to an embodiment of this application, a method embodiment for optimizing operation and maintenance decisions is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0027] The methods and embodiments provided in this application can be executed on mobile terminals, computer terminals, or similar computing devices. Figure 1 A hardware block diagram of a computer terminal for implementing an operation and maintenance decision optimization method is shown. Figure 1As shown, the computer terminal 10 may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0028] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10. As involved in the embodiments of this application, the data processing circuits serve as processor control (e.g., selection of a variable resistor termination path connected to an interface).

[0029] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the operation and maintenance decision optimization method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the above-mentioned operation and maintenance decision optimization method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0030] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0031] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10.

[0032] Under the above operating environment, this application provides an operation and maintenance decision optimization method, such as... Figure 2 As shown, the method includes the following steps:

[0033] Step S202: Obtain the image to be inspected taken by the maintenance personnel. The image to be inspected includes the target device, which is the device that needs to be inspected as recorded in the maintenance decision. The maintenance decision is generated by the maintenance decision generation model. The maintenance decision record includes the target device and the first device fault corresponding to the target device. The first device fault is predicted by the maintenance decision generation model.

[0034] As an optional implementation, before acquiring the image to be inspected taken by the maintenance personnel, the method further includes: inputting historical equipment data into the maintenance decision generation model, wherein the historical equipment data includes at least one of the following: historical equipment maintenance records, historical equipment fault data; acquiring the target equipment output by the maintenance decision generation model, and the first equipment fault corresponding to the target equipment; generating a maintenance work order based on the target equipment and the first equipment fault, wherein the maintenance work order is used to instruct the maintenance personnel to inspect the target equipment; and sending the maintenance work order to the maintenance personnel.

[0035] Optionally, the operation and maintenance decision generation model is trained based on historical data. It uses collected historical equipment data for fault prediction and decision generation, including equipment operation and maintenance records, fault history, and other equipment-related business activity data. This data undergoes preprocessing, such as denoising, feature extraction, and data standardization, to eliminate redundant information and improve model training efficiency. The operation and maintenance decision generation model is trained using deep learning architectures such as Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN) to capture the dynamic characteristics of equipment status changes over time and potential fault modes. After training, the operation and maintenance decision generation model can accept historical equipment data as input and output predicted fault probabilities and fault types. If the fault probability exceeds a preset probability threshold, the corresponding equipment is designated as the target equipment, the corresponding fault type as the first equipment fault, and an operation and maintenance work order is generated, instructing operation and maintenance personnel to perform operation and maintenance on the target equipment with the first equipment fault.

[0036] Optionally, maintenance work orders are automatically generated based on the first device fault prediction result output by the model. The work orders detail the target equipment information requiring inspection, the predicted fault type, the recommended inspection time, the inspection route, and the tools and spare parts that may be needed. These work orders not only guide the work of maintenance personnel but also provide a preliminary assessment of fault prediction, reducing blind inspections and improving maintenance efficiency.

[0037] Finally, the system sends the generated maintenance work order to the relevant maintenance personnel via mobile devices, email, or system notifications to ensure timely receipt and execution. Upon receiving the work order, the maintenance personnel will proceed to the site to inspect the target equipment, capture images of the equipment to be inspected, and verify the model's predictive accuracy through subsequent image recognition processes. This forms a decision feedback loop, continuously optimizing the decision model and ensuring the timeliness and effectiveness of maintenance decisions.

[0038] Step S204: Detect the image to be detected and determine the second device fault present in the target device.

[0039] As an optional implementation, detecting the image to be detected and determining the second device fault in the image includes: extracting features from the image to be detected to obtain image features; determining the fault identification type of the target device in the image to be detected and the fault confidence level corresponding to the fault identification type based on the image features; and determining the fault identification type corresponding to the fault confidence level as the second device fault when the fault confidence level is greater than a preset confidence threshold.

[0040] Optionally, a deep feature extraction is performed on the image to be detected using an image recognition model. This process involves using a pre-trained convolutional neural network (CNN) model to automatically identify and extract key information in the image, such as the structural features of the device, surface defects, and operating status, through multi-layer convolution and pooling operations, and converting it into a digital image feature vector.

