Equipment operation and maintenance data management system and method based on intelligent computing

Abstract

The invention discloses an equipment operation and maintenance data management system and method based on intelligent computing, and relates to the technical field of equipment operation and maintenance management.The system comprises a multi-modal data acquisition module, a data analysis module, a data processing module, a data processing module and a data processing module, and the multi-modal data acquisition module is used for collecting structured and unstructured equipment operation and maintenance data from different data sources; the preprocessing and feature extraction module is used for carrying out preprocessing and feature extraction on the collected data to generate a preliminary analysis result, the dynamic updating module is used for constructing an incremental knowledge unit according to the preliminary analysis result and an existing operation and maintenance knowledge graph, and the knowledge graph updating module is used for dynamically adjusting updating rules of the knowledge graph based on the incremental knowledge unit. The fault processing experience is optimized; according to the equipment operation and maintenance data management system and method based on intelligent computing, the dynamic updating rule of the operation and maintenance knowledge graph is regulated and controlled so as to solve the problems of historical fault processing experience inheritance and insufficient new problem adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment operation and maintenance management, and in particular to an equipment operation and maintenance data management system and method based on intelligent computing. Background Art

[0002] In the field of equipment operation and maintenance management, the equipment operation and maintenance data management system based on intelligent computing provides comprehensive support for equipment status monitoring and fault diagnosis by integrating multiple data sources such as sensor data, equipment logs, text reports, images and videos. As one of the core technologies, the operation and maintenance knowledge graph can build the relationship between equipment, failure modes, and operation and maintenance measures, and realize knowledge reasoning and intelligent decision-making. However, the existing knowledge graph update mechanism mainly relies on rule-driven or regular batch updates, which makes it difficult to fully explore and utilize historical fault handling experience, resulting in the failure to effectively pass on some key knowledge. At the same time, in the case of constantly changing equipment operating environments, newly emerging failure modes often differ greatly from existing knowledge. Traditional update methods are difficult to adjust in a timely manner, affecting the adaptability and accuracy of operation and maintenance decisions.

[0003] Furthermore, existing knowledge graph update methods lack intelligent dynamic control capabilities and rely primarily on manually set update rules. This makes it difficult to flexibly adjust update strategies based on actual O&M needs, resulting in untimely knowledge updates or information redundancy. Furthermore, due to the complex sources of O&M data and differences in data formats, update frequencies, and reliability, traditional data fusion methods face challenges in processing inconsistent and incomplete information, impacting the accuracy and usability of knowledge. Therefore, there is an urgent need for an optimization method that can dynamically control knowledge graph update rules to achieve efficient inheritance of historical experience, enhance adaptability to new problems, and improve the intelligent level of equipment O&M management. Summary of the Invention

[0004] The purpose of the present invention is to provide an equipment operation and maintenance data management system and method based on intelligent computing, which regulates the dynamic update rules of the operation and maintenance knowledge graph to solve the problem of insufficient inheritance of historical fault handling experience and adaptability to new problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an equipment operation and maintenance data management system based on intelligent computing, the system comprising: Multimodal data acquisition module, used to collect structured and unstructured equipment operation and maintenance data from different data sources; A data analysis module connected to the multimodal data acquisition module is used to preprocess and extract features from the collected data to generate preliminary analysis results; The dynamic update module connected to the data analysis module is used to construct incremental knowledge units based on the preliminary analysis results and the existing operation and maintenance knowledge graph, including counting the number of new knowledge units and the number of existing knowledge units, and calculating the relative advantages of incremental knowledge units. The specific formula is: ; Among them, A represents the relative advantage of incremental knowledge units, B represents the number of newly added knowledge units, and D represents the number of existing knowledge units; The knowledge graph update module, connected to the dynamic update module, is used to dynamically adjust the update rules of the knowledge graph based on incremental knowledge units to optimize fault handling experience. This includes determining the frequency of new knowledge and the degree of obsolescence of current knowledge, and calculating the necessity value of knowledge graph updates. The specific formula is: C=G / H; Among them, C represents the necessity value of knowledge graph update, G represents the frequency of occurrence of new knowledge, and H represents the degree of obsolescence of current knowledge.

