Maintenance decision system and method based on big data driving

Abstract

The invention belongs to the technical field of thermal power auxiliary equipment maintenance, and particularly relates to a maintenance decision system and method based on big data driving, and the system comprises an asset structure tree module, a fault knowledge base management module, an equipment defect management module, a maintenance operation standardization module, a grading push module, and a strategy reliability analysis module. The data acquisition module is used for capturing equipment main data, spare part resource data, human resource data and equipment operation data, storing equipment structure decomposition, fault modes, maintenance strategies and health indexes and providing reference basis for equipment digital modeling, and is used for acquiring contents in a maintenance process so that a user can browse and view the contents from a big data platform; the design module is used for performing standardized design on contents of the large and small repair file packages; the method comprises steps 1-6. According to the method, the reliability attenuation rule of the equipment is visually quantified, and a statistical basis is provided for dynamically adjusting maintenance decisions so as to deploy the maintenance decisions or priorities in a grading manner.

Description

Technical Field

[0001] The present invention belongs to the technical field of maintenance of thermal power auxiliary equipment, and in particular relates to a maintenance decision-making system and method driven by big data. Background Art

[0002] Thermal power auxiliary equipment includes water feed pumps, induced draft fans, forced draft fans, primary fans, air preheaters, dust removal equipment, soot blowers, condensers, condensate pumps, circulating water pumps, cooling towers, vacuum pumps, drain pumps, etc. Not only are there large numbers and many types, but the working environment temperature is high, which brings certain troubles to the maintenance staff.

[0003] These auxiliary equipment can occasionally malfunction due to high loads or changes in the operating environment, requiring either scheduled maintenance or emergency repairs. For example, China's invention, published under the patent number CN119624426A, describes an AI-based safety monitoring and early warning management method and system for thermal power plants. This system uses an AI clustering algorithm to analyze and extract features from equipment vibration data, better capturing the complex patterns and inherent structures within the data. This provides more accurate information for subsequent model training and fault prediction, thereby improving the accuracy and reliability of safety monitoring and early warning.

[0004] Although vibration data is analyzed and features are extracted through clustering algorithms, there is still a lack of a corresponding evaluation system for maintenance decision-making. To this end, the present invention proposes a maintenance decision-making system, method, computer equipment and storage medium based on big data, which can intuitively quantify the reliability attenuation law of equipment and provide a statistical basis for dynamically adjusting maintenance decisions. Summary of the Invention

[0005] The purpose of the present invention is to provide a maintenance decision-making system, method, computer equipment and storage medium driven by big data, which can intuitively quantify the reliability attenuation law of equipment, provide a statistical basis for dynamically adjusting maintenance decisions, and allocate maintenance decisions or priorities in a hierarchical manner.

[0006] The technical solutions adopted by the present invention are as follows:

[0007] A maintenance decision-making system driven by big data, including:

[0008] The asset structure tree module is used to capture equipment master data, spare parts resource data, human resources data, and equipment operation data;

[0009] The fault knowledge base management module stores device structure decomposition, failure modes, maintenance strategies, and health indicators, providing a reference for digital modeling of equipment;

[0010] The equipment defect management module is used to collect content during the maintenance process for users to browse and view on the big data platform;

[0011] The maintenance operation standardization module is used to standardize the content of the maintenance document package to guide the execution process of major and minor maintenance and comprehensively record the maintenance work results;

[0012] The hierarchical push module uses a hierarchical evaluation algorithm to analyze the captured maintenance process content and data captured by the asset structure tree module, and calls the fault knowledge base management module to determine the equipment maintenance priority or strategy;

[0013] The strategy reliability analysis module estimates the survival function and performs statistics to obtain a survival curve that reflects the typical trouble-free cycle of the equipment, guides the preventive maintenance time point, and is used to optimize the maintenance strategy.