[0041] The extracted image feature vectors are then input into a fault identification model, which is also a deep learning model, used to identify image features corresponding to different fault types. The model calculates a series of possible fault types and their corresponding fault confidence scores based on the input feature vectors; that is, the probability that the model determines a specific fault type exists. Fault confidence scores reflect the reliability and accuracy of the identification results and are an important indicator for evaluating the model's fault identification performance.

[0042] In the fault identification results, the system filters out fault identification types with high confidence based on a preset confidence threshold. The preset confidence threshold is a key parameter used to distinguish between genuine and false identification results, preventing low-confidence erroneous identifications from being mistaken for real equipment faults. Only when the confidence level corresponding to a fault identification type is higher than the preset confidence threshold is the identification result considered a valid secondary equipment fault, subsequently recorded in the operation and maintenance decision system's database, and used to update the target equipment's operation and maintenance records.

[0043] For example, if the model identifies a potential "insulation aging" fault in a device within an image to be inspected, with a confidence level of 85% and a preset confidence threshold of 70%, then "insulation aging" will be identified as a valid second device fault. The system will automatically record this fault information and associate it with the target device's ID. Simultaneously, it will update the device's maintenance records with detailed information such as the fault type, fault confidence level, and the date and time of fault identification. This process not only ensures the accurate recording of fault information but also provides valuable data for subsequent equipment status assessments, maintenance plan development, and optimization of decision-making models.

[0044] As an optional implementation, after determining the existence of a second equipment fault in the target equipment, the method further includes: recording the second equipment fault in the operation and maintenance record, wherein the operation and maintenance record is used to generate operation and maintenance decisions.

[0045] Optionally, after identifying a secondary equipment fault in the target device using image recognition technology, it is entered into the maintenance record system. This operation aims to ensure the completeness and timeliness of all fault information, providing accurate data support for subsequent maintenance decisions. Maintenance records are not limited to faults discovered during the current inspection, but also include the equipment's maintenance history, fault frequency, repair records, and information on replaced parts. Each new fault entry automatically updates the equipment's historical data, forming a continuously enriched and detailed equipment profile, providing comprehensive historical references for future decisions based on equipment status. Furthermore, the maintenance record system supports multi-dimensional data querying and analysis. Maintenance managers can quickly query historical fault information and analyze fault trends through the system, providing data support for developing long-term equipment maintenance strategies and budget planning. The system may also generate maintenance reports based on maintenance records, automatically calculating fault types, frequencies, and maintenance costs, helping the maintenance team identify problems in maintenance and further optimize maintenance processes and resource allocation.

[0046] Step S206: Determine whether the faults of the first device and the second device match, and obtain the fault matching result, wherein the fault matching result includes fault matching and fault mismatch.

[0047] Optionally, if the fault types of the first equipment fault and the second equipment fault are the same, the fault matching result is determined to be fault matching; if the fault types of the first equipment fault and the second equipment fault are different, the fault matching result is determined to be fault mismatch.

[0048] Step S208: If the judgment result is a mismatch, report the decision error information. The decision error information includes the image to be detected, the operation and maintenance decision and the second device failure. The decision error information is used to optimize the operation and maintenance decision generation model.

[0049] As an optional implementation, after reporting decision error information, the method further includes: adding the decision error information to the training dataset; and training the operation and maintenance decision generation model based on the updated training dataset.

[0050] Optionally, training the operation and maintenance decision generation model based on the updated training dataset includes: determining the decision error rate, wherein the decision error rate is the proportion of fault-mismatched operation and maintenance decisions to all operation and maintenance decisions; determining the training parameters of the operation and maintenance decision generation model based on the decision error rate, wherein the training parameters include the learning rate, and the larger the decision error rate, the larger the learning rate; and training the operation and maintenance decision generation model based on the updated training parameters.

