An intelligent fault analysis and management system and method based on big data

Abstract

The application provides a big data-based intelligent fault analysis and management system and method, which comprises the following steps: using a Spring Cloud technology stack to develop an architecture, generating a multidimensional fault management platform corresponding to each service station, using the multidimensional fault management platform to edit and create a manual library of the service station, generating a fault identification list corresponding to each service station, determining the fault characteristics of the corresponding service station according to real-time data corresponding to each service station, identifying the corresponding fault characteristics in the fault identification list, determining the fault information corresponding to each service station, deeply mining the fault information in the multidimensional fault management platform, determining the remediation method corresponding to each service station and displaying the remediation method, effectively solving the problems of fault handling, drawing management, data analysis and operation and maintenance efficiency in the prior art, and providing strong support for intelligent operation and maintenance transformation.

Description

Technical Field

[0001] This invention relates to the field of fault analysis and management technology, and in particular to an intelligent fault analysis and management system and method based on big data. Background Technology

[0002] In current management systems, traditional fault handling methods have shown significant limitations in the face of increasingly large and complex power equipment networks. These methods often rely on the experience and manual records of maintenance personnel, which is not only inefficient but also fails to guarantee the timeliness and accuracy of fault handling. Manuals and drawings are typically managed in a fragmented manner, lacking a unified online platform, causing maintenance personnel to spend a significant amount of time searching for relevant information during fault handling. Furthermore, the statistical analysis of equipment fault data largely depends on manual compilation, making it difficult to achieve in-depth data mining and intelligent analysis, thus failing to provide strong support for maintenance decision-making. Moreover, traditional methods of querying based on paper fault manuals or electronic documents are no longer sufficient for rapid equipment fault handling. Paper fault manuals are cumbersome and inconvenient to search, while electronic documents suffer from inconvenient searching, poor user experience, and low query efficiency, increasing the difficulty of maintenance and repair for new employees and reducing maintenance efficiency.

[0003] Therefore, the present invention provides an intelligent fault analysis and management system and method based on big data. Summary of the Invention

[0004] This invention provides an intelligent fault analysis and management system and method based on big data, which effectively solves the problems existing in the prior art in terms of fault handling, drawing management, data analysis and operation and maintenance efficiency, and provides strong support for intelligent operation and maintenance transformation.

[0005] This invention provides an intelligent fault analysis and management system based on big data, comprising:

[0006] The multi-dimensional management module is used to develop the architecture using the Spring Cloud technology stack and generate a multi-dimensional fault management platform for each service station.

[0007] The data management module is used to edit and create the manual library of the service station using the multi-dimensional fault management platform, and generate a fault identification list corresponding to each service station.

[0008] The analysis and implementation module is used to determine the fault characteristics of the corresponding service station based on the real-time data of each service station, identify the corresponding fault characteristics in the fault identification list, and determine the fault information corresponding to each service station.

[0009] The fault management module is used to perform in-depth analysis of the fault information in the multi-dimensional fault management platform, determine the corresponding compensation method for each service station, and display it.

[0010] In one feasible approach

[0011] The multi-dimensional fault management platform includes:

[0012] The manual management sub-platform is used to classify the corresponding manual library according to the prescribed document structure using semantic analysis, and generate manual documents for each of the service stations.

[0013] The instruction response sub-platform is used to locate and display the corresponding target manual in the manual document based on the search instructions issued by the operation and maintenance personnel;

[0014] The offline buffering sub-platform is used to buffer the aforementioned manual documents offline.

[0015] In one feasible approach

[0016] The multidimensional management module includes:

[0017] The microservice architecture unit is used to obtain the site location information and working attribute information corresponding to each service station, use the site location information and working attribute information to create the initial analysis management platform, and use the Spring Cloud technology stack to divide the initial analysis management platform into several microservices;

[0018] The functional analysis unit is used to trace the source of each microservice, determine the service object corresponding to each microservice, collect the functional attribute information corresponding to each service object, analyze the topological relationship between different functional attribute information according to the distribution of the microservice in the initial analysis management platform, and generate a functional topology network.

[0019] The functional deployment unit is used to generate functional annotations for corresponding service objects based on the node data corresponding to each topology node in the functional topology network, determine the docking function corresponding to each service object in combination with the network logic of the functional topology network, and deploy the corresponding microservices using the docking function to obtain the available microservices corresponding to each service station.

[0020] The platform construction unit is used to identify high-frequency faults corresponding to each service station based on the site location information and working attribute information of each service station, enhance the functionality of the available microservices using the high-frequency faults, and establish a multi-dimensional fault management platform for each service station in combination with preset platform construction technology.

[0021] In one feasible approach

[0022] The platform construction unit includes:

[0023] The fault identification subunit is used to draw a service station functional structure diagram based on the initial analysis management platform, determine the dynamic and static working information of the corresponding service station according to the site location information and working attribute information of each service station, and sequentially delete each dynamic and static working information in the service station functional structure diagram to obtain the dynamic fault threshold and static fault threshold corresponding to each service station.

[0024] The fault mining subunit is used to generate several dynamic and static faults corresponding to each service station by combining each available microservice with the corresponding dynamic fault threshold and static fault threshold, retrieve historical faults of similar service stations in the big data center based on the working attribute information of the service station, and determine several high-frequency faults of each service station based on the first similarity between each dynamic fault and the historical fault and the second similarity between each static fault and the historical fault.

[0025] The service enhancement subunit is used to obtain the fault level and fault performance corresponding to each of the high-frequency faults, establish mirror service information using the fault level and fault performance, enhance the corresponding available microservices using the mirror service information, and obtain the preferred microservices corresponding to each service station.