[0014] As an optional solution, the hierarchical evaluation algorithm is specifically as follows:

[0015] First, during the maintenance operation, the failure consequences, failure probability and maintenance time are scored separately;

[0016] Then make a comprehensive risk score, recorded as R:

[0017] R=ω1·C+ω2·P+ω3·T r ①

[0018] Where C is the failure consequence score; P is the failure probability score; T r is the maintenance time score, ω1, ω2 and ω3 are weight coefficients, satisfying ω1+ω2+ω3=1;

[0019] At the same time, a hierarchical decision index is given, denoted as DI:

[0020] DI=R·M②

[0021] Among them, M is an adjustment factor used to consider other dynamic factors; and the maintenance strategy is divided according to the DI threshold.

[0022] As an optional solution, the survival function is as follows:

[0023] Assume that the survival probability S(t) represents the probability that the equipment has not failed before time t, and the calculation formula is:

[0024]

[0025] Among them, t i represents the time point when the i-th fault is observed, d i Indicates that at time t i The number of failed devices, n i Indicates that at time t i The number of devices that were previously at risk;

[0026] And, the key calculation steps:

[0027] S1. Data preparation:

[0028] The observation data of each device includes time and event status;

[0029] S2. Sorting and calculation:

[0030] Arrange the failure time of all devices in ascending order as t1<t2<…<t k ;

[0031] For each t i Calculate d i and n i ;

[0032] S3. Survival probability update:

[0033] Initialize S(0)=1;

[0034] For each t i , update the survival probability:

[0035]

[0036] Among them, deletion only affects n at subsequent time points i , do not participate in d i calculate;

[0037] And through mathematical supplementation, we find the median line of the maintenance strategy update as follows:

[0038] Variance estimation:

[0039]

[0040] The median survival time is the time t that satisfies S(t) = 0.5, which represents the average reliable life of the equipment. It can intuitively quantify the reliability attenuation law of the equipment and provide a statistical basis for dynamically adjusting the maintenance cycle.

[0041] As an optional solution, the asset structure tree module includes:

[0042] The equipment master data management unit interfaces with external systems to obtain equipment master data, build a complete asset digital model, and store and manage enterprise equipment master data information;

[0043] Spare parts resource data management unit, which centrally manages and maintains spare parts resources, supports querying spare parts resource information, and provides data support for maintenance cost estimation;

[0044] The human resources data management unit manages and maintains basic human resources data, supports querying human resources information, and provides data support for maintenance cost estimation;

[0045] The equipment operation data management unit collects physical signals from thermal power auxiliary equipment in real time through sensors, stores raw data and result data, and digitally models thermal power auxiliary equipment.

[0046] As an optional solution, the fault knowledge base management module includes:

[0047] Equipment structure decomposition library, providing structural decomposition references for corresponding equipment in the fault knowledge base;

[0048] Fault mode library, which provides references to the fault causes and fault modes of corresponding components in the fault knowledge base;

[0049] Maintenance strategy library, which provides references to the failure causes, maintenance measures, and maintenance cycles of corresponding components in the fault knowledge base;

[0050] The health indicator library provides a reference for the corresponding health indicators of the devices in the fault knowledge base.

[0051] As an optional solution, the equipment defect management module includes:

[0052] The defect data collection unit includes defect registration, defect confirmation, operation information, and defect elimination process content. Defect information is automatically collected, filled in, and improved through each page field and defect category to achieve standardization of all defect data.

[0053] Defect query unit, which allows users to browse and view all defects synchronized from the big data platform, including defect information completion status, number, defect description, unit to which it belongs, and defect category;

[0054] The defect information improvement unit can call the failure mode and effect analysis results or fault knowledge base data in the defect information improvement interface, improve the defects according to the defect description information, select the component information where the defect occurs, and associate the fault mode and fault information.