[0051] Optionally, decision-making error information can be used to improve the prediction accuracy and decision-making performance of the operation and maintenance decision generation model. Specifically, the system first adds confirmed decision-making error information, i.e., operation and maintenance decision cases of "fault mismatch," to the model's training dataset, expanding the coverage of the training data and ensuring that the model can learn from and adjust its decision logic from more real-world cases. This dataset update process is continuous; each new decision-making error information enriches the dataset, allowing the model to encounter more types of fault information and specific equipment conditions in subsequent training, thereby enhancing its ability to identify and predict complex fault scenarios.

[0052] Subsequently, the decision error rate in the updated training dataset is calculated, which is the proportion of fault-mismatched operational decisions out of all operational decisions. This proportion reflects the model's prediction accuracy in practical applications and is a key indicator for evaluating model performance. A high decision error rate means the model's predictive ability needs improvement, while a low decision error rate indicates good prediction accuracy. The system automatically adjusts the training parameters of the operational decision generation model based on the decision error rate, with the learning rate being a crucial adjustment factor. The learning rate determines how quickly the model updates its parameters during training. A larger learning rate means the model adjusts parameters more significantly in each iteration, enabling faster learning and adaptation, but may lead to model instability during training; while a smaller learning rate helps the model converge more stably, but the learning speed is relatively slower.

[0053] After determining the decision error rate, the system adjusts the learning rate accordingly. Generally, the higher the decision error rate, the larger the learning rate is set, thereby accelerating the model's learning process and enabling it to quickly adapt to new failure modes and operational scenarios. However, the adjustment of the learning rate must also consider the stability of model training to avoid excessively large learning rates that could lead to model divergence. The system automatically finds a balance point through algorithms that both accelerates the learning speed and ensures the stability of the training process and the continuous improvement of model performance.

[0054] Optionally, a baseline for the frequency of decision errors is set. When the actual frequency is higher than the baseline, the learning rate is increased in a direct proportional relationship (e.g., the learning rate increases by 5% for every 10% increase in frequency), and when the frequency is lower than the baseline, the learning rate is decreased in an inverse proportional relationship (e.g., the learning rate decreases by 5% for every 10% decrease in frequency).

[0055] Next, the system retrains the operation and maintenance decision generation model using the updated training dataset and adjusted learning rate. After each training iteration, the model's prediction accuracy improves accordingly, and the decision error rate gradually decreases, forming a continuously optimizing decision model training loop.

[0056] For example, suppose that in the initial stage, the decision error rate is high, reaching 20%. The system might adjust the learning rate upward to accelerate the model's learning of new data and expect to significantly improve prediction accuracy in subsequent training. As time goes on and the model's predictive ability improves, the decision error rate gradually decreases, and the learning rate may be gradually reduced to ensure that the model converges more stably as it approaches its optimal state, ultimately achieving comprehensive optimization of model performance.

[0057] By feeding back decision-making error information to the model training process, the system can not only continuously enrich the training dataset to ensure that the model can learn from actual operation and maintenance scenarios, but also dynamically adjust key training parameters such as the learning rate to accelerate the model's absorption of new knowledge and optimization of decision-making logic. Ultimately, this improves the prediction accuracy and decision-making effect of the operation and maintenance decision generation model, and realizes the intelligent and efficient operation and maintenance decision-making of the power distribution network.

[0058] Optionally, before reporting decision error information, the method further includes: obtaining feedback information from maintenance personnel, wherein the feedback information includes text information and voice information; extracting semantic information from the feedback information, wherein the semantic information includes at least one of the following: evaluation information on maintenance decisions, evaluation information on second equipment failures, and evaluation information on failure matching results; and adding the semantic information to the decision error information.

[0059] Optionally, in the intelligent operation and maintenance decision-making process, before reporting decision error information, the system proactively collects on-site feedback from operation and maintenance personnel. This feedback, consisting of text and voice information, comprehensively reflects the personnel's direct observation of the decision and equipment status. By employing natural language processing technology, the system can automatically extract semantic information from the feedback, including satisfaction with the operation and maintenance decision, evaluation of the accuracy of the second equipment fault identification, and confirmation of the fault matching results. This semantic information is considered an important component of decision error information, providing a deeper understanding of the fault identification context and operation and maintenance decision-making situation, helping the system to deeply understand the root causes of decision errors.