[0026] The platform construction subunit is used to build the platform structure based on the interaction between the front-end framework, back-end technology and database corresponding to the initial analysis management platform. Each of the preferred microservices is input into the platform structure to generate the preferred management platform. The preferred management platform is used to find the sub-platform corresponding to each service station and perform functional repair to obtain the multi-dimensional fault management platform corresponding to each service station.

[0027] In one feasible approach

[0028] The data management module includes:

[0029] The data collection unit is used to search for the operation and maintenance data corresponding to each similar service station in the big data center, establish several operation and maintenance data information corresponding to the service station based on the operation and maintenance data, and determine the data description features corresponding to each operation and maintenance data information based on semantic analysis technology.

[0030] The data filtering unit is used to locate contradictory operation and maintenance data information with opposite data description characteristics, upload the contradictory operation and maintenance information to the big data service center for information search, obtain the objective support quantity corresponding to each contradictory operation and maintenance data information, eliminate unqualified operation and maintenance data information based on the objective support quantity, obtain several qualified operation and maintenance data information corresponding to each service station, and build a manual library corresponding to each service station respectively.

[0031] The fault identification unit is used to generate several fault descriptions for corresponding service stations in the multi-dimensional fault management platform, search for several matching information in the manual library according to the keywords corresponding to each fault description, and determine the response information of the corresponding service station to each fault description based on the matching information, thus forming a fault identification list.

[0032] In one feasible approach

[0033] The analysis implementation module includes:

[0034] The data processing unit is used to collect real-time data generated by each of the service stations, obtain the data period corresponding to each of the real-time data, determine the truncation distance corresponding to each of the real-time data according to the data period, use the truncation distance to cut the corresponding real-time data into several single data segments, and perform data sampling in each of the single data segments corresponding to the same real-time data segment to obtain several sampled data.

[0035] The feature generation unit is used to obtain the data similarity between different sampled data, determine the correlation features between different single data segments, perform clustering processing on the sub-data contained in each single data segment to obtain the outlier corresponding to each single data segment, analyze the correlation outlier corresponding to each outlier based on the correlation features, and establish the fault features of the corresponding service station based on the distribution information of the correlation outlier in the real-time data.

[0036] The fault identification unit is used to search for matching information corresponding to each fault feature in the fault identification list, establish a fault knowledge graph for each service station based on the matching information corresponding to each service station, and generate several fault information of the service station based on the fault level corresponding to each fault feature and the fault knowledge graph.

[0037] In one feasible approach

[0038] The fault management module includes:

[0039] The fault decomposition unit is used to perform initial identification of each fault information in the multi-dimensional fault management platform, obtain the dimension coverage information corresponding to each fault information, and perform multi-scale coarse-grained processing on each fault information based on the dimension coverage information to obtain several single-dimensional fault information corresponding to each service station.

[0040] The fault mining unit is used to establish a fault damage model corresponding to the service station based on the fault dimension corresponding to each single-dimensional fault information, and input the single-dimensional fault information corresponding to the same service station into the fault damage model for deep training to obtain the explicit fault and implicit fault corresponding to each service station.

[0041] The fault compensation unit is used to trace the source of each explicit fault in the fault damage model, obtain the first fault device corresponding to each explicit fault, enhance each implicit fault, run the fault damage model to determine the second fault device corresponding to each enhanced implicit fault, obtain the manual information corresponding to each first fault device and the second fault device from the manual library, establish the compensation method corresponding to each fault information and display it.

[0042] In one feasible approach

[0043] The fault compensation unit includes:

[0044] The first fault tracing subunit is used to run the fault damage model to obtain the first fault device corresponding to each of the explicit faults, obtain the fault level classification corresponding to each of the implicit faults, determine the maximum fault level classification corresponding to each of the implicit faults, and enhance the implicit faults to the corresponding maximum fault level classification to obtain the enhanced implicit faults.

[0045] The second fault tracing subunit is used to treat the enhanced latent fault as the model running result, adjust the model running data in the fault damage model using the model running result, and determine the second fault device corresponding to the latent fault based on the adjustment direction corresponding to the model running data.

[0046] The fault compensation execution subunit is used to count a number of first faulty devices and second faulty devices corresponding to each service station, filter the manual information corresponding to each first faulty device and second faulty device in the corresponding manual library, determine the maintenance method corresponding to each faulty device based on the manual information, merge the maintenance methods to obtain the compensation method corresponding to each service station and display it.

[0047] In one feasible approach

[0048] Also includes:

[0049] The fusion execution subunit is used to match priority weights for each of the first faulty devices and to match non-priority weights for each of the second faulty devices.

[0050] Based on the priority weight and non-priority weight, a fusion threshold corresponding to the maintenance method is determined, and the maintenance methods are fused to generate a compensation method based on the fusion threshold.

[0051] This invention provides an intelligent fault analysis and management method based on big data, comprising:

[0052] Step 1: Develop the architecture using the Spring Cloud technology stack to generate a multi-dimensional fault management platform for each service station;

[0053] Step 2: Use the multi-dimensional fault management platform to edit and create the manual library of the service station, and generate a fault identification list for each service station;

[0054] Step 3: Determine the fault characteristics of each service station based on the real-time data of each service station, identify the corresponding fault characteristics in the fault identification list, and determine the fault information corresponding to each service station.

[0055] Step 4: In the multi-dimensional fault management platform, perform in-depth mining of the fault information, determine the corresponding compensation method for each service station, and display it.

