Equipment Intelligent Operation and Maintenance Management Method Based on Fault Knowledge Base

Abstract

The present invention relates to the technical field of equipment operation and maintenance management, and discloses an intelligent equipment operation and maintenance management method based on a fault knowledge base. By collecting equipment data, storing the equipment data in a data warehouse, and preprocessing the equipment data, centralized management and easy access to the equipment data are ensured; a fault knowledge base is constructed based on historical equipment data to help understand different types of faults and their causes, providing support for the diagnosis of fault states; a status monitoring model is constructed based on historical equipment data and real-time equipment data to identify the fault states of the equipment, helping to timely discover potential problems and prevent the occurrence of fault states; if the equipment shows a fault state, based on the fault knowledge base and real-time equipment data, the fault location is diagnosed and maintenance measures are proposed, reducing the maintenance time and improving the reliability of the equipment; continuously monitor the fault state after equipment maintenance, and optimize the fault knowledge base and maintenance measures.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment operation and maintenance management, and particularly to an intelligent operation and maintenance management method for equipment based on a fault knowledge base. Background Art

[0002] In the prior art, machine learning and big data analysis are used to predict equipment failures, realize automatic diagnosis and prediction of equipment failures, realize real-time monitoring and adjustment of equipment status, construct a fault knowledge base including fault modes, solutions, etc., provide real-time problem-solving support for operation and maintenance personnel, realize the intelligence of operation and maintenance decisions based on big data analysis, and improve equipment reliability and efficiency. However, it does not solve the problem of how to identify the fault status and fault location of equipment by collecting equipment data to construct a fault knowledge base and a status monitoring model, and monitor whether there is a fault status after repairing the fault location.

[0003] For example, the patent with the publication number CN117474525A discloses a collaborative operation and maintenance method based on the full life cycle of shield TBM equipment. This method fuses the data of the full life cycle of equipment design, manufacturing, production, operation, and maintenance, and performs data mining and analysis through algorithms such as statistical analysis, classification analysis, correlation analysis, and machine learning to provide comprehensive support for the operation and maintenance of shield TBM engineering projects. This method can realize the comprehensive intelligent control of shield TBM engineering tunnels, geology, equipment, business information, etc., provide operation decisions such as operation optimization, attitude control, abnormal warning, and parameter warning during the tunneling process of shield TBM; effectively guarantee the tunneling efficiency of the equipment through the evaluation and analysis of equipment tunneling efficiency, key parameters, tunneling processes, etc.; realize the management and guidance of operation and maintenance, shutdown maintenance, overhaul maintenance, etc. through collaborative maintenance to accelerate the maintenance efficiency, ensure the normal operation of the equipment, and improve the risk prevention and control ability.

[0004] For example, the Chinese patent with the authorization announcement number CN113254541B discloses a theater operation and maintenance management method based on cluster analysis, including: setting attention with equipment view volume X and time period T as the main components; performing attention time clustering according to the same time period on different days to obtain N data clusters of different equipment; setting maintainability with equipment repair volume Y and time period T as the main components; performing repair time clustering according to the same time period on different days to obtain M data clusters of different equipment; performing comparative analysis on the N data clusters of attention time clustering and the M data clusters of repair time clustering; realizing targeted operation and maintenance management of equipment according to the analysis results. The above technical solution uses the behavior and maintenance record data of relevant equipment in the browsing platform of operation and maintenance management personnel as a data warehouse, and realizes the effective management of theater performing arts equipment by operation and maintenance management personnel through combining the clustering grouping results with historical judgments, optimizing operation and maintenance management data, simplifying operation and maintenance management behaviors, and realizing efficient and accurate operation and maintenance.