[0060] For example, after completing an equipment inspection, maintenance personnel can record the difference between the actual equipment status and the predictions of the decision-making model via voice, and describe in detail the specific manifestations of equipment failures and the reasons for the discrepancies with predictions in text. The system converts this feedback into semantic information, associates it with specific cases of decision-making errors, and forms decision-making error records containing contextual details, which are used for subsequent model training and optimization. This decision-making error information, containing the intuitive evaluations of maintenance personnel, can help the model more accurately identify equipment failure modes, optimize decision-making logic, and thus improve the accuracy of fault prediction and the rationality of maintenance strategies in future maintenance decisions.

[0061] This feedback mechanism ensures that the operation and maintenance decision generation model can continuously evolve, better meet actual operation and maintenance needs, and improve the intelligence level and efficiency of distribution network equipment operation and maintenance.

[0062] Through the above steps, automatic detection of equipment faults can be achieved and compared with the fault predictions of operation and maintenance decisions, thus achieving the purpose of timely identification and reporting of decision-making errors. This enables the technical effect of rapidly optimizing the operation and maintenance decision generation model, thereby solving the technical problems of low feedback efficiency and slow update and optimization of operation and maintenance decisions caused by the use of manual feedback in the existing operation and maintenance decision optimization process.

[0063] This application provides an operation and maintenance decision optimization device. Figure 3 This is a schematic diagram of the device, as shown below. Figure 3 As shown, the device includes: an acquisition module 30, used to acquire an image to be inspected taken by maintenance personnel, wherein the image to be inspected includes a target device, which is a device that needs to be inspected as recorded in the maintenance decision. The maintenance decision is generated by the maintenance decision generation model, and the maintenance decision record includes the target device and a first device fault corresponding to the target device. The first device fault is predicted by the maintenance decision generation model; a detection module 32, used to detect the image to be inspected and determine the second device fault existing in the target device; a judgment module 34, used to judge whether the first device fault and the second device fault match and obtain a fault matching result, wherein the fault matching result includes fault matching and fault mismatch; and a reporting module 36, used to report decision error information when the judgment result is mismatch, wherein the decision error information includes the image to be inspected, the maintenance decision, and the second device fault, and the decision error information is used to optimize the maintenance decision generation model.

[0064] In some embodiments of this application, after the reporting module 36 reports the decision error information, the method further includes: adding the decision error information to the training dataset; and training the operation and maintenance decision generation model based on the updated training dataset.

[0065] In some embodiments of this application, training an operation and maintenance decision generation model based on an updated training dataset includes: determining a decision error rate, wherein the decision error rate is the proportion of fault-mismatched operation and maintenance decisions to all operation and maintenance decisions; determining training parameters of the operation and maintenance decision generation model based on the decision error rate, wherein the training parameters include a learning rate, and the higher the decision error rate, the higher the learning rate; and training the operation and maintenance decision generation model based on the updated training parameters.

[0066] In some embodiments of this application, the detection module 32 detects the image to be detected and determines the second device fault in the image to be detected by: extracting features from the image to be detected to obtain image features; determining the fault identification type of the target device in the image to be detected and the fault confidence level corresponding to the fault identification type based on the image features; and determining the fault identification type corresponding to the fault confidence level as the second device fault when the fault confidence level is greater than a preset confidence threshold.

[0067] In some embodiments of this application, before the acquisition module 30 acquires the image to be detected taken by the maintenance personnel, the method further includes: inputting historical equipment data into the maintenance decision generation model, wherein the historical equipment data includes at least one of the following: historical equipment maintenance records, historical equipment fault data; acquiring the target device output by the maintenance decision generation model, and the first equipment fault corresponding to the target device; generating a maintenance work order based on the target device and the first equipment fault, wherein the maintenance work order is used to instruct the maintenance personnel to inspect the target device; and sending the maintenance work order to the maintenance personnel.

[0068] In some embodiments of this application, before the reporting module 36 reports the decision error information, the method further includes: obtaining feedback information from maintenance personnel, wherein the feedback information includes text information and voice information; extracting semantic information from the feedback information, wherein the semantic information includes at least one of the following: evaluation information on maintenance decisions, evaluation information on second equipment failures, and evaluation information on failure matching results; and adding the semantic information to the decision error information.