[0055] As an optional solution, the maintenance operation standardization module includes:

[0056] Maintenance time planning unit, which lists the maintenance work to be carried out, time plan and arrangement of related resources;

[0057] Maintenance work instruction unit, which provides maintenance work steps, operating procedures, and safety measures;

[0058] Maintenance record and report unit, which records key information during the maintenance process, including equipment status, maintenance details, work time and personnel, as a reference for subsequent maintenance;

[0059] The maintenance tools and spare parts unit lists the tools and equipment that need to be used, as well as the spare parts and consumables that need to be replaced or used later.

[0060] A maintenance decision-making method based on big data driving includes the following steps:

[0061] Step 1: Call the asset structure tree module to capture equipment master data, spare parts resource data, human resources data, and equipment operation data;

[0062] Step 2: Call the storage device structure decomposition, failure mode, maintenance strategy, and health indicators to digitally model the device and identify the data corresponding to step 1;

[0063] Step 3: Collect maintenance process content for users to browse and view on the big data platform, and update it to the equipment digital model in real time;

[0064] Step 4: Standardize the contents of the maintenance and overhaul document packages to guide the execution of the maintenance and overhaul process, and comprehensively record the maintenance results for users to retrieve and review from the equipment digital model;

[0065] Step 5: Using a hierarchical evaluation algorithm, the captured maintenance process content and data captured by the asset structure tree module are analyzed, and the fault knowledge base management module is called to determine the equipment maintenance priority or strategy;

[0066] Step 6: By estimating the survival function and performing statistics, a survival curve is obtained to reflect the typical trouble-free period of the equipment, guide the preventive maintenance time point, and optimize the maintenance strategy.

[0067] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the maintenance decision-making method when executing the computer program.

[0068] A computer-readable storage medium stores a computer program, which implements the steps of the maintenance decision-making method when executed by a processor.

[0069] The technical effects achieved by the present invention are:

[0070] In the maintenance of thermal power auxiliary equipment, the present invention captures real-time data, structures the asset structure tree to form a supporting fault knowledge base, and provides a standardized maintenance work plan. Based on big data-driven technical means, a hierarchical push standard and strategy reliability evaluation system is systematically constructed, which can intuitively quantify the reliability attenuation law of the equipment and provide a statistical basis for dynamically adjusting maintenance decisions.

[0071] The present invention combines the asset structure tree, fault knowledge base and equipment defects to form a relatively complete equipment digital model, which is convenient for users to browse and view, and displays the maintenance work plan at the corresponding node, updates data in real time, and completes the status of each device.

[0072] The hierarchical evaluation algorithm proposed in this invention provides a scientific and flexible maintenance strategy hierarchical evaluation theory by quantifying the consequences of failures, failure probabilities, and repair times, combined with weights and adjustment factors. In practical applications, parameters and weights can be adjusted according to specific needs to ensure the rationality and effectiveness of decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 is a system block diagram of a maintenance decision system in a first embodiment of the present invention;

[0074] Figure 2 This is a flowchart of the maintenance decision-making method in the second embodiment of the present invention. DETAILED DESCRIPTION

[0075] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is described in detail below with reference to the following examples. It should be understood that the following text is only used to describe one or more specific embodiments of the present invention and does not strictly limit the scope of protection of the present invention.

[0076] Example 1:

[0077] like Figure 1 As shown, a maintenance decision system based on big data driving includes:

[0078] (1) Asset structure tree

[0079] (1) Equipment master data management

[0080] Establish interfaces with external systems to obtain equipment master data, build a complete asset digital model, store and manage enterprise equipment master data, and support manual import of enterprise equipment master data. By sorting out the hierarchical relationships of enterprise assets, establish a tree structure of units, systems, and equipment, and build an asset structure tree. This helps enterprises clarify the relationships between units, systems, and equipment, and provides a data foundation for further reliability analysis.