[0056] The beneficial effects of the above technical solution are as follows: To improve the efficiency of fault analysis and management, the architecture is developed using the easy integration features of the Spring Cloud technology stack. A multi-dimensional fault management platform is established for each service station. Within this platform, manuals are edited to create a fault identification list for each service station. During service station operation, real-time data generated by each station is collected and used to construct fault characteristics for each station. Corresponding faults are then identified in the fault identification list, determining the fault information for each service station. Further in-depth analysis of the fault information is conducted within the multi-dimensional fault management platform. Finally, fault mitigation methods are established for each service station. This technical approach enables efficient guidance and management of power operation and maintenance faults, effectively addressing issues in fault handling, drawing management, data analysis, and operation and maintenance efficiency. It provides strong support for intelligent operation and maintenance transformation and allows for intuitive and convenient display of fault trends across various dimensions to maintenance management personnel, enabling precise guidance for maintenance operations, proactive planning, and improved service station availability.

[0057] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0058] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0059] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0060] Figure 1 This is a schematic diagram illustrating the composition of an intelligent fault analysis and management system based on big data, as described in an embodiment of the present invention.

[0061] Figure 2 This is a schematic diagram illustrating the workflow of an intelligent fault analysis and management method based on big data, as described in an embodiment of the present invention. Detailed Implementation

[0062] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0063] Example 1

[0064] This embodiment provides an intelligent fault analysis and management system based on big data, such as... Figure 1 As shown, it includes:

[0065] The multi-dimensional management module is used to develop the architecture using the Spring Cloud technology stack and generate a multi-dimensional fault management platform for each service station.

[0066] The data management module is used to edit and create the manual library of the service station using the multi-dimensional fault management platform, and generate a fault identification list corresponding to each service station.

[0067] The analysis and implementation module is used to determine the fault characteristics of the corresponding service station based on the real-time data of each service station, identify the corresponding fault characteristics in the fault identification list, and determine the fault information corresponding to each service station.

[0068] The fault management module is used to perform in-depth analysis of the fault information in the multi-dimensional fault management platform, determine the corresponding compensation method for each service station, and display it.

[0069] In this example, the Spring Cloud technology stack is a tool for building microservices;

[0070] In this example, the service station represents a standalone wind power station;

[0071] In this example, the manual library represents a database that stores files on various devices and components in a service station, and each service station is equipped with one manual library;

[0072] In this example, the fault identification list contains all the faults that a service station may experience;

[0073] In this example, real-time data represents data generated by a server during operation;

[0074] In this example, the editing and creation process is the process of organizing the manual library;

[0075] In this example, the multidimensional fault management platform refers to the platform used to analyze and manage service stations. Each service station is configured with a multidimensional fault management platform, and the multidimensional fault management platform supports web and mobile application access, ensuring that maintenance personnel can perform fault troubleshooting and handling anytime and anywhere.

[0076] In this example, the fault characteristics represent the features exhibited at the corresponding fault location when a service station experiences a fault;

[0077] In this example, the fault information represents the descriptive information about the fault contained in the service station;

[0078] In this example, deep mining refers to the process of locating the location of existing faults in a service station and uncovering potential faults.

[0079] In this example, the multi-dimensional fault management platform can be used to display the remediation methods for each service station.

[0080] The working principle and beneficial effects of the above technical solution are as follows: To improve the efficiency of fault analysis and management, the architecture is developed using the easy integration features of the Spring Cloud technology stack. A multi-dimensional fault management platform is established for each service station. Within this platform, manuals are edited to create a fault identification list for each service station. During service station operation, real-time data generated by each station is collected and used to construct fault characteristics for each station. Corresponding faults are then identified in the fault identification list, determining the fault information for each service station. Further in-depth analysis of the fault information is conducted within the multi-dimensional fault management platform, and finally, fault mitigation methods are established for each service station. This technical approach enables efficient guidance and management of power operation and maintenance faults, effectively addressing problems in fault handling, drawing management, data analysis, and operation and maintenance efficiency. It provides strong support for intelligent operation and maintenance transformation and can intuitively and conveniently display fault trends across various dimensions to maintenance management personnel, accurately guiding maintenance operations, proactively planning for future issues, and improving the availability of service stations.

[0081] Example 2

[0082] Based on Example 1, the intelligent fault analysis and management system based on big data, the multi-dimensional fault management platform, includes:

[0083] The manual management sub-platform is used to classify the corresponding manual library according to the prescribed document structure using semantic analysis, and generate manual documents for each of the service stations.

[0084] The instruction response sub-platform is used to locate and display the corresponding target manual in the manual document based on the search instructions issued by the operation and maintenance personnel;

[0085] The offline buffering sub-platform is used to buffer the aforementioned manual documents offline.

[0086] In this example, the fault handling manuals are classified and managed according to factors such as manufacturer, model, and type of equipment failure, forming a structured manual library;

[0087] In this example, to address the network instability issue in offshore wind farms, the system added a file caching function, allowing maintenance personnel to cache troubleshooting manuals and drawing libraries to mobile devices for offline viewing. Furthermore, the system also added a batch import function for standard templates of troubleshooting manuals to save data entry time and improve user experience.

[0088] In this example, operations and maintenance personnel can quickly find the required documents and perform approval operations using tags or keywords.

[0089] The working principle and beneficial effects of the above technical solution are as follows: In order to improve the standardization and efficiency of document management and solve some problems in actual use, the manual library has been organized so that maintenance personnel can respond quickly when looking up manuals, thereby improving the work efficiency of maintenance personnel. This makes the system not only suitable for ordinary power plants, but also for power plants in various complex environments.