[0069] In some embodiments of this application, after the detection module 32 determines that a second device fault exists in the target device, it further includes: recording the second device fault into the operation and maintenance record, wherein the operation and maintenance record is used to generate operation and maintenance decisions.

[0070] It should be noted that each module in the above-mentioned operation and maintenance decision optimization device can be a program module (for example, a set of program instructions to implement a certain function) or a hardware module. For the latter, it can be manifested in the following forms, but is not limited to them: each of the above modules is manifested as a processor, or the functions of each of the above modules are implemented by a processor.

[0071] This application provides a non-volatile storage medium storing a program. During program execution, the program controls the device containing the non-volatile storage medium to perform the following operation and maintenance decision optimization method: acquiring an image to be inspected taken by maintenance personnel, wherein the image includes a target device, which is a device recorded in the operation and maintenance decision requiring inspection. The operation and maintenance decision is generated by an operation and maintenance decision generation model, and the operation and maintenance decision record includes the target device and a first device fault corresponding to the target device, which is predicted by the operation and maintenance decision generation model; inspecting the image to be inspected to determine a second device fault existing in the target device; determining whether the first device fault and the second device fault match, and obtaining a fault matching result, wherein the fault matching result includes fault matching and fault mismatch; if the determination result is a mismatch, reporting decision error information, wherein the decision error information includes the image to be inspected, the operation and maintenance decision, and the second device fault, and the decision error information is used to optimize the operation and maintenance decision generation model.

[0072] This application provides an electronic device, including a memory and a processor. The processor runs a program stored in the memory, wherein the program executes the following operation and maintenance decision optimization method: acquiring an image to be inspected taken by operation and maintenance personnel, wherein the image to be inspected includes a target device, the target device being a device recorded in the operation and maintenance decision that needs to be inspected, the operation and maintenance decision being generated by an operation and maintenance decision generation model, the operation and maintenance decision recording the target device and a first device fault corresponding to the target device, the first device fault being predicted by the operation and maintenance decision generation model; inspecting the image to be inspected to determine a second device fault existing in the target device; determining whether the first device fault and the second device fault match, and obtaining a fault matching result, wherein the fault matching result includes fault matching and fault mismatch; if the determination result is a mismatch, reporting decision error information, wherein the decision error information includes the image to be inspected, the operation and maintenance decision, and the second device fault, and the decision error information is used to optimize the operation and maintenance decision generation model.

[0073] This application provides a computer program product, including a computer program that, when executed by a processor, implements the following operation and maintenance decision optimization method: acquiring an image to be inspected taken by operation and maintenance personnel, wherein the image to be inspected includes a target device, the target device being a device recorded in the operation and maintenance decision that needs to be inspected, the operation and maintenance decision being generated by an operation and maintenance decision generation model, the operation and maintenance decision recording including the target device and a first device fault corresponding to the target device, the first device fault being predicted by the operation and maintenance decision generation model; inspecting the image to be inspected to determine a second device fault existing in the target device; determining whether the first device fault and the second device fault match, and obtaining a fault matching result, wherein the fault matching result includes fault matching and fault mismatch; if the determination result is a mismatch, reporting decision error information, wherein the decision error information includes the image to be inspected, the operation and maintenance decision, and the second device fault, and the decision error information is used to optimize the operation and maintenance decision generation model.

[0074] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0075] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0076] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0077] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0078] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to related technologies, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0079] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for optimizing operation and maintenance decisions, characterized in that, include: Acquire images to be inspected taken by maintenance personnel, wherein the images to be inspected include target devices, the target devices are devices that need to be inspected as recorded in the maintenance decision, the maintenance decision is generated by the maintenance decision generation model, the maintenance decision records the target devices and a first device fault corresponding to the target devices, and the first device fault is predicted by the maintenance decision generation model; The image to be detected is inspected to determine a second equipment fault in the target device; Determine whether the fault of the first device matches the fault of the second device, and obtain the fault matching result, wherein the fault matching result includes fault matching and fault mismatch; If the judgment result is a mismatch, a decision error information is reported. The decision error information includes the image to be detected, the operation and maintenance decision and the second equipment failure. The decision error information is used to optimize the operation and maintenance decision generation model.