[0081] The asset structure tree and asset master data can share a set of equipment codes (KKS codes) with the smart enterprise management and control platform, using the same coding system;

[0082] (2) Spare parts resource data management

[0083] By linking the spare parts management function within the smart enterprise management and control platform, unified management and maintenance of spare parts resources are achieved, including resource names, resource descriptions, resource codes, procurement price information, spare parts model information, etc.; keyword queries can be used to quickly query spare parts resource information, providing data support for maintenance cost estimation;

[0084] (3) Human Resources Data Management

[0085] Manage and maintain basic human resources data, including human resources name, resource description, number of internal staff, number of external staff, fixed costs, time costs, model information, and other dimensions. Quickly query human resources information through keyword queries, providing data support for maintenance cost estimation.

[0086] (4) Equipment operation data management

[0087] Multiple sensors are used to collect physical signals from thermal power auxiliary equipment in real time, such as vibration, temperature, ultrasound, sound, angle, and image. These signals are stored in both raw and resulting data using technologies such as time-series databases or object databases. Digital modeling of thermal power auxiliary equipment is performed (e.g., using laser scanning, point cloud reconstruction, VR, or cloud computing platforms) to facilitate on-duty personnel in the control room to understand the operating status of each auxiliary equipment.

[0088] (2) Fault knowledge base management

[0089] Leveraging enterprise asset data, we build our own fault knowledge base, which includes pre-built equipment structure decomposition libraries, fault mode libraries, and maintenance strategy libraries. This quickly establishes a data foundation for the company's reliability-centric equipment management system. This database can also be combined with past user fault and inspection and maintenance records as auxiliary input data for failure mode and effects analysis.

[0090] By accessing the industry knowledge base, the fault knowledge base provides users with a reference for equipment digital modeling and failure mode and effect analysis. Through continued use of the software, users can define their own private fault knowledge base and accumulate equipment maintenance experience. The fault knowledge base displays the device type, device name, device description, and device knowledge base content. The device fault knowledge base contains the following content:

[0091] Equipment structure decomposition library, providing structural decomposition references for corresponding equipment in the fault knowledge base;

[0092] Fault mode library, which provides references to the fault causes and fault modes of corresponding components in the fault knowledge base;

[0093] Maintenance strategy library, which provides references to the failure causes, maintenance measures, and maintenance cycles of corresponding components in the fault knowledge base;

[0094] Health indicator library, which provides references to the corresponding health indicators of devices in the fault knowledge base;

[0095] The fault knowledge base is filtered based on the knowledge base type, device type, and device keyword conditions, allowing users to select and view device details within the fault knowledge base. Once the target device is determined, the device knowledge base is called to view the device structure decomposition library, failure mode library, maintenance strategy library, and health indicator library. During the device structure decomposition phase and the early stages of the failure mode and impact analysis phase, the initial stage of device structure decomposition and failure modes is completed through the call and reference of the knowledge base.

[0096] (3) Equipment defect management

[0097] The main functions of equipment defect management are as follows:

[0098] Defect data collection includes content from the business process, including defect registration, defect confirmation, operation information, and defect elimination. Defect information completion automatically collects, fills in, and completes defect content through each page field and defect category, achieving standardization of all defect data.

[0099] Defect query allows you to browse and view all defects synchronized from the big data platform, including defect information completion status, number, defect description, unit to which it belongs, defect category, and other information;

[0100] Defect information improvement: In the defect information improvement interface, you can call the failure mode and effect analysis results or fault knowledge base data, improve the defect according to the defect description information, select the component information where the defect occurs, and associate the failure mode and fault information;

[0101] (4) Standardization of maintenance operations

[0102] Maintenance operation standardization is used to standardize the content of maintenance document packages in the power industry, guide the execution process of maintenance, and comprehensively record maintenance results. This system will establish standard maintenance operation standards for different types of equipment. After the maintenance project is established, the standard maintenance operation standardization can be used as the basis for planning the maintenance work. Users can adjust the standard maintenance operation standardization according to the situation. Maintenance operation standardization mainly includes the following contents:

[0103] (4) Maintenance time plan

[0104] List in detail the maintenance work to be carried out, the time plan and the arrangement of related resources, which should include a list of equipment, specific maintenance steps and requirements, and the division of labor and responsibilities of personnel;

[0105] (5) Maintenance work instructions

[0106] Provide specific guidance and requirements, including maintenance work steps, operating procedures, safety measures, etc. These instructions should describe the operating requirements of each work step in detail to ensure the correctness and safety of maintenance work;

[0107] (6) Maintenance records and reports

[0108] Record key information during the maintenance process, including equipment status, maintenance details, work time, and personnel. These records can be used to monitor equipment maintenance and track problems, and serve as a reference for future maintenance.

[0109] (7) Maintenance tools and spare parts

[0110] List the tools and equipment that will be used, as well as a list of spare parts and consumables that may need to be replaced or used;

[0111] Maintenance work standardization supports APP (mobile application), which can push relevant maintenance work content and procedures to mobile terminals to guide maintenance personnel in on-site operations. After the maintenance work is completed, the relevant maintenance information and inspection results can be directly uploaded and archived through the mobile terminal. The system's structured design can convert some content into reliability data, which serves as the data foundation for fault pattern identification and strategy optimization, further enhancing the later value of maintenance data.

[0112] (5) Tiered push

[0113] During the maintenance process, the key information recorded is used to provide a hierarchical evaluation algorithm to determine the equipment maintenance priority or strategy (such as preventive maintenance, corrective maintenance, condition monitoring, etc.), as follows:

[0114] (I) Parameter scoring

[0115] Consequences of failure:

[0116] 1 to 3 points: low consequences (such as minor downtime, low repair costs);

[0117] 4 to 6 points: moderate consequences (e.g., partial production interruption, moderate repair costs);

[0118] 7 to 10 points: high consequences (such as safety risks, significant economic losses);

[0119] Failure probability:

[0120] 1 to 3 points: low probability (e.g., new equipment, stable operation);

[0121] 4 to 6 points: medium probability (e.g., aging equipment, poor operating environment);

[0122] 7 to 10 points: high probability (e.g., frequent failures, poor historical record);

[0123] Maintenance time:

[0124] 1 to 3 points: short time (such as simple replacement of spare parts);

[0125] 4 to 6 points: medium time (such as partial disassembly and debugging);

[0126] 7-10 points: long time (e.g., major repairs, need for external support);

[0127] (II) Comprehensive risk score, denoted as R:

[0128] R=ω1·C+ω2·P+ω3·T r ①

[0129] Where C is the failure consequence score (such as safety risk, economic loss, environmental impact, etc.); P is the failure probability score (such as historical failure frequency, equipment aging degree, etc.); T r is the maintenance time score (such as maintenance complexity, spare parts availability, etc.), ω1, ω2 and ω3 are weight coefficients, satisfying ω1+ω2+ω3=1. The comprehensive risk score is calculated using several equipment as examples, as shown in Table 1 below;

[0130] Table 1 Comprehensive risk score of equipment

[0131] equipment C (Failure Consequences) P (failure probability) <![CDATA[T r (Maintenance Time)]]> <![CDATA[ω1=0.4,ω2=0.3,ω3=0.3]]> R water pump 8 7 6 0.4·8+0.3·7+0.3·6＝7.1 7.1 induced draft fan 5 4 3 0.4·5+0.3·4+0.3·3＝4.1 4.1 vacuum pump 2 2 2 0.4·2+0.3·2+0.3·2＝2.0 2.0

[0132] (III) Hierarchical decision index, denoted as DI:

[0133] DI=R·M②

[0134] Where M is an adjustment factor, which is used to consider other dynamic factors (such as equipment criticality and production impact range); and the maintenance strategy is divided according to the DI threshold as follows:

[0135] Low risk (DI < 30), no immediate repair required, planned maintenance possible;

[0136] Medium risk (30≤DI<60), it is recommended to strengthen monitoring and perform preventive maintenance when necessary;

[0137] High risk (DI ≥ 60), take corrective maintenance or emergency measures immediately;

[0138] This algorithm quantifies the consequences of failure, failure probability, and repair time, and combines weights and adjustment factors to provide a scientific and flexible maintenance strategy grading evaluation theory. In practical applications, parameters and weights can be adjusted according to specific needs to ensure the rationality and effectiveness of decision-making.