[0005] The above patents have the problems raised in this background art: The collaborative operation and maintenance method for the entire life cycle of the above equipment fuses the data of the entire life cycle of equipment design, manufacturing, production, operation, and maintenance, and conducts data mining and analysis through algorithms such as statistical analysis, classification analysis, correlation analysis, and machine learning for the comprehensive intelligent control of shield TBM engineering tunnels, geology, equipment, business information, etc.; The above theater operation and maintenance management method sets the attention degree with the equipment view volume X and time period T as the main components, sets the maintenance degree with the equipment repair volume Y and time period T as the main components, conducts attention time clustering and maintenance time clustering according to the same time period of different days, obtains N data clusters and M data clusters of different equipment, and conducts comparative analysis on the N data clusters of attention time clustering and the M data clusters of maintenance time clustering to achieve targeted operation and maintenance management of the equipment. The above patents do not solve the problem of how to construct a fault knowledge base and a status monitoring model by collecting equipment data to identify the fault status and fault location of the equipment, and monitor whether there is a fault status after repairing the fault location. To solve this problem, the present invention proposes an intelligent operation and maintenance management method for equipment based on a fault knowledge base. Summary of the Invention

[0006] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the specification of this application, to avoid obscuring the purpose of this part, the abstract of the specification, and the title of the invention, and such simplifications or omissions shall not be used to limit the scope of the present invention.

[0007] In view of the problems existing in the above-mentioned existing intelligent operation and maintenance management method for equipment based on a fault knowledge base, the present invention is proposed.

[0008] Therefore, the purpose of the present invention is to provide an intelligent operation and maintenance management method for equipment based on a fault knowledge base.

[0009] To solve the above technical problems, the present invention provides the following technical solutions:

[0010] An intelligent operation and maintenance management method for equipment based on a fault knowledge base, comprising:

[0011] Collect equipment data, store the equipment data in a data warehouse, and preprocess the equipment data;

[0012] Construct a fault knowledge base based on historical equipment data;

[0013] Construct a status monitoring model based on historical equipment data and real-time equipment data to identify the fault status of the equipment;

[0014] If the equipment has a fault status, diagnose the fault location based on the fault knowledge base and real-time equipment data, and propose repair measures;

[0015] Continuously monitor the fault status of the equipment after maintenance, and optimize the fault knowledge base and maintenance measures.

[0016] As a preferred solution of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention, wherein: the construction strategy of the fault knowledge base includes:

[0017] Determine the entities and relationships of the fault knowledge base, populate the entities of the fault knowledge base based on historical equipment data, determine the relationships between the entities, configure the relationship threshold, and if the degree of relationship between the entities is less than the relationship threshold, replace the relationship between the entities;

[0018] If the degree of relationship between the entities is greater than or equal to the relationship threshold, the relationship between the entities remains unchanged.

[0019] As a preferred solution of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention, wherein: the state monitoring model includes a normal recognition strategy and a fault detection strategy. The normal recognition strategy includes learning the normal state of the equipment through the normal recognition formula, and the calculation of the normal recognition formula is as follows:

[0020] y i = g(w2 × (f(w1 × x i + b1)) + b2);

[0021] In the formula, y i represents the recognition result of the i-th historical equipment data output, g(·) represents the activation function from the hidden layer to the output layer, w2 represents the weight coefficient from the hidden layer to the output layer, f(·) represents the activation function from the input layer to the hidden layer, w1 represents the weight coefficient from the input layer to the hidden layer, x i represents the i-th historical equipment data input, b1 represents the bias coefficient from the input layer to the hidden layer, and b2 represents the bias coefficient from the hidden layer to the output layer;

[0022] Judge the accuracy of the recognition result output by the normal recognition formula through the loss function, and the function expression of the loss function is as follows:

[0023]

[0024] In the formula, L represents the loss value of the recognition result, n represents the total number of historical equipment data, and ||·|| represents the norm calculation.

[0025] As a preferred solution of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention, wherein: the fault detection strategy includes calculating the covariance matrix and mean vector of the historical equipment data, and calculating the distance between the real-time equipment data and the normal state data in the historical equipment data. The distance formula between the real-time equipment data and the normal state data in the historical equipment data is as follows:

[0026]

[0027] Wherein, d m represents the distance between the m-th real-time equipment data and the normal state data in the historical equipment data, represents the m-th real-time equipment data, μ represents the mean value of the normal state data in the historical equipment data, and T represents the transpose of , and ∑ represents the covariance matrix of the normal state data in the historical equipment data;

[0028] Configure a distance threshold. If the distance between the real-time equipment data and the normal state data in the historical equipment data is less than or equal to the distance threshold, the real-time equipment data is in a normal state;

[0029] If the distance between the real-time equipment data and the normal state data in the historical equipment data is greater than the distance threshold, the real-time equipment data is in a fault state.