[0090] Example 3

[0091] Based on Example 1, the intelligent fault analysis and management system based on big data, wherein the multi-dimensional management module includes:

[0092] The microservice architecture unit is used to obtain the site location information and working attribute information corresponding to each service station, use the site location information and working attribute information to create the initial analysis management platform, and use the Spring Cloud technology stack to divide the initial analysis management platform into several microservices;

[0093] The functional analysis unit is used to trace the source of each microservice, determine the service object corresponding to each microservice, collect the functional attribute information corresponding to each service object, analyze the topological relationship between different functional attribute information according to the distribution of the microservice in the initial analysis management platform, and generate a functional topology network.

[0094] The functional deployment unit is used to generate functional annotations for corresponding service objects based on the node data corresponding to each topology node in the functional topology network, determine the docking function corresponding to each service object in combination with the network logic of the functional topology network, and deploy the corresponding microservices using the docking function to obtain the available microservices corresponding to each service station.

[0095] The platform construction unit is used to identify high-frequency faults corresponding to each service station based on the site location information and working attribute information of each service station, enhance the functionality of the available microservices using the high-frequency faults, and establish a multi-dimensional fault management platform for each service station in combination with preset platform construction technology.

[0096] In this example, a microservice represents an independent service system, with one service station corresponding to one microservice;

[0097] In this example, the service object refers to the object of the microservice, that is, the device or equipment that the microservice serves in the service station;

[0098] In this example, the functional attribute information represents the executable functions and basic attributes of the service;

[0099] In this example, the functional topology network represents a functional cross-network composed of the relationships between various functional attribute information;

[0100] In this example, a topology node corresponds to a service object. A topology node represents a node in a functional topology network and contains corresponding functional attribute information and the associations between functions.

[0101] In this example, functional annotations represent a way to explain the corresponding function of a service using concise language;

[0102] In this example, the docking function refers to the ability of a service object to dock with other service objects;

[0103] In this example, microservices can be used to represent microservices that can perform corresponding analysis on the functions of the service station;

[0104] In this example, high-frequency faults refer to faults that frequently occur at the service station;

[0105] In this example, the platform building technology can be the platform orchestration technology process;

[0106] In this example, the purpose of enhancing the available microservices is to strengthen their functionality so that they are better suited to high-frequency failures.

[0107] The working principle and beneficial effects of the above technical solution are as follows: To enable each service station to manage itself, a preliminary analysis and management platform for all service stations is first established based on the site location and work attribute information of each service station. This platform is then further divided into several microservices using a technology stack, assigning a microservice to each service station. The microservices are then traced to determine their service objects. A functional topology network is constructed based on the functional attribute information of each service object. This network is then analyzed and functionally deployed to obtain the available microservices for each service station. Finally, high-frequency faults for each service station are identified, and these faults are used to enhance the functionality of available microservices. Finally, a multi-dimensional fault management platform is established for each service station using platform building technology. This not only provides each service station with a self-managing multi-dimensional fault management platform but also ensures that each platform can accurately manage the service station, reducing redundant functions and improving the efficiency and quality of the multi-dimensional fault management platform.

[0108] Example 4

[0109] Based on Example 3, the platform construction unit of the intelligent fault analysis and management system based on big data includes:

[0110] The fault identification subunit is used to draw a service station functional structure diagram based on the initial analysis management platform, determine the dynamic and static working information of the corresponding service station according to the site location information and working attribute information of each service station, and sequentially delete each dynamic and static working information in the service station functional structure diagram to obtain the dynamic fault threshold and static fault threshold corresponding to each service station.

[0111] The fault mining subunit is used to generate several dynamic and static faults corresponding to each service station by combining each available microservice with the corresponding dynamic fault threshold and static fault threshold, retrieve historical faults of similar service stations in the big data center based on the working attribute information of the service station, and determine several high-frequency faults of each service station based on the first similarity between each dynamic fault and the historical fault and the second similarity between each static fault and the historical fault.

[0112] The service enhancement subunit is used to obtain the fault level and fault performance corresponding to each of the high-frequency faults, establish mirror service information using the fault level and fault performance, enhance the corresponding available microservices using the mirror service information, and obtain the preferred microservices corresponding to each service station.

[0113] The platform construction subunit is used to build the platform structure based on the interaction between the front-end framework, back-end technology and database corresponding to the initial analysis management platform. Each of the preferred microservices is input into the platform structure to generate the preferred management platform. The preferred management platform is used to find the sub-platform corresponding to each service station and perform functional repair to obtain the multi-dimensional fault management platform corresponding to each service station.

[0114] In this example, the service station functional structure diagram shows the result of arranging all the designed functions of the service station according to its structure;

[0115] In this example, dynamic work information refers to the information generated when dynamic work is performed in the service station, such as the information generated when the motor rotates;

[0116] In this example, static work information refers to the information generated when static work is performed at the service station, such as the place of origin information when using a display screen;

[0117] In this example, the dynamic fault threshold represents the threshold at which a dynamic fault occurs in the service station;

[0118] In this example, the static fault threshold represents the threshold at which a static fault occurs in the service station;

[0119] In this example, "same type service station" refers to other service stations that correspond to the service station type.

[0120] In this example, historical faults refer to faults exhibited by similar service stations;

[0121] In this example, the first similarity represents the similarity between the dynamic faults and historical faults of the service station, and the second similarity represents the similarity between the static faults and historical faults of the service station.