[0139] (6) Strategy reliability analysis

[0140] Based on captured equipment failure data and maintenance strategies, we estimate the survival function and perform statistics to generate a survival curve. The horizontal axis (time) reflects the typical failure-free period of the equipment, guiding the timing of preventive maintenance and optimizing maintenance strategies. If the curve drops sharply after time t', preventive maintenance should be scheduled before t'. We also compare the survival curves of different equipment / failure modes to identify high-risk components.

[0141] Assume that the survival probability S(t) represents the probability that the equipment has not failed before time t, and the calculation formula is:

[0142]

[0143] Among them, t i represents the time point when the i-th fault is observed (arranged in ascending order), d i Indicates that at time t i The number of failed devices, n i Indicates that at time t i The number of devices that were previously at risk;

[0144] Key calculation steps:

[0145] S1. Data preparation:

[0146] The observation data for each device include time (how long the device was in operation until failure or censoring) and event status (1 = failure, 0 = censoring, e.g., no failure before maintenance);

[0147] S2. Sorting and calculation:

[0148] Arrange the failure time (non-censored time) of all devices in ascending order as t1<t2<…<t k ;

[0149] For each t i Calculate d i and n i ;

[0150] S3. Survival probability update:

[0151] Initialization S(0)=1 (all devices are initially fault-free);

[0152] For each ti , update the survival probability:

[0153]

[0154] Among them, deletion (such as planned downtime without failure) only affects n at subsequent time points i , do not participate in d i calculate;

[0155] And through mathematical supplementation, we find the median line of the maintenance strategy update as follows:

[0156] Variance estimation:

[0157]

[0158] The median survival time is the time t that satisfies S(t) = 0.5, which represents the average reliable life of the equipment. It can intuitively quantify the reliability attenuation law of the equipment and provide a statistical basis for dynamically adjusting the maintenance cycle.

[0159] Example 2:

[0160] like Figure 2 As shown, a maintenance decision-making method based on big data driving includes the following steps:

[0161] Step 1: Call the asset structure tree module to capture equipment master data, spare parts resource data, human resources data, and equipment operation data;

[0162] Step 2: Call the storage device structure decomposition, failure mode, maintenance strategy, and health indicators to digitally model the device and identify the data corresponding to step 1;

[0163] Step 3: Collect maintenance process content for users to browse and view on the big data platform, and update it to the equipment digital model in real time;

[0164] Step 4: Standardize the contents of the maintenance and overhaul document packages to guide the execution of the maintenance and overhaul process, and comprehensively record the maintenance results for users to retrieve and review from the equipment digital model;

[0165] Step 5: Using a hierarchical evaluation algorithm, the captured maintenance process content and data captured by the asset structure tree module are analyzed, and the fault knowledge base management module is called to determine the equipment maintenance priority or strategy;

[0166] Step 6: By estimating the survival function and performing statistics, a survival curve is obtained to reflect the typical trouble-free period of the equipment, guide the preventive maintenance time point, and optimize the maintenance strategy.

[0167] Example 3:

[0168] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the maintenance decision-making method in embodiment 2 when executing the computer program.

[0169] Among them, the memory and processor are set up on the data server, equipped with a CPU (Central Processing Unit) with more than 40 cores, more than 256G of memory, dual gigabit network cards, 2*800W power supplies, and a hard disk of no less than 960G*2SSD (Solid State Disk) + 4TB*6SSDRAID (Redundant Arrays of Independent Disks), which realizes interface docking and integration with the SIS real-time data monitoring system, obtains relevant data to realize business functions; in Example 1, all interface development work between the intelligent perception system and the SIS real-time data monitoring system fully considers integration and convenience, and automatically realizes the unification of business processes and data, as well as the sharing and communication of data.