[0030] As a preferred solution of the equipment intelligent operation and maintenance management method based on the fault knowledge base according to the present invention, wherein: if the equipment is in a fault state, diagnose the fault location through a location diagnosis strategy, and the location diagnosis strategy includes:

[0031] By comparing the real-time equipment data with the fault knowledge base, identify the fault type of the real-time equipment data in the fault state. By matching with the fault types in the fault knowledge base, if the fault type of the real-time equipment data in the fault state matches the fault types in the fault knowledge base successfully, then locate the fault cause and equipment components of the real-time equipment data in the fault state according to the fault knowledge base;

[0032] If the fault type of the real-time equipment data in the fault state does not match the fault types in the fault knowledge base, then troubleshoot the fault cause of the real-time equipment data in the fault state and update the fault knowledge base.

[0033] As a preferred solution of the equipment intelligent operation and maintenance management method based on the fault knowledge base according to the present invention, wherein: diagnose the fault location based on the fault knowledge base and the real-time equipment data, repair the fault location according to the repair measures, continuously monitor the fault state of the equipment after repair. If there is no fault state after the equipment is repaired, record the repair process in the fault knowledge base;

[0034] If there is a fault state after the equipment is repaired, update the fault knowledge base and repair measures according to the location diagnosis strategy after determining the fault location.

[0035] As a preferred solution of the equipment intelligent operation and maintenance management method based on the fault knowledge base according to the present invention, wherein: the equipment data includes equipment status, operating environment, and historical fault records;

[0036] The collected equipment data is stored in the data warehouse through a storage policy, and the storage policy includes:

[0037] Judge the update method of the equipment data, configure the frequency threshold. If the update frequency of the equipment data is less than or equal to the frequency threshold, the update method of the equipment data is low-frequency update;

[0038] If the update frequency of the equipment data is greater than the frequency threshold, the update method of the equipment data is high-frequency update;

[0039] Judge the storage method of the equipment data, configure the stock threshold. If the storage volume of the equipment data is less than or equal to the stock threshold, the storage method of the equipment data is low-frequency storage;

[0040] If the storage volume of the equipment data is greater than the stock threshold, the storage method of the equipment data is high-frequency storage;

[0041] Comprehensively judge the loading method of the equipment data by combining the update method and the storage method of the equipment data. If the update method of the equipment data is low-frequency update and the storage method of the equipment data is high-frequency storage, the loading method of the equipment data is full-volume loading;

[0042] If the update method of the equipment data is high-frequency update and the storage method of the equipment data is low-frequency storage, the loading method of the equipment data is incremental loading;

[0043] Dynamically adjust the resource allocation of the data warehouse according to the loading method of the equipment data, configure the resource volume, and the resource volume includes the baseline resource and the extreme value resource. If the loading method of the equipment data is full-volume loading, allocate the extreme value resource;

[0044] If the loading method of the equipment data is incremental loading, allocate the baseline resource.

[0045] A computer device includes a memory for storing instructions; a processor for executing the instructions, so that the device executes an equipment intelligent operation and maintenance management method based on a fault knowledge base.

[0046] A computer-readable storage medium stores a computer program, and when the computer program is executed, an equipment intelligent operation and maintenance management method based on a fault knowledge base is implemented.

[0047] Advantages of the present invention: By collecting equipment data, storing the equipment data in a data warehouse, and preprocessing the equipment data, the present invention ensures the centralized management and easy access of the equipment data; constructs a fault knowledge base based on historical equipment data to help understand different types of faults and their causes, and provides support for the diagnosis of fault states; constructs a status monitoring model based on historical equipment data and real-time equipment data to identify the fault states of the equipment, which helps to timely discover potential problems and prevent the occurrence of fault states; if the equipment appears in a fault state, based on the fault knowledge base and real-time equipment data, diagnoses the fault location and proposes maintenance measures, reducing the maintenance time and improving the equipment reliability; continuously monitors the fault state after equipment maintenance, and optimizes the fault knowledge base and maintenance measures. Through accurate fault diagnosis and maintenance measures, it is possible to better allocate maintenance resources, avoid unnecessary inspections and maintenance, and save costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. Among them:

[0049] Figure 1 is the method flow chart of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention;

[0050] Figure 2 is the storage strategy flow chart of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention;

[0051] Figure 3 is the construction strategy flow chart of the fault knowledge base of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention;

[0052] Figure 4 is the location diagnosis strategy flow chart of the equipment intelligent operation and maintenance management method based on the fault knowledge base of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] In order to make the above objects, features and advantages of the present invention more obvious and understandable, the following will give a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings of the specification.