[0122] In this example, high-frequency faults represent dynamic / static faults with a similarity greater than 45%;

[0123] In this example, the fault level represents the number of levels that high-frequency faults can be divided into;

[0124] In this example, fault performance refers to the performance of the service station under the influence of a high-frequency fault.

[0125] In this example, the mirror service information represents the information obtained after mirroring the fault level and fault performance. The essence of the mirror service information is information that is completely opposite to that of high-frequency faults.

[0126] In this example, preferred microservices refer to microservices that can better manage service stations;

[0127] In this example, the front-end framework, back-end technology, and database are all part of the preset platform building technology. The front-end framework can be React or Vue, the back-end technology can be Node.js or Django, and the database can be MySQL or MongoDB.

[0128] In this example, the platform architecture represents the design of the system architecture, including the front-end, back-end, database, and the ways in which they interact.

[0129] In this example, the purpose of the feature fix is ​​to improve the functionality of each sub-platform.

[0130] In this example, one service station corresponds to one sub-platform.

[0131] The working principle and beneficial effects of the above technical solution are as follows: First, the service station functional structure diagram is drawn using the initial analysis management platform to determine the dynamic and static working information of each service station. Dynamic and static fault thresholds for each service station are determined through deletion simulation. Then, the dynamic and static faults of the service station are analyzed based on available microservices. Next, the historical faults of similar service stations retrieved from the big data center are compared to identify the high-frequency faults of each service station. By establishing mirror service information for each high-frequency fault, the functionality of available microservices is enhanced, generating optimal microservices for each service station. Finally, platform building technology is used to repair the functionality of the sub-platforms corresponding to each service station, ultimately resulting in a multi-dimensional fault management platform for each service station. This approach allows for functional analysis and fault diagnosis of each service station, with a focus on analyzing and processing high-frequency faults. Building a multi-dimensional fault management platform for each service station improves its self-management efficiency.

[0132] Example 5

[0133] Based on Example 1, the data management module of the intelligent fault analysis and management system based on big data includes:

[0134] The data collection unit is used to search for the operation and maintenance data corresponding to each similar service station in the big data center, establish several operation and maintenance data information corresponding to the service station based on the operation and maintenance data, and determine the data description features corresponding to each operation and maintenance data information based on semantic analysis technology.

[0135] The data filtering unit is used to locate contradictory operation and maintenance data information with opposite data description characteristics, upload the contradictory operation and maintenance information to the big data service center for information search, obtain the objective support quantity corresponding to each contradictory operation and maintenance data information, eliminate unqualified operation and maintenance data information based on the objective support quantity, obtain several qualified operation and maintenance data information corresponding to each service station, and build a manual library corresponding to each service station respectively.

[0136] The fault identification unit is used to generate several fault descriptions for corresponding service stations in the multi-dimensional fault management platform, search for several matching information in the manual library according to the keywords corresponding to each fault description, and determine the response information of the corresponding service station to each fault description based on the matching information, thus forming a fault identification list.

[0137] In this example, the data period refers to the period at which relevant data is repeated in real-time data;

[0138] In this example, the truncation distance is consistent with the length of the data period. The truncation distance represents the standard for dividing real-time data into several single data segments. The length of each single data segment is consistent with the truncation distance. Since the data period may be different, the truncation distance corresponding to each real-time data may also be different.

[0139] In this example, the sampling method is as follows: random sampling is performed in each single data segment, and the number of samples is 1.

[0140] In this example, clustering refers to the process of grouping sub-data with the same characteristics into one class;

[0141] In this example, outliers represent data values ​​corresponding to sub-data that failed to be grouped with other sub-data.

[0142] In this example, the fault knowledge graph represents a graph used to describe the relationships between various faults presented by the service station;

[0143] In this example, the fault information represents the information presented when a service station experiences a failure.

[0144] The working principle and beneficial effects of the above technical solution are as follows: To perform corresponding fault analysis on each service station, real-time data generated during the service station's operation is collected. Then, based on the data period corresponding to each real-time data point, a truncation distance is determined, thus truncating the real-time data into several single data segments. Further data sampling is performed, and the correlation characteristics between single data segments are determined based on the data similarity between different sampled data. Then, the sub-data within each single data segment are clustered, and outliers in each single data segment are filtered out, thereby identifying the associated outliers for each service station. Combining the distribution of associated outliers in the real-time data, the fault characteristics of each service station are determined. Then, by searching for matching information for each fault characteristic, a fault knowledge graph of the service station is established. Combined with the corresponding fault characteristics and fault levels, fault information for the service station is generated. In this way, real-time analysis of the service station can be performed to determine its fault characteristics at different times and to identify its fault information. This lays the foundation for subsequent fault management and allows maintenance personnel to promptly detect and handle faults, reducing their impact.

[0145] Example 7

[0146] Based on Example 1, the intelligent fault analysis and management system based on big data, wherein the fault management module includes:

[0147] The fault decomposition unit is used to perform initial identification of each fault information in the multi-dimensional fault management platform, obtain the dimension coverage information corresponding to each fault information, and perform multi-scale coarse-grained processing on each fault information based on the dimension coverage information to obtain several single-dimensional fault information corresponding to each service station.

[0148] The fault mining unit is used to establish a fault damage model corresponding to the service station based on the fault dimension corresponding to each single-dimensional fault information, and input the single-dimensional fault information corresponding to the same service station into the fault damage model for deep training to obtain the explicit fault and implicit fault corresponding to each service station.