[0170] Moreover, the data server reads the device-associated measurement point information and measurement point data through the interface, receives the alarm information of the device-associated measurement point in the SIS system status monitoring, obtains the characteristic quantity of the equipment asset health assessment, and regularly obtains and analyzes the alarm information pushed by the comprehensive status monitoring of the equipment.

[0171] Also includes:

[0172] The application server is configured with a 20-core CPU, 128GB of memory, a hard drive of no less than 2TB*2 SSD RAID, dual Gigabit network cards, and 2*800W power supplies. It is used to build various functional modules and is developed based on a modular and componentized architecture. Users can complete the intelligent perception system operations in Example 1 using a web browser without installing other client software.

[0173] Among them, the maintenance decision system in Example 1 adopts a back-end web application, which is developed based on SpringBoot (an open source application framework on the Java platform) and SpringSecurity (a security framework that can provide descriptive security access control solutions for Spring-based enterprise application systems), and is combined with gateway layer components such as Nginx to realize reverse proxy and traffic load balancing; the front-end browser page is developed based on UI frameworks such as Vue, ElementUI, and EChart, and can adapt to all browsers based on the Chrome kernel.

[0174] NAS storage, equipped with 32G running memory, 4 cores, 2*NVME disk slots, and 8*18TNAS dedicated 3.5-inch mechanical keyboard.

[0175] Data confidentiality, document information, business secrets and other information learned during the intelligent perception process must be kept strictly confidential, and appropriate confidentiality measures must be taken.

[0176] Example 4:

[0177] A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the maintenance decision-making method in Example 2 are implemented, and the storage medium includes at least one or more of TTL (bipolar transistor), MOS (metal oxide semiconductor), HDD / SSD (disk), CD / DVD (optical disk), USB flash drive (USB flash drive), SD card (Secure Digital Memory Card), CF card (Compact Flash), RAM (random access memory), ROM (read-only memory), and LiDAR (lidar and camera joint calibration storage medium).

[0178] The foregoing merely represents optional embodiments of the present invention. It should be noted that those skilled in the art may make various improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained herein shall, unless otherwise specified or limited, be implemented in accordance with conventional means in the art.

Claims

1. A maintenance decision system driven by big data, characterized by: include: The asset structure tree module is used to capture equipment master data, spare parts resource data, human resources data, and equipment operation data; The fault knowledge base management module stores device structure decomposition, failure modes, maintenance strategies, and health indicators, providing a reference for digital modeling of equipment; The equipment defect management module is used to collect content during the maintenance process for users to browse and view on the big data platform; The maintenance operation standardization module is used to standardize the content of the maintenance document package to guide the execution process of major and minor maintenance and comprehensively record the maintenance work results; The hierarchical push module uses a hierarchical evaluation algorithm to analyze the captured maintenance process content and data captured by the asset structure tree module, and calls the fault knowledge base management module to determine the equipment maintenance priority or strategy; The strategy reliability analysis module estimates the survival function and performs statistics to obtain a survival curve that reflects the typical trouble-free cycle of the equipment, guides the preventive maintenance time point, and is used to optimize the maintenance strategy.

2. The maintenance decision system based on big data drive according to claim 1, characterized in that: The grading evaluation algorithm, during the maintenance operation, scores the consequences of failure, the probability of failure and the maintenance time respectively; Then a comprehensive risk score is performed, and at the same time, a hierarchical decision index is given, recorded as DI, and the maintenance strategy is divided according to the DI threshold.