[0054] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0055] Second, the "one embodiment" or "embodiment" mentioned herein refers to specific features, structures, or characteristics that may be included in at least one implementation manner of the present invention. The phrase "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments.

[0056] Embodiment 1

[0057] In this embodiment, a method flowchart of an equipment intelligent operation and maintenance management method based on a fault knowledge base is provided, as Figure 1 shown. The equipment intelligent operation and maintenance management method based on a fault knowledge base includes:

[0058] S1. Collect equipment data, store the equipment data in a data warehouse, and preprocess the equipment data.

[0059] The equipment data includes equipment status, operating environment, and historical fault records.

[0060] Store the collected equipment data in the data warehouse through a storage policy. The storage policy is as Figure 2 shown and specifically includes:

[0061] Judge the update method of the equipment data, configure a frequency threshold. If the update frequency of the equipment data is less than or equal to the frequency threshold, the update method of the equipment data is low-frequency update;

[0062] If the update frequency of the equipment data is greater than the frequency threshold, the update method of the equipment data is high-frequency update.

[0063] Judge the storage method of the equipment data, configure a stock threshold. If the storage amount of the equipment data is less than or equal to the stock threshold, the storage method of the equipment data is low-frequency storage;

[0064] If the storage amount of the equipment data is greater than the stock threshold, the storage method of the equipment data is high-frequency storage.

[0065] Comprehensively judge the loading method of the equipment data in combination with the update method and storage method of the equipment data. If the update method of the equipment data is low-frequency update and the storage method of the equipment data is high-frequency storage, the loading method of the equipment data is full-volume loading;

[0066] If the update method of the equipment data is high-frequency update and the storage method of the equipment data is low-frequency storage, the loading method of the equipment data is incremental loading;

[0067] Dynamically adjust the resource allocation of the data warehouse according to the loading method of the equipment data, configure the resource amount. The resource amount includes a benchmark resource and an extreme value resource. If the loading method of the equipment data is full-volume loading, allocate the extreme value resource;

[0068] If the loading method of equipment data is incremental loading, allocate benchmark resources.

[0069] Among them, full - volume loading means transmitting all equipment data, and incremental loading means transmitting only the equipment data that has changed since the last loading. Here, it can be judged that the loading methods of equipment status and operating environment in equipment data are incremental loading, while the loading method of historical fault records in equipment data is full - volume loading. Extreme resources refer to a large demand for data warehouse resources when all equipment data needs to be transmitted, and benchmark resources refer to only requiring less data warehouse resources when only transmitting the equipment data that has changed since the last loading. At the same time, when the loading method of equipment data is full - volume loading, the storage performance requirements for the data warehouse are relatively low, while when the loading method of equipment data is incremental loading, the storage performance requirements for the data warehouse are relatively high. Classifying equipment data according to the loading method of equipment data can ensure that the response time when accessing equipment data in the data warehouse can meet business requirements, and at the same time can optimize the cost of storing equipment data.

[0070] Check the errors and inconsistencies of the equipment data stored in the data warehouse and correct them, including deleting duplicate data, filling in missing values, and handling outliers. Through standardization processing, ensure the consistency of equipment data at different scales. Among them, equipment status includes equipment parameters and fault displays, operating environment includes temperature, humidity, and air quality during equipment operation, and these environmental factors may affect the performance and failure rate of the equipment. Historical fault records include all historical fault events, including the time, type, cause, impact, and repair measures of the fault.

[0071] S2. Build a fault knowledge base based on historical equipment data.