[0006] Preferably, the multimodal data acquisition module collects structured and unstructured equipment operation and maintenance data from different data sources, including determining the signal change amplitude of the data source and calculating the data acquisition intensity. The specific formula is: ; Among them, L represents the data collection intensity, and E represents the variation range of the collected data.

[0007] Preferably, the data analysis module preprocesses and extracts features from the collected data to generate preliminary analysis results, including determining the memory occupancy rate and data analysis rate of the system's current data analysis task, and calculating the data throughput, specifically using the formula: Z=W×U; Among them, Z represents data analysis throughput, W represents memory usage, and U represents data analysis speed.

[0008] Preferably, the multimodal data acquisition module includes a structured data acquisition unit and an unstructured data acquisition unit, which are respectively used to acquire different types of data from sensors, logs, texts and images.

[0009] Preferably, the knowledge graph update module further includes a threshold adjustment unit for automatically adjusting the threshold of knowledge update based on historical data to optimize the knowledge iteration frequency.

[0010] Preferably, the knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on the incremental knowledge unit to optimize the calculation formula of the occurrence frequency G of new knowledge in the fault handling experience: G= ; Among them, G represents the frequency of occurrence of new knowledge, It represents the number of new knowledge units added during the statistical time period T, where T represents the statistical time period.

[0011] Preferably, the knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on the incremental knowledge unit to optimize the degree of obsolescence H of the current knowledge in the fault handling experience. The calculation formula is: H=N1 / N2; Among them, H represents the degree of obsolescence of current knowledge, N1 represents the number of obsolete knowledge units in the knowledge base, and N2 represents the total number of knowledge units in the knowledge base.

[0012] Preferably, the calculation formula for the data variation range E collected by the multimodal data acquisition module from structured and unstructured equipment operation and maintenance data collected from different data sources is: E= ; Among them, E represents the variation range of the collected data, P represents the number of data points in the statistical time window, represents the data value at the i-th time point, represents the data mean, and i represents the index number of the data point.

[0013] Preferably, the calculation formula of the data analysis speed U of the structured and unstructured equipment operation and maintenance data collected by the multimodal data acquisition module from different data sources is: U=R / S; Among them, U represents the data analysis speed, R represents the total amount of data analyzed, and S represents the time taken for analysis.

[0014] The equipment operation and maintenance data management method based on multimodal data fusion adopts the equipment operation and maintenance data management system based on intelligent computing, and the method includes: S1, the multimodal data acquisition module collects structured and unstructured equipment operation and maintenance data from different sources; S2. The data analysis engine preprocesses and extracts features from the collected data to generate preliminary analysis results; S3. Dynamic update model builds incremental knowledge units based on preliminary analysis results and existing operation and maintenance knowledge graph; S4. Dynamically adjust the update rules of the knowledge graph based on incremental knowledge units to optimize fault handling experience.