3. The maintenance decision system based on big data drive according to claim 1, characterized in that: The survival function assumes that the survival probability S(t) represents the probability that the equipment has not failed before time t, t i Indicates the time point when the i-th fault is observed, and prepares the observation data of each device including time and event status; Arrange the failure time of all devices in ascending order as t1<t2<…<t k ; Initialize S(0)=1; for each t i , update the survival probability; And through mathematical supplementation, the median line of maintenance strategy update is found to intuitively quantify the reliability attenuation law of equipment and provide a statistical basis for dynamically adjusting the maintenance cycle.

4. The maintenance decision system based on big data drive according to claim 1, characterized in that: The asset structure tree module includes: The equipment master data management unit interfaces with external systems to obtain equipment master data, build a complete asset digital model, and store and manage enterprise equipment master data information; Spare parts resource data management unit, which centrally manages and maintains spare parts resources, supports querying spare parts resource information, and provides data support for maintenance cost estimation; The human resources data management unit manages and maintains basic human resources data, supports querying human resources information, and provides data support for maintenance cost estimation; The equipment operation data management unit collects physical signals from thermal power auxiliary equipment in real time through sensors, stores raw data and result data, and digitally models thermal power auxiliary equipment.

5. The maintenance decision system based on big data drive according to claim 1, characterized in that: The fault knowledge base management module includes: Equipment structure decomposition library, providing structural decomposition references for corresponding equipment in the fault knowledge base; Fault mode library, which provides references to the fault causes and fault modes of corresponding components in the fault knowledge base; Maintenance strategy library, which provides references to the failure causes, maintenance measures, and maintenance cycles of corresponding components in the fault knowledge base; The health indicator library provides a reference for the corresponding health indicators of the devices in the fault knowledge base.

6. The maintenance decision system based on big data drive according to claim 1, characterized in that: The equipment defect management module includes: The defect data collection unit includes defect registration, defect confirmation, operation information, and defect elimination process content. Defect information is automatically collected, filled in, and improved through each page field and defect category to achieve standardization of all defect data. Defect query unit, which allows users to browse and view all defects synchronized from the big data platform, including defect information completion status, number, defect description, unit to which it belongs, and defect category; The defect information improvement unit can call the failure mode and effect analysis results or fault knowledge base data in the defect information improvement interface, improve the defects according to the defect description information, select the component information where the defect occurs, and associate the fault mode and fault information.

7. The maintenance decision system based on big data drive according to claim 1, characterized in that: The maintenance operation standardization module includes: Maintenance time planning unit, which lists the maintenance work to be carried out, time plan and arrangement of related resources; Maintenance work instruction unit, which provides maintenance work steps, operating procedures, and safety measures; Maintenance record and report unit, which records key information during the maintenance process, including equipment status, maintenance details, work time and personnel, as a reference for subsequent maintenance; The maintenance tools and spare parts unit lists the tools and equipment that need to be used, as well as the spare parts and consumables that need to be replaced or used later.

8. A maintenance decision method based on big data driving, applied to a maintenance decision system based on big data driving according to any one of claims 1 to 7, characterized in that: The following steps are involved: Step 1: Call the asset structure tree module to capture equipment master data, spare parts resource data, human resources data, and equipment operation data; Step 2: Call the storage device structure decomposition, failure mode, maintenance strategy, and health indicators to digitally model the device and identify the data corresponding to step 1; Step 3: Collect maintenance process content for users to browse and view on the big data platform, and update it to the equipment digital model in real time; Step 4: Standardize the contents of the maintenance and overhaul document packages to guide the execution of the maintenance and overhaul process, and comprehensively record the maintenance results for users to retrieve and review from the equipment digital model; Step 5: Using a hierarchical evaluation algorithm, the captured maintenance process content and data captured by the asset structure tree module are analyzed, and the fault knowledge base management module is called to determine the equipment maintenance priority or strategy; Step 6: By estimating the survival function and performing statistics, a survival curve is obtained to reflect the typical trouble-free period of the equipment, guide the preventive maintenance time point, and optimize the maintenance strategy.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to claim 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to claim 8 are implemented.

Images