[0072] The construction strategy of the fault knowledge base is as Figure 3 shown, specifically including:

[0073] Determine the entities and relationships of the fault knowledge base. The entities of the fault knowledge base include fault types, fault causes, and equipment components, etc. The relationships between the entities of the fault knowledge base are configured with influences, belong to, and cause, etc. Fill the entities of the fault knowledge base based on historical equipment data, and determine the relationships between the entities, and configure the relationship threshold. If the relationship degree between the entities is less than the relationship threshold, replace the relationship between the entities;

[0074] If the degree of relationship between entities is greater than or equal to the relationship threshold, the relationship between the entities remains unchanged. For example, there are many failure causes that can lead to faults in pump components. Each failure cause has a different degree of causing faults in pump components. The decrease or increase in the flow rate of the pump will lead to a decrease or increase in the operating efficiency of the pump. The increase or decrease in the working pressure of the pump may affect the performance and stability of the pump. The current fluctuation of the pump may indicate that the pump is being interfered with. The pump in a high-temperature environment may cause component expansion, a decrease in the performance of lubricating oil, and even damage to internal components of the pump. While the pump in a low-temperature environment may cause some components to become brittle or solidify, affecting the normal operation of the pump. In a high-humidity environment, it may cause internal components of the pump to rust and corrode, reducing the life and performance of the pump. In a low-humidity environment, it may cause the sealing components of the pump to crack and harden, increasing the risk of leakage of the pump. Here, it is necessary to first determine the degree of causing faults in pump components, and then determine whether this failure cause leads to the occurrence of this failure type or affects the occurrence of this failure type, so as to determine the relationship between entities. Here, the maintenance measures for different failure types are sorted and analyzed by referring to the maintenance methods of domain experts for different faults.

[0075] S3. Construct a condition monitoring model based on historical equipment data and real-time equipment data to identify the fault status of the equipment.

[0076] The condition monitoring model includes a normal recognition strategy and a fault detection strategy. The normal recognition strategy includes learning the normal state of the equipment through a normal recognition formula. The calculation of the normal recognition formula is as follows:

[0077] y i = g(w2 × (f(w1 × x i + b1)) + b2);

[0078] In the formula, y i represents the recognition result of the i-th historical equipment data output. g(·) represents the activation function from the hidden layer to the output layer. w2 represents the weight coefficient from the hidden layer to the output layer. f(·) represents the activation function from the input layer to the hidden layer. w1 represents the weight coefficient from the input layer to the hidden layer. x i represents the i-th historical equipment data input. b1 represents the bias coefficient from the input layer to the hidden layer. b2 represents the bias coefficient from the hidden layer to the output layer.

[0079] Judge the accuracy of the recognition result output by the normal recognition formula through the loss function. The function expression of the loss function is as follows:

[0080]

[0081] In the formula, L represents the loss value of the recognition result. n represents the total number of historical equipment data. ||·|| represents the norm calculation.

[0082] It should be explained that the normal recognition strategy is to learn the normal state of the equipment through historical equipment data, while the fault detection strategy is to detect faults in real-time equipment data and identify the fault state of the equipment.

[0083] The fault detection strategy includes calculating the covariance matrix and mean vector of historical equipment data, and calculating the distance between real-time equipment data and the normal state data in historical equipment data. The distance formula between real-time equipment data and the normal state data in historical equipment data is as follows:

[0084]

[0085] In the formula, d m represents the distance between the m-th real-time equipment data and the normal state data in historical equipment data. represents the m-th real-time equipment data, μ represents the mean of the normal state data in historical equipment data, T represents the transpose of and ∑ represents the covariance matrix of the normal state data in historical equipment data.

[0086] Configure a distance threshold. If the distance between real-time equipment data and the normal state data in historical equipment data is less than or equal to the distance threshold, the real-time equipment data is in a normal state;

[0087] If the distance between real-time equipment data and the normal state data in historical equipment data is greater than the distance threshold, the real-time equipment data is in a fault state.

[0088] Combining neural networks with anomaly detection algorithms can effectively build a state monitoring model to identify the fault state of the pump and achieve accurate fault state recognition in actual application scenarios.

[0089] S4. If the equipment is in a fault state, based on the fault knowledge base and real-time equipment data, diagnose the fault location and propose repair measures.

[0090] If the equipment is in a fault state, diagnose the fault location through the location diagnosis strategy. The location diagnosis strategy is as Figure 4 shown and specifically includes:

[0091] By comparing the real-time equipment data with the fault knowledge base, identify the fault type of the real-time equipment data as a fault state. By matching with the fault types in the fault knowledge base, if the fault type of the real-time equipment data as a fault state matches successfully with the fault types in the fault knowledge base, then locate the fault cause and equipment components of the real-time equipment data as a fault state according to the fault knowledge base;

[0092] If the fault type of the real-time equipment data in the fault state does not match the fault type in the fault knowledge base, then troubleshoot the cause of the fault of the real-time equipment data in the fault state and update the fault knowledge base.