2. The operation and maintenance decision optimization method according to claim 1, characterized in that, After reporting the decision-making error information, the method further includes: Add the decision-making error information to the training dataset; The operation and maintenance decision generation model is trained based on the updated training dataset.

3. The operation and maintenance decision optimization method according to claim 2, characterized in that, Training the operation and maintenance decision generation model based on the updated training dataset includes: Determine the decision error rate, wherein the decision error rate is the proportion of fault-mismatched operation and maintenance decisions to all operation and maintenance decisions; The training parameters of the operation and maintenance decision generation model are determined based on the decision error rate, wherein the training parameters include the learning rate, and the larger the decision error rate, the larger the learning rate. The operation and maintenance decision generation model is trained based on the updated training parameters.

4. The operation and maintenance decision optimization method according to claim 1, characterized in that, Detecting the image to be inspected and determining a second equipment fault in the device within the image includes: Feature extraction is performed on the image to be detected to obtain image features; Based on the image features, determine the fault identification type of the target device in the image to be detected and the fault confidence level corresponding to the fault identification type; If the fault confidence level is greater than a preset confidence threshold, the fault identification type corresponding to the fault confidence level will be determined as a second equipment fault.

5. The operation and maintenance decision optimization method according to claim 1, characterized in that, Before acquiring the image to be inspected taken by the maintenance personnel, the method further includes: The historical equipment data is input into the operation and maintenance decision generation model, wherein the historical equipment data includes at least one of the following: historical equipment operation and maintenance records, and historical equipment fault data; Obtain the target device output by the operation and maintenance decision generation model, and the first device fault corresponding to the target device; A maintenance work order is generated based on the target device and the fault of the first device, wherein the maintenance work order is used to instruct the maintenance personnel to perform an inspection of the target device; Send the maintenance work order to the maintenance personnel.

6. The operation and maintenance decision optimization method according to claim 1, characterized in that, Before reporting decision-making errors, the method further includes: Obtain feedback information from the maintenance personnel, wherein the feedback information includes text information and voice information; Extract the semantic information from the feedback information, wherein the semantic information includes at least one of the following: evaluation information on the operation and maintenance decision, evaluation information on the second equipment failure, and evaluation information on the failure matching result; The semantic information is added to the decision error information.

7. The operation and maintenance decision optimization method according to claim 1, characterized in that, After determining the second equipment fault in the target device, the method further includes: recording the second equipment fault in the operation and maintenance record, wherein the operation and maintenance record is used to generate operation and maintenance decisions.

8. An operation and maintenance decision optimization device, characterized in that, include: The acquisition module is used to acquire images to be inspected taken by maintenance personnel. The images to be inspected include target devices, which are devices that need to be inspected as recorded in the maintenance decision. The maintenance decision is generated by the maintenance decision generation model. The maintenance decision records the target devices and a first device fault corresponding to the target devices. The first device fault is predicted by the maintenance decision generation model. The detection module is used to detect the image to be detected and determine the second device fault present in the target device; The judgment module is used to determine whether the fault of the first device matches the fault of the second device and obtain the fault matching result, wherein the fault matching result includes fault matching and fault mismatch; The reporting module is used to report decision error information when the judgment result is a mismatch. The decision error information includes the image to be detected, the operation and maintenance decision and the second equipment failure. The decision error information is used to optimize the operation and maintenance decision generation model.

9. A non-volatile storage medium, characterized in that, The non-volatile storage medium stores a program, wherein when the program is executed, it controls the device where the non-volatile storage medium is located to execute the operation and maintenance decision optimization method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, include: A memory and a processor, the processor being configured to run a program stored in the memory, wherein the program, when running, executes the operation and maintenance decision optimization method according to any one of claims 1 to 7.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the operation and maintenance decision optimization method according to any one of claims 1 to 7.

Images