[0015] It can be seen from the above technical solution that the present invention has the following beneficial effects: This equipment operation and maintenance data management system and method based on intelligent computing collects structured and unstructured equipment operation and maintenance data from different data sources through a multimodal data acquisition module. The data analysis module preprocesses and extracts features from the collected data to generate preliminary analysis results. The dynamic update module constructs incremental knowledge units based on the preliminary analysis results and the existing operation and maintenance knowledge graph. The knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on the incremental knowledge units to optimize fault handling experience. It can adjust the knowledge update strategy in real time according to the equipment operating status, fault mode changes and operation and maintenance needs, avoid the knowledge lag problem caused by rule solidification or batch update methods, improve the accuracy and adaptability of operation and maintenance decisions, and ensure that key operation and maintenance knowledge can be effectively used. Efficient accumulation and inheritance, reducing misjudgment of repeated faults and inefficient repairs, improving the equipment's ability to respond to sudden problems and unknown faults, avoiding inefficient redundant updates caused by manually set rules in traditional methods, improving knowledge storage and computing efficiency, enabling the equipment operation and maintenance management system to manage massive operation and maintenance data in a lighter and more efficient manner, ensuring the accuracy of knowledge reasoning and intelligent decision-making, reducing erroneous operation and maintenance judgments caused by data inconsistency, reducing dependence on manually set rules, improving the autonomous learning ability of knowledge management, making the equipment operation and maintenance system more adaptable and intelligent decision-making, promoting the development of intelligent operation and maintenance, and regulating the dynamic update rules of the operation and maintenance knowledge graph to solve the problem of insufficient inheritance of historical fault handling experience and adaptability to new problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a connection diagram of the system modules of the present invention; Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, the present invention provides a technical solution: an equipment operation and maintenance data management system based on intelligent computing, the system comprising: Multimodal data acquisition module, used to collect structured and unstructured equipment operation and maintenance data from different data sources; A data analysis module connected to the multimodal data acquisition module is used to preprocess and extract features from the collected data to generate preliminary analysis results; The dynamic update module connected to the data analysis module is used to construct incremental knowledge units based on the preliminary analysis results and the existing operation and maintenance knowledge graph, including counting the number of new knowledge units and the number of existing knowledge units, and calculating the relative advantages of incremental knowledge units. The specific formula is: ; Among them, A represents the relative advantage of incremental knowledge units, B represents the number of newly added knowledge units, and D represents the number of existing knowledge units; The knowledge graph update module, connected to the dynamic update module, is used to dynamically adjust the update rules of the knowledge graph based on incremental knowledge units to optimize fault handling experience. This includes determining the frequency of new knowledge and the degree of obsolescence of current knowledge, and calculating the necessity value of knowledge graph updates. The specific formula is: C=G / H; Among them, C represents the necessity value of knowledge graph update, G represents the frequency of occurrence of new knowledge, and H represents the degree of obsolescence of current knowledge.

[0019] The system first collects equipment operation and maintenance data from various sources through a multimodal data acquisition module. The data analysis module then performs preprocessing, including data cleaning, noise reduction, standardization, and feature extraction, to ensure data quality and usability. The processed data then enters the dynamic update module, which compares the analysis results with the existing operation and maintenance knowledge graph, identifies newly added knowledge units, and calculates their relative strengths to determine their importance within the overall knowledge system. After statistically evaluating these incremental knowledge units, the knowledge graph update module dynamically adjusts the knowledge graph to reflect the latest equipment operation and maintenance experiences and patterns. During the knowledge update process, the system calculates the necessity value for knowledge graph updates based on the frequency of new knowledge and the degree of obsolescence of existing knowledge, ensuring the accuracy and rationality of the update strategy. Through this mechanism, the system continuously optimizes equipment operation and maintenance management, improves its ability to identify failure modes, and enhances its predictive and decision-making capabilities, ultimately achieving more efficient equipment management and maintenance. This system significantly enhances the intelligence level of equipment operation and maintenance in multiple aspects. First, through multimodal data fusion, it is possible to collect and integrate operation and maintenance information from structured and unstructured data sources, improving data integrity and accuracy and providing a richer foundation for subsequent analysis. Second, based on a dynamically updated knowledge graph, the system can learn new knowledge in real time, compensating for the limitations of traditional static knowledge bases in operation and maintenance management, and ensuring the timeliness and accuracy of the knowledge base. Furthermore, by utilizing incremental knowledge calculation methods, the system can effectively screen new knowledge, ensuring that only high-value information is incorporated into the knowledge graph, avoiding data redundancy and improving the efficiency of knowledge management. Through an adaptive knowledge update mechanism, the system can quickly respond to abnormal conditions in equipment operation, optimize fault diagnosis and maintenance strategies, thereby improving equipment reliability and availability and reducing maintenance costs. At the same time, the operation and maintenance model based on the knowledge graph can be promoted across different equipment, working conditions, and industries, enabling knowledge sharing and reuse, further enhancing the enterprise's operation and maintenance capabilities and competitiveness.