[0093] Take corresponding maintenance measures according to the fault type and cause. Before repairing the pump, immediate emergency measures such as shutdown or power-off should be taken according to the fault type and cause to avoid more serious losses caused by the expansion of the fault. At the same time, set necessary warning signs and isolation areas to prevent the risks brought by misoperation, ensure the safety of operators, and back up the key data before and after the fault occurs.

[0094] Decide whether to perform on-site repair or replace new equipment parts according to the damage degree and repairability of the pump components. For the fault type, update the preventive maintenance plan to reduce the probability of future faults. After completing the maintenance measures, conduct a function test to ensure that all pump components return to normal operation, and record the maintenance process and results in the fault knowledge base for reference in future fault state diagnosis and repair.

[0095] S5. Continuously monitor the fault state of the equipment after maintenance, and optimize the fault knowledge base and maintenance measures.

[0096] Diagnose the fault location based on the fault knowledge base and real-time equipment data, repair the fault location according to the maintenance measures, and continuously monitor the fault state of the equipment after maintenance. If there is no fault state after the equipment is repaired, record the maintenance process in the fault knowledge base;

[0097] If there is a fault state after the equipment is repaired, update the fault knowledge base and maintenance measures according to the location diagnosis strategy after determining the fault location.

[0098] Continuously collect pump fault cases, maintenance manuals and solutions to enrich the fault knowledge base, carry out regular skills training for maintenance personnel to improve their application ability of new technologies and new methods, and regularly review and optimize the content of the fault knowledge base to ensure that the fault knowledge base reflects the latest maintenance measures.

[0099] Embodiment 2

[0100] In this embodiment, a computer device is provided, including a memory and a processor. The memory is used to store instructions, and the processor is used to execute the instructions, so that the computer device executes the steps of implementing the above-mentioned equipment intelligent operation and maintenance management method based on the fault knowledge base.

[0101] Embodiment 3

[0102] In this embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed, the steps of implementing the above-mentioned equipment intelligent operation and maintenance management method based on the fault knowledge base are realized.

[0103] The computer-readable storage medium includes various media for storing program codes, such as USB flash drives, mobile hard disks, read-only memories, random access memories, magnetic disks, or optical discs.

[0104] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An intelligent equipment operation and maintenance management method based on a fault knowledge base, characterized in that: include: Collect equipment data, store the equipment data in the data warehouse, and pre-process the equipment data; Equipment data includes equipment status, operating environment and historical fault records. The collected equipment data is stored in the data warehouse through storage strategies, which include: Determine the update mode of equipment data and configure a frequency threshold. If the update frequency of equipment data is less than or equal to the frequency threshold, the update mode of equipment data is low-frequency update. If the update frequency of the equipment data is greater than the frequency threshold, the update mode of the equipment data is high-frequency update; Determine the storage mode of equipment data and configure the stock threshold. If the storage amount of equipment data is less than or equal to the stock threshold, the storage mode of equipment data is low-frequency storage. If the storage amount of equipment data is greater than the storage threshold, the storage method of equipment data is high-frequency storage; The equipment data loading method is comprehensively determined based on the equipment data update method and storage method. If the equipment data update method is low-frequency update and the equipment data storage method is high-frequency storage, the equipment data loading method is full loading. If the update mode of equipment data is high-frequency update and the storage mode of equipment data is low-frequency storage, the loading mode of equipment data is incremental loading; Dynamically adjust the resource allocation of the data warehouse according to the loading mode of equipment data, and configure the resource quantity, which includes baseline resources and extreme resources. If the loading mode of equipment data is full loading, extreme resources are allocated; If the loading mode of equipment data is incremental loading, the base resource is allocated; Build a fault knowledge base based on historical equipment data; A condition monitoring model is constructed based on historical equipment data and real-time equipment data to identify the fault status of the equipment; the condition monitoring model includes a normal identification strategy and a fault detection strategy, and the fault detection strategy includes calculating the covariance matrix and mean vector of the historical equipment data and calculating the distance between the real-time equipment data and the normal status data in the historical equipment data; The distance formula between real-time equipment data and normal status data in historical equipment data is as follows: Where, d m Indicates the distance between the mth real-time equipment data and the normal state data in the historical equipment data. represents the mth real-time equipment data, μ represents the mean of the normal state data in the historical equipment data, and T represents the The transpose of , ∑ represents the covariance matrix of normal state data in historical equipment data; If the equipment fails, diagnose the fault location and propose maintenance measures based on the fault knowledge base and real-time equipment data; Continuously monitor the fault status of equipment after maintenance, and optimize the fault knowledge base and maintenance measures.