[0149] The fault compensation unit is used to trace the source of each explicit fault in the fault damage model, obtain the first fault device corresponding to each explicit fault, enhance each implicit fault, run the fault damage model to determine the second fault device corresponding to each enhanced implicit fault, obtain the manual information corresponding to each first fault device and the second fault device from the manual library, establish the compensation method corresponding to each fault information and display it.

[0150] In this example, the dimension coverage information represents the dimensions involved in a fault message;

[0151] In this example, the processing scale of multi-scale coarse-graining is the same as the coverage of information in the dimension.

[0152] In this example, the fault damage model represents a model of the damage caused to a service station by various faults.

[0153] In this example, a visible fault represents a fault that the service station is currently exhibiting, while a hidden fault represents a potential fault that the service station may have. That is, when a visible fault does not exist, a hidden fault may disappear.

[0154] In this example, the first faulty device refers to the device that is currently faulty, and the second faulty device refers to the device that is not currently faulty but has the potential to be faulty.

[0155] The working principle and beneficial effects of the above technical solution are as follows: To facilitate timely handling of service station faults by maintenance personnel, each fault information is initially identified in the multi-dimensional fault management platform to determine its corresponding dimensional information. Then, the fault information undergoes multi-scale coarsening processing to identify several single-dimensional fault information within the service station. Furthermore, a fault damage model for the service station is established by comprehensively considering the different fault dimensions corresponding to each service station. Explicit and implicit faults of the service station are determined through deep training. Faulty components are identified through source tracing, and relevant manual information is retrieved from a manual library. Finally, a mitigation method is established based on the retrieved manual information. This allows for simultaneous mitigation of multiple faults, improving the efficiency of maintenance personnel in resolving problems, ensuring the quality of problem-solving, and reducing losses.

[0156] Example 8

[0157] Based on Example 7, the fault compensation unit of the intelligent fault analysis and management system based on big data includes:

[0158] The first fault tracing subunit is used to run the fault damage model to obtain the first fault device corresponding to each of the explicit faults, obtain the fault level classification corresponding to each of the implicit faults, determine the maximum fault level classification corresponding to each of the implicit faults, and enhance the implicit faults to the corresponding maximum fault level classification to obtain the enhanced implicit faults.

[0159] The second fault tracing subunit is used to treat the enhanced latent fault as the model running result, adjust the model running data in the fault damage model using the model running result, and determine the second fault device corresponding to the latent fault based on the adjustment direction corresponding to the model running data.

[0160] The fault compensation execution subunit is used to count a number of first faulty devices and second faulty devices corresponding to each service station, filter the manual information corresponding to each first faulty device and second faulty device in the corresponding manual library, determine the maintenance method corresponding to each faulty device based on the manual information, merge the maintenance methods to obtain the compensation method corresponding to each service station and display it.

[0161] In this example, the maximum fault level classification represents the maximum level of a latent fault;

[0162] In this example, the purpose of generating enhanced impact faults is to facilitate better device traceability.

[0163] The working principle and beneficial effects of the above technical solution are as follows: When forming a compensation method, the first faulty device and the latent fault corresponding to the manifest fault are first determined according to the fault damage model. At the same time, the quality of tracing is improved by strengthening the latent fault, and the second faulty device corresponding to the latent fault is determined. Then, the manual library is searched to construct the compensation method of the service station. In this way, multiple maintenance methods can be integrated to handle multiple faults at one time and reduce unnecessary processing steps.

[0164] Example 9

[0165] Based on Example 8, the intelligent fault analysis and management system based on big data further includes:

[0166] The fusion execution subunit is used to match priority weights for each of the first faulty devices and to match non-priority weights for each of the second faulty devices.

[0167] Based on the priority weight and non-priority weight, a fusion threshold corresponding to the maintenance method is determined, and the maintenance methods are fused to generate a compensation method based on the fusion threshold.

[0168] In this example, during the fusion process, the maintenance methods containing non-priority weights are adjusted first, and then the maintenance methods containing priority weights are adjusted to ensure that the quality of maintenance methods containing priority weights is reduced.

[0169] The working principle and beneficial effects of the above technical solution are as follows: When merging, the maintenance method of the first faulty device is fine-tuned, and the maintenance method of the second faulty device is adjusted in multiple ways to ensure the quality of handling obvious faults.

[0170] Example 10

[0171] This example provides an intelligent fault analysis and management method based on big data, including:

[0172] Step 1: Develop the architecture using the Spring Cloud technology stack to generate a multi-dimensional fault management platform for each service station;

[0173] Step 2: Use the multi-dimensional fault management platform to edit and create the manual library of the service station, and generate a fault identification list for each service station;

[0174] Step 3: Determine the fault characteristics of each service station based on the real-time data of each service station, identify the corresponding fault characteristics in the fault identification list, and determine the fault information corresponding to each service station.

[0175] Step 4: In the multi-dimensional fault management platform, perform in-depth mining of the fault information, determine the corresponding compensation method for each service station, and display it.

[0176] In this example, the Spring Cloud technology stack is a tool for building microservices;

[0177] In this example, the service station represents a standalone wind power station;

[0178] In this example, the manual library represents a database that stores files on various devices and components in a service station, and each service station is equipped with one manual library;

[0179] In this example, the fault identification list contains all the faults that a service station may experience;

[0180] In this example, real-time data represents data generated by a server during operation;

[0181] In this example, the editing and creation process is the process of organizing the manual library;

[0182] In this example, the multidimensional fault management platform refers to the platform used to analyze and manage service stations. Each service station is configured with a multidimensional fault management platform, and the multidimensional fault management platform supports web and mobile application access, ensuring that maintenance personnel can perform fault troubleshooting and handling anytime and anywhere.