[0020] The multimodal data acquisition module collects structured and unstructured equipment operation and maintenance data from different data sources, including determining the signal change amplitude of the data source and calculating the data collection intensity. The specific formula is: ; Among them, L represents the data collection intensity, and E represents the variation range of the collected data.

[0021] The system's multimodal data acquisition module can simultaneously collect equipment operation and maintenance data from various sources, including sensor data, log information, historical records, and remote monitoring data. Because different data sources exhibit varying signal fluctuations, to more scientifically assess data collection quality, the system quantifies the signal fluctuations of each source, calculates the data variation amplitude E, and uses a logarithmic transformation to derive the data collection intensity L. This approach effectively measures the contribution of different data sources to operation and maintenance analysis, screens for high-value data, and improves the system's data utilization efficiency. Furthermore, this calculation method adapts to the data collection needs of diverse equipment and environments, ensuring accurate measurement of data variation even under complex operating conditions, providing a reliable basis for subsequent data analysis and knowledge updating. This system utilizes multimodal data fusion and data variation amplitude calculation to effectively improve the accuracy and reliability of data collection. By calculating the data collection intensity L, the system can identify the contribution of different data sources, enabling the data analysis module to prioritize high-value data with significant variations, thereby improving data utilization. Furthermore, this method adapts to fluctuations in different equipment operating environments, improving the ability to detect early signs of failure and enhancing the effectiveness of predictive maintenance. Compared with the traditional uniform collection strategy, this system can dynamically adjust the collection strategy, reduce the computing resource consumption of redundant data storage and processing, while ensuring the integrity and timeliness of key data, and ultimately improve the intelligence level of equipment operation and maintenance management.

[0022] The data analysis module preprocesses and extracts features from the collected data to generate preliminary analysis results, including determining the memory usage and data analysis rate of the system's current data analysis task and calculating the data throughput. The specific formula is: Z=W×U; Among them, Z represents data analysis throughput, W represents memory usage, and U represents data analysis speed.

[0023] The system first uses a multimodal data acquisition module to collect data from various data sources and inputs it into the data analysis module for preprocessing. This module performs data cleaning, noise reduction, normalization, and feature extraction to ensure that data quality meets analysis requirements. During data processing, the system monitors the current memory usage W and data analysis speed U and calculates the data analysis throughput Z based on these data to assess the system's analytical capabilities. If throughput Z is too low, the system can dynamically adjust computing resource allocation, such as by optimizing task scheduling or adding computing nodes, to improve analysis efficiency. Furthermore, the system supports continuous optimization of data at each stage of the analysis process to reduce computing resource consumption and improve the real-time and stability of data analysis. The system utilizes a data analysis throughput calculation method that enables real-time assessment and optimization of the system's data processing capabilities, ensuring efficient and stable operation even under high-load conditions. By monitoring memory usage and data analysis speed, the system can dynamically adjust computing resources to avoid data processing delays caused by memory overflow or computing bottlenecks. Compared to traditional methods that use fixed computing resource allocation, this system exhibits greater adaptability and can maintain efficient operation in diverse operational scenarios. In addition, this method improves the response speed and accuracy of the data analysis module, enabling equipment operation and maintenance management to more quickly identify fault signs and optimize maintenance strategies, ultimately improving equipment availability and operation and maintenance efficiency.