2. The equipment intelligent operation and maintenance management method based on the fault knowledge base according to claim 1, characterized in that: The strategies for building the fault knowledge base include: Determine the entities and relationships of the fault knowledge base, fill the entities of the fault knowledge base based on historical equipment data, determine the relationships between the entities, configure the relationship threshold, and if the degree of relationship between the entities is less than the relationship threshold, replace the relationship between the entities; If the degree of relationship between entities is greater than or equal to the relationship threshold, the relationship between the entities remains unchanged.

3. The equipment intelligent operation and maintenance management method based on the fault knowledge base according to claim 2 is characterized in that: The normal recognition strategy includes learning the normal state of the equipment through the normal recognition formula. The calculation of the normal recognition formula is as follows: <h2 style=";text-align:left;direction:ltr">y<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> =g(w2×(f(w1×x<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> +b1))+b2); In the formula, y i represents the recognition result of the i-th historical equipment data output, g(·) represents the activation function from the hidden layer to the output layer, w2 represents the weight coefficient from the hidden layer to the output layer, f(·) represents the activation function from the input layer to the hidden layer, w1 represents the weight coefficient from the input layer to the hidden layer, x i represents the i-th historical equipment data input, b1 represents the bias coefficient from the input layer to the hidden layer, and b2 represents the bias coefficient from the hidden layer to the output layer; The accuracy of the recognition result output by the normal recognition formula is judged by the loss function. The function expression of the loss function is as follows: Where L represents the loss value of the recognition result, n represents the total number of historical equipment data, and ||·|| represents the norm calculation.

4. The equipment intelligent operation and maintenance management method based on the fault knowledge base according to claim 3 is characterized in that: Configure a distance threshold. If the distance between the real-time equipment data and the normal state data in the historical equipment data is less than or equal to the distance threshold, the real-time equipment data is in a normal state. If the distance between the real-time equipment data and the normal state data in the historical equipment data is greater than the distance threshold, the real-time equipment data is in a fault state.

5. The equipment intelligent operation and maintenance management method based on the fault knowledge base according to claim 4 is characterized in that: If the equipment fails, the fault location is diagnosed through the location diagnosis strategy, which includes: By comparing the real-time equipment data with the fault knowledge base, the fault type of the real-time equipment data in the fault state is identified, and by matching with the fault type in the fault knowledge base, if the fault type of the real-time equipment data in the fault state successfully matches the fault type in the fault knowledge base, the fault cause and equipment component of the real-time equipment data in the fault state are located according to the fault knowledge base; If the fault type of the real-time equipment data being in a fault state fails to match the fault type in the fault knowledge base, the fault cause of the real-time equipment data being in a fault state is checked and the fault knowledge base is updated.

6. The equipment intelligent operation and maintenance management method based on the fault knowledge base according to claim 5 is characterized in that: Diagnose the fault location based on the fault knowledge base and real-time equipment data, repair the fault location according to the maintenance measures, and continuously monitor the fault status of the equipment after maintenance. If the equipment no longer has a fault status after maintenance, the maintenance process will be recorded in the fault knowledge base; If the equipment is in a fault state after maintenance, the fault location is determined and the fault knowledge base and maintenance measures are updated according to the location diagnosis strategy.

7. A computer device, characterized in that: include, A memory for storing instructions; A processor is used to execute the instruction so that the computer device executes the equipment intelligent operation and maintenance management method based on the fault knowledge base as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the equipment intelligent operation and maintenance management method based on the fault knowledge base as described in any one of claims 1 to 6 is implemented.

Images