[0183] In this example, the fault characteristics represent the features exhibited at the corresponding fault location when a service station experiences a fault;

[0184] In this example, the fault information represents the descriptive information about the fault contained in the service station;

[0185] In this example, deep mining refers to the process of locating the location of existing faults in a service station and uncovering potential faults.

[0186] In this example, the multi-dimensional fault management platform can be used to display the remediation methods for each service station.

[0187] The working principle and beneficial effects of the above technical solution are as follows: To improve the efficiency of fault analysis and management, the architecture is developed using the easy integration features of the Spring Cloud technology stack. A multi-dimensional fault management platform is established for each service station. Within this platform, manuals are edited to create a fault identification list for each service station. During service station operation, real-time data generated by each station is collected and used to construct fault characteristics for each station. Corresponding faults are then identified in the fault identification list, determining the fault information for each service station. Further in-depth analysis of the fault information is conducted within the multi-dimensional fault management platform, and finally, fault mitigation methods are established for each service station. This technical approach enables efficient guidance and management of power operation and maintenance faults, effectively addressing problems in fault handling, drawing management, data analysis, and operation and maintenance efficiency. It provides strong support for intelligent operation and maintenance transformation and can intuitively and conveniently display fault trends across various dimensions to maintenance management personnel, accurately guiding maintenance operations, proactively planning for future issues, and improving the availability of service stations.

[0188] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent fault analysis and management system based on big data, characterized in that, include: The multi-dimensional management module is used to develop the architecture using the Spring Cloud technology stack and generate a multi-dimensional fault management platform for each service station. The data management module is used to edit and create the manual library of the service station using the multi-dimensional fault management platform, and generate a fault identification list corresponding to each service station. The analysis and implementation module is used to determine the fault characteristics of the corresponding service station based on the real-time data of each service station, identify the corresponding fault characteristics in the fault identification list, and determine the fault information corresponding to each service station. The fault management module is used to perform in-depth mining of the fault information in the multi-dimensional fault management platform, determine the corresponding compensation method for each service station, and display it. The multidimensional management module includes: The microservice architecture unit is used to obtain the site location information and working attribute information corresponding to each service station, use the site location information and working attribute information to create the initial analysis management platform, and use the Spring Cloud technology stack to divide the initial analysis management platform into several microservices; The functional analysis unit is used to trace the source of each microservice, determine the service object corresponding to each microservice, collect the functional attribute information corresponding to each service object, analyze the topological relationship between different functional attribute information according to the distribution of the microservice in the initial analysis management platform, and generate a functional topology network. The functional deployment unit is used to generate functional annotations for corresponding service objects based on the node data corresponding to each topology node in the functional topology network, determine the docking function corresponding to each service object in combination with the network logic of the functional topology network, and deploy the corresponding microservices using the docking function to obtain the available microservices corresponding to each service station. The platform construction unit is used to identify high-frequency faults corresponding to each service station based on the site location information and working attribute information of each service station, enhance the functionality of the available microservices using the high-frequency faults, and establish a multi-dimensional fault management platform for each service station in combination with preset platform construction technology.

2. The intelligent fault analysis and management system based on big data as described in claim 1, characterized in that, The multi-dimensional fault management platform includes: The manual management sub-platform is used to classify the corresponding manual library according to the prescribed document structure using semantic analysis, and generate manual documents for each of the service stations. The instruction response sub-platform is used to locate and display the corresponding target manual in the manual document based on the search instructions issued by the operation and maintenance personnel; The offline buffering sub-platform is used to buffer the aforementioned manual documents offline.

3. The intelligent fault analysis and management system based on big data as described in claim 1, characterized in that, The platform construction unit includes: The fault identification subunit is used to draw a service station functional structure diagram based on the initial analysis management platform, determine the dynamic and static working information of the corresponding service station according to the site location information and working attribute information of each service station, and sequentially delete each dynamic and static working information in the service station functional structure diagram to obtain the dynamic fault threshold and static fault threshold corresponding to each service station. The fault mining subunit is used to generate several dynamic and static faults corresponding to each service station by combining each available microservice with the corresponding dynamic fault threshold and static fault threshold, retrieve historical faults of similar service stations in the big data center based on the working attribute information of the service station, and determine several high-frequency faults of each service station based on the first similarity between each dynamic fault and the historical fault and the second similarity between each static fault and the historical fault. The service enhancement subunit is used to obtain the fault level and fault performance corresponding to each of the high-frequency faults, establish mirror service information using the fault level and fault performance, enhance the corresponding available microservices using the mirror service information, and obtain the preferred microservices corresponding to each service station. The platform construction subunit is used to build the platform structure based on the interaction between the front-end framework, back-end technology and database corresponding to the initial analysis management platform. Each of the preferred microservices is input into the platform structure to generate the preferred management platform. The preferred management platform is used to find the sub-platform corresponding to each service station and perform functional repair to obtain the multi-dimensional fault management platform corresponding to each service station.

4. The intelligent fault analysis and management system based on big data as described in claim 1, characterized in that, The data management module includes: The data collection unit is used to search for the operation and maintenance data corresponding to each similar service station in the big data center, establish several operation and maintenance data information corresponding to the service station based on the operation and maintenance data, and determine the data description features corresponding to each operation and maintenance data information based on semantic analysis technology. The data filtering unit is used to locate contradictory operation and maintenance data information with opposite data description characteristics, upload the contradictory operation and maintenance information to the big data service center for information search, obtain the objective support quantity corresponding to each contradictory operation and maintenance data information, eliminate unqualified operation and maintenance data information based on the objective support quantity, obtain several qualified operation and maintenance data information corresponding to each service station, and build a manual library corresponding to each service station respectively. The fault identification unit is used to generate several fault descriptions for corresponding service stations in the multi-dimensional fault management platform, search for several matching information in the manual library according to the keywords corresponding to each fault description, and determine the response information of the corresponding service station to each fault description based on the matching information, thus forming a fault identification list.