[0024] The multimodal data acquisition module consists of a structured data acquisition unit and an unstructured data acquisition unit, respectively, for acquiring different types of data from sensors, logs, text, and images. This system's multimodal data acquisition module utilizes these two distinct data acquisition units to efficiently integrate structured and unstructured data. The structured data acquisition unit connects to sensors, device management systems, or data warehouses using standard interfaces or protocols, periodically collecting equipment operating data and storing it in a table or database format. Simultaneously, the unstructured data acquisition unit utilizes technologies such as natural language processing (NLP), optical character recognition (OCR), and computer vision (CV) to extract key information from log text, operation and maintenance manuals, equipment photos, and video streams, and performs feature analysis and classification. These two data acquisition units work together to ensure multimodal integration at the data acquisition level, providing a complete and accurate data foundation for subsequent data analysis and knowledge building. This system utilizes a method of synchronously collecting structured and unstructured data, achieving comprehensive awareness of equipment operation and maintenance data and improving its comprehensiveness and accuracy. Compared to traditional single-type data collection methods, this system effectively integrates sensor data, text logs, and image information to analyze equipment operating status from multiple perspectives, improving the accuracy of fault prediction and health assessment. Furthermore, using technologies such as Natural Language Processing (NLP), Optical Character Recognition (OCR), and Computer Vision (CV), this system automatically parses and extracts key information from unstructured data, reducing manual intervention and improving data processing efficiency. Ultimately, this system can enhance the intelligence of equipment management, support more accurate decision-making and analysis, and enhance the automation and intelligent diagnostic capabilities of equipment operation and maintenance.

[0025] The knowledge graph update module further includes a threshold adjustment unit, which automatically adjusts the knowledge update threshold based on historical data to optimize the frequency of knowledge iteration. The knowledge graph update module in this system utilizes this threshold adjustment unit to intelligently optimize the knowledge update strategy. The system continuously monitors knowledge change trends in historical data, including the frequency of new knowledge and the decay rate of existing knowledge. It dynamically adjusts the knowledge update threshold based on fluctuations in the device operation and maintenance environment. For example, when the frequency of new knowledge is high, the system lowers the knowledge update threshold to quickly introduce new knowledge. When the stability of existing knowledge is high, the system raises the threshold to reduce unnecessary updates and maintain the stability of the knowledge graph. Furthermore, the threshold adjustment unit automatically optimizes the knowledge update frequency using machine learning algorithms (such as adaptive threshold adjustment models or Bayesian updating methods), ensuring that the knowledge graph remains efficient and accurate. Through this automated threshold adjustment mechanism, this system achieves precise control over the knowledge graph update frequency, enhancing the intelligent level of knowledge management. Compared to traditional fixed-threshold knowledge update methods, this system can adaptively adjust update strategies to ensure the real-time and rationality of knowledge updates, avoiding data inflation caused by overly rapid updates and knowledge lag caused by overly slow updates. Furthermore, based on dynamic analysis of historical data, the system can flexibly adjust knowledge update parameters for different devices and scenarios, thereby improving the adaptability of the knowledge graph, making it more consistent with actual operation and maintenance needs, and ultimately enhancing the accuracy of device management and intelligent decision-making capabilities.

[0026] The knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on incremental knowledge units to optimize the frequency of occurrence of new knowledge in fault handling experience. The calculation formula G is: G= ; Among them, G represents the frequency of occurrence of new knowledge, It represents the number of new knowledge units added during the statistical time period T, where T represents the statistical time period.