5. The intelligent fault analysis and management system based on big data as described in claim 1, characterized in that, The analysis implementation module includes: The data processing unit is used to collect real-time data generated by each of the service stations, obtain the data period corresponding to each of the real-time data, determine the truncation distance corresponding to each of the real-time data according to the data period, use the truncation distance to cut the corresponding real-time data into several single data segments, and perform data sampling in each of the single data segments corresponding to the same real-time data segment to obtain several sampled data. The feature generation unit is used to obtain the data similarity between different sampled data, determine the correlation features between different single data segments, perform clustering processing on the sub-data contained in each single data segment to obtain the outlier corresponding to each single data segment, analyze the correlation outlier corresponding to each outlier based on the correlation features, and establish the fault features of the corresponding service station based on the distribution information of the correlation outlier in the real-time data. The fault identification unit is used to search for matching information corresponding to each fault feature in the fault identification list, establish a fault knowledge graph for each service station based on the matching information corresponding to each service station, and generate several fault information of the service station based on the fault level corresponding to each fault feature and the fault knowledge graph.

6. The intelligent fault analysis and management system based on big data as described in claim 1, characterized in that, The fault management module includes: The fault decomposition unit is used to perform initial identification of each fault information in the multi-dimensional fault management platform, obtain the dimension coverage information corresponding to each fault information, and perform multi-scale coarse-grained processing on each fault information based on the dimension coverage information to obtain several single-dimensional fault information corresponding to each service station. The fault mining unit is used to establish a fault damage model corresponding to the service station based on the fault dimension corresponding to each single-dimensional fault information, and input the single-dimensional fault information corresponding to the same service station into the fault damage model for deep training to obtain the explicit fault and implicit fault corresponding to each service station. The fault compensation unit is used to trace the source of each explicit fault in the fault damage model, obtain the first fault device corresponding to each explicit fault, enhance each implicit fault, run the fault damage model to determine the second fault device corresponding to each enhanced implicit fault, obtain the manual information corresponding to each first fault device and the second fault device from the manual library, establish the compensation method corresponding to each fault information and display it.

7. The intelligent fault analysis and management system based on big data as described in claim 6, characterized in that, The fault compensation unit includes: The first fault tracing subunit is used to run the fault damage model to obtain the first fault device corresponding to each of the explicit faults, obtain the fault level classification corresponding to each of the implicit faults, determine the maximum fault level classification corresponding to each of the implicit faults, and enhance the implicit faults to the corresponding maximum fault level classification to obtain the enhanced implicit faults. The second fault tracing subunit is used to treat the enhanced latent fault as the model running result, adjust the model running data in the fault damage model using the model running result, and determine the second fault device corresponding to the latent fault based on the adjustment direction corresponding to the model running data. The fault compensation execution subunit is used to count a number of first faulty devices and second faulty devices corresponding to each service station, filter the manual information corresponding to each first faulty device and second faulty device in the corresponding manual library, determine the maintenance method corresponding to each faulty device based on the manual information, merge the maintenance methods to obtain the compensation method corresponding to each service station and display it.

8. The intelligent fault analysis and management system based on big data as described in claim 7, characterized in that, Also includes: The fusion execution subunit is used to match priority weights for each of the first faulty devices and to match non-priority weights for each of the second faulty devices. Based on the priority weight and non-priority weight, a fusion threshold corresponding to the maintenance method is determined, and the maintenance methods are fused to generate a compensation method based on the fusion threshold.

9. A method for intelligent fault analysis and management based on big data, characterized in that, include: Step 1: Develop the architecture using the Spring Cloud technology stack to generate a multi-dimensional fault management platform for each service station; Step 2: Use the multi-dimensional fault management platform to edit and create the manual library of the service station, and generate a fault identification list for each service station; Step 3: Determine the fault characteristics of each service station based on the real-time data of each service station, identify the corresponding fault characteristics in the fault identification list, and determine the fault information corresponding to each service station. Step 4: In the multi-dimensional fault management platform, perform in-depth mining of the fault information, determine the corresponding compensation method for each service station, and display it; The process of developing the architecture using the Spring Cloud technology stack to generate a multi-dimensional fault management platform for each service station includes: Obtain the site location information and work attribute information corresponding to each service station, create an initial analysis management platform using the site location information and work attribute information, and divide the initial analysis management platform into several microservices using the Spring Cloud technology stack; Each microservice is traced to determine the service object corresponding to each microservice, and the functional attribute information corresponding to each service object is collected. Based on the distribution of the microservices in the initial analysis and management platform, the topological relationship between different functional attribute information is analyzed to generate a functional topology network. Based on the node data corresponding to each topology node in the functional topology network, functional annotations for corresponding service objects are generated. Combined with the network logic of the functional topology network, the docking function corresponding to each service object is determined. The docking function is used to deploy the corresponding microservices to obtain the available microservices corresponding to each service station. Based on the site location information and working attribute information of each service station, high-frequency faults corresponding to each service station are identified. The high-frequency faults are used to enhance the functionality of the available microservices. A multi-dimensional fault management platform is established for each service station in combination with the preset platform construction technology.

Images