[0027] The knowledge graph update module of this system continuously monitors the number of newly added knowledge units, calculates their frequency of occurrence G within a set time range, and dynamically optimizes the knowledge graph update rules based on this value. Specifically, the system will first collect the newly added knowledge units generated during the equipment operation and maintenance process, and classify and count them according to the set statistical time period T to obtain the number of newly added knowledge units. . Subsequently, the system calculates the frequency of occurrence G of the new knowledge and compares it with the preset knowledge update threshold. When G exceeds a certain set threshold, the system will speed up the update frequency of the knowledge graph to ensure that new knowledge can be incorporated in a timely manner and optimize the operation and maintenance experience; conversely, when G is lower than the set threshold, the system will reduce the update frequency to maintain the stability of the knowledge base and avoid system overhead caused by unnecessary updates. In addition, this module can combine machine learning algorithms and historical data analysis to automatically adjust the statistical time period T to adapt to the knowledge evolution pattern under different equipment and operation and maintenance scenarios. This system realizes adaptive optimization of the knowledge graph through the calculation and dynamic adjustment strategy of the frequency of occurrence G of new knowledge, ensuring the real-time and rationality of knowledge updates. Compared with the traditional periodic update strategy, this system can make intelligent adjustments based on the actual knowledge growth, which not only avoids the problem of knowledge lag caused by too low update frequency, but also prevents data redundancy and waste of computing resources caused by excessive updates. In addition, this method can improve the system's response speed to new knowledge, enabling it to absorb and apply new experiences more quickly when faced with sudden failures or new operation and maintenance modes, thereby improving the intelligence level of equipment management, optimizing operation and maintenance efficiency, and enhancing knowledge sharing and reuse capabilities.

[0028] The knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on incremental knowledge units to optimize the degree of obsolescence H of current knowledge in fault handling experience. The calculation formula is: H=N1 / N2; Among them, H represents the degree of obsolescence of current knowledge, N1 represents the number of obsolete knowledge units in the knowledge base, and N2 represents the total number of knowledge units in the knowledge base.

[0029] The knowledge graph update module of this system continuously monitors the knowledge units in the knowledge base and evaluates the degree of obsolescence H of the knowledge based on historical data and actual operation and maintenance conditions. Specifically, the system will regularly check the knowledge units to determine which knowledge is no longer applicable to the current operation and maintenance environment (such as knowledge that has been eliminated due to equipment upgrades, changes in working conditions, or the introduction of new technologies), and mark them as obsolete knowledge units N1. At the same time, the system counts the total number of all existing knowledge units N2 in the knowledge base and calculates the degree of obsolescence H of the current knowledge. When H exceeds the set threshold, the system will accelerate the update frequency of the knowledge graph, clean up outdated knowledge in a timely manner, and introduce new knowledge to maintain the effectiveness and cutting-edge nature of the knowledge graph. In addition, the system can combine machine learning algorithms and expert feedback mechanisms to dynamically adjust the knowledge elimination criteria to ensure the accuracy of knowledge updates and avoid the accidental deletion of still valid knowledge units. By dynamically calculating the degree of obsolescence H of knowledge, this system ensures that the knowledge graph always remains efficient, accurate, and applicable. Compared to traditional knowledge management methods that rely on manual updating, this system automatically identifies and eliminates outdated knowledge, reducing knowledge base redundancy and improving system response speed, while ensuring the real-time and accurate nature of troubleshooting experience. Furthermore, this method optimizes knowledge base management strategies based on the actual use of knowledge, ensuring that equipment operation and maintenance teams always rely on the latest and most effective knowledge for decision-making. Ultimately, this system can enhance the intelligence level of equipment maintenance, reduce misdiagnosis of faults caused by outdated knowledge, and improve operation and maintenance efficiency and equipment stability.

[0030] The calculation formula for the data variation E collected by the multimodal data acquisition module from different data sources, including structured and unstructured equipment operation and maintenance data, is: E= ; Among them, E represents the variation range of the collected data, P represents the number of data points in the statistical time window, represents the data value at the i-th time point, represents the data mean, and i represents the index number of the data point.

[0031] The multimodal data acquisition module of this system will regularly collect multiple data points within the set statistical time window and calculate their variation E to evaluate the volatility of the data. Specifically, the system will extract a series of data values from multiple sensors, logs or other data sources and calculate their mean , then calculates the absolute deviation between each data point and the mean, and then takes the average to quantify data fluctuation. This method effectively measures the stability of equipment status and assists in subsequent data analysis and knowledge graph updates. For example, when E is high, it indicates significant fluctuations in equipment status, potentially requiring further analysis of potential anomalies. When E remains low, it indicates relatively stable equipment status. The system can adjust data processing strategies based on this data, such as increasing the data sampling frequency in cases of high fluctuation or reducing redundant data storage in cases of low fluctuation, to optimize data management efficiency. This system utilizes a data change amplitude calculation method to accurately measure fluctuations in equipment operating data, improving the accuracy and sensitivity of data collection. Compared to traditional single-point data analysis methods, this system eliminates random errors through multi-point statistical calculations, making data analysis results more stable and reliable. Furthermore, this method can be used for equipment anomaly detection. When equipment operating status fluctuates significantly, the system can provide early warnings, reducing the probability of equipment failure. Furthermore, by adjusting the size of the statistical time window P, the system can adapt to different types of equipment and operation and maintenance scenarios, making data processing more flexible and efficient. Ultimately, this system can improve the intelligence level of equipment operation and maintenance management, optimize fault prediction and maintenance strategies, improve operation and maintenance efficiency, and reduce maintenance costs.

[0032] The calculation formula for the data analysis speed U of the multimodal data acquisition module that collects structured and unstructured equipment operation and maintenance data from different data sources is: U=R / S; Among them, U represents the data analysis speed, R represents the total amount of data analyzed, and S represents the time taken for analysis.

[0033] After completing data collection, the system's multimodal data acquisition module performs preprocessing, feature extraction, format conversion, and analysis on the acquired data. The system monitors the total amount of analyzed data (R) (e.g., the number of processed log entries, the amount of sensor data analyzed, the number of converted image frames, etc.) and the analysis time (S) in real time to calculate the data analysis speed (U) to assess the system's data processing efficiency. If U falls below a set threshold, the system can take optimization measures, such as increasing computing resources, optimizing algorithm parallelism, or adjusting the data preprocessing process, to improve data analysis efficiency. Furthermore, this calculation method can be used for dynamic load balancing. When the system detects a decrease in U, it automatically allocates more computing resources to maintain data processing stability and ensure efficient system operation. By calculating the data analysis speed (U), the system can evaluate and optimize data processing capabilities in real time, improving the system's data analysis efficiency and responsiveness. Compared to traditional static computing resource allocation methods, this system can dynamically adjust computing resources based on the actual analysis load, ensuring efficient operation even under high data traffic. This approach also helps operations teams identify and analyze bottlenecks, such as data processing delays caused by storage I / O limitations or insufficient computing power, and provides optimization strategies. By continuously monitoring U trends, the system can proactively predict potential data processing issues, optimize data flow management, enhance the intelligence of equipment operations and management, and improve the accuracy and real-time nature of data-driven decision-making.

[0034] like Figure 2 As shown, a device operation and maintenance data management method based on multimodal data fusion is also provided, which adopts the device operation and maintenance data management system based on intelligent computing, and the method includes: S1, the multimodal data acquisition module collects structured and unstructured equipment operation and maintenance data from different sources; S2. The data analysis engine preprocesses and extracts features from the collected data to generate preliminary analysis results; S3. Dynamic update model builds incremental knowledge units based on preliminary analysis results and existing operation and maintenance knowledge graph; S4. Dynamically adjust the update rules of the knowledge graph based on incremental knowledge units to optimize fault handling experience.

[0035] This method leverages multimodal data fusion technology to enable automated collection, intelligent analysis, dynamic knowledge updating, and optimization of equipment operation and maintenance data. The system first collects operation and maintenance information from multiple data sources and uses a data analysis engine to extract key features and generate preliminary analysis results. Then, based on incremental knowledge computation methods, it dynamically identifies and evaluates newly added knowledge units. In combination with historical data, it adjusts the knowledge graph's update rules to adapt to changes in equipment operation and maintenance scenarios. Through this mechanism, the system continuously optimizes equipment management processes, improving both operation and maintenance efficiency and intelligence.

[0036] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. Equipment operation and maintenance data management system based on intelligent computing, characterized by: The system comprises: Multimodal data acquisition module, used to collect structured and unstructured equipment operation and maintenance data from different data sources; A data analysis module connected to the multimodal data acquisition module is used to preprocess and extract features from the collected data to generate preliminary analysis results; The dynamic update module connected to the data analysis module is used to construct incremental knowledge units based on the preliminary analysis results and the existing operation and maintenance knowledge graph, including counting the number of new knowledge units and the number of existing knowledge units, and calculating the relative advantages of incremental knowledge units. The specific formula is: ; Among them, A represents the relative advantage of incremental knowledge units, B represents the number of newly added knowledge units, and D represents the number of existing knowledge units; The knowledge graph update module, connected to the dynamic update module, is used to dynamically adjust the update rules of the knowledge graph based on incremental knowledge units to optimize fault handling experience. This includes determining the frequency of new knowledge and the degree of obsolescence of current knowledge, and calculating the necessity value of knowledge graph updates. The specific formula is: C=G / H; Among them, C represents the necessity value of knowledge graph update, G represents the frequency of occurrence of new knowledge, and H represents the degree of obsolescence of current knowledge.

2. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The multimodal data acquisition module collects structured and unstructured equipment operation and maintenance data from different data sources, including determining the signal change amplitude of the data source and calculating the data acquisition intensity. The specific formula is: ; Among them, L represents the data collection intensity, and E represents the variation range of the collected data.

3. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The data analysis module preprocesses and extracts features from the collected data to generate preliminary analysis results, including determining the memory occupancy and data analysis rate of the system's current data analysis task and calculating the data throughput. The specific formula is: Z=W×U; Among them, Z represents data analysis throughput, W represents memory usage, and U represents data analysis speed.

4. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The multimodal data acquisition module includes a structured data acquisition unit and an unstructured data acquisition unit, which are respectively used to acquire different types of data from sensors, logs, texts and images.

5. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The knowledge graph update module further includes a threshold adjustment unit for automatically adjusting the threshold of knowledge update based on historical data to optimize the knowledge iteration frequency.

6. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on the incremental knowledge units to optimize the calculation formula of the frequency of occurrence G of new knowledge in the fault handling experience: G= ； Among them, G represents the frequency of occurrence of new knowledge, It represents the number of new knowledge units added during the statistical time period T, where T represents the statistical time period.

7. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The knowledge graph update module dynamically adjusts the update rules of the knowledge graph based on the incremental knowledge units to optimize the degree of obsolescence H of the current knowledge in the fault handling experience. The calculation formula is: H=N1 / N2; Among them, H represents the degree of obsolescence of current knowledge, N1 represents the number of obsolete knowledge units in the knowledge base, and N2 represents the total number of knowledge units in the knowledge base.

8. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The calculation formula for the data variation range E collected by the multimodal data acquisition module from structured and unstructured equipment operation and maintenance data collected from different data sources is: E= ； Among them, E represents the variation range of the collected data, P represents the number of data points in the statistical time window, represents the data value at the i-th time point, represents the data mean, and i represents the index number of the data point.

9. The equipment operation and maintenance data management system based on intelligent computing according to claim 1, characterized in that: The calculation formula of the data analysis speed U of the multimodal data acquisition module in collecting structured and unstructured equipment operation and maintenance data from different data sources is: U=R / S; Among them, U represents the data analysis speed, R represents the total amount of data analyzed, and S represents the time taken for analysis.

10. A method for managing equipment operation and maintenance data based on intelligent computing, comprising: The method comprises: S1, the multimodal data acquisition module collects structured and unstructured equipment operation and maintenance data from different sources; S2. The data analysis engine preprocesses and extracts features from the collected data to generate preliminary analysis results; S3. Dynamic update model builds incremental knowledge units based on preliminary analysis results and existing operation and maintenance knowledge graph; S4. Dynamically adjust the update rules of the knowledge graph based on incremental knowledge units to optimize fault handling experience.

Images