Equipment failure maintenance method and device, computer equipment and readable storage medium

Abstract

The invention relates to an equipment failure maintenance method and device, computer equipment and a readable storage medium. The method comprises the following steps: under the condition that an equipment operation state index value of a generator is out of a preset safety threshold range, performing fault detection on the generator based on equipment operation data to obtain equipment fault information; detecting local health states of the plurality of auxiliary substructures based on the equipment operation data, and detecting an overall health state of the generator based on the plurality of local health states; predicting the equipment residual life of the generator based on the equipment operation data, the plurality of local health states and the overall health state; according to the equipment residual life, the equipment fault information, the local health state of the plurality of auxiliary substructures and the overall health state of the generator, the fault maintenance priority of the generator is evaluated; and performing fault maintenance on the generator according to the equipment fault information and the fault maintenance priority. By adopting the method, equipment fault maintenance can be accurately carried out.

Description

Technical Field

[0001] The present application relates to the technical field of fault handling, and in particular to a method and apparatus for repairing equipment faults, a computer device, and a readable storage medium. Background Art

[0002] With the development of science and technology, repairing equipment failures is a very important part of the equipment's life cycle. By repairing equipment failures, the equipment's service life can be extended and its healthy operation can be guaranteed.

[0003] At present, the operation and maintenance of generators mainly rely on planned maintenance and technical transformation and upgrading of online monitoring systems. First of all, generators mainly carry out A-level maintenance work according to the plan of 5-6 years / time, carry out comprehensive maintenance, or during daily inspections and B and C-level maintenance of units, if the generator has non-stop failures, performance parameters or preventive test data have major abnormalities, the generator will undergo certain unplanned or expanded maintenance.

[0004] However, the current equipment failure repair method is not accurate enough. Summary of the Invention

[0005] Based on this, it is necessary to provide an accurate equipment fault repair method, device, computer equipment, computer-readable storage medium and computer program product to address the above technical problems.

[0006] In a first aspect, the present application provides a method for repairing equipment failure, comprising:

[0007] Acquiring equipment operating data of the generator, wherein the equipment operating data includes an equipment operating status indicator value;

[0008] When the equipment operating status indicator value is outside the preset safety threshold range, the generator fault detection is performed based on the equipment operating data to obtain equipment fault information;

[0009] Based on the equipment operation data, the local health status of multiple auxiliary substructures in the generator is detected, and based on the local health status of the multiple auxiliary substructures, the overall health status of the generator is detected;

[0010] Predict the remaining life of the generator based on equipment operating data, the local health status of multiple attached substructures, and the overall health status of the generator;

[0011] Evaluate the generator's fault repair priority based on the remaining life of the equipment, equipment fault information, the local health status of multiple attached substructures, and the overall health status of the generator;

[0012] Perform fault repairs on the generator based on equipment fault information and fault repair priority.

[0013] In one embodiment, detecting the local health status of multiple auxiliary substructures in a generator based on equipment operation data includes:

[0014] Extracting local operating data of multiple auxiliary substructures in the generator from equipment operating data;

[0015] For each subsidiary substructure, indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators is generated according to local operation data, and the local health status of the subsidiary substructure is detected based on all indicator degradation information.

[0016] In one embodiment, detecting the local health status of the attached substructure based on all indicator degradation information includes:

[0017] Obtaining first weights and second weights of a plurality of preset health status assessment indicators, wherein the first weights represent a theoretical importance of the preset health status assessment indicators for assessing the local health status of the accessory substructure, and the second weights represent an actual importance of the preset health status assessment indicators for assessing the local health status of the accessory substructure;

[0018] For each preset health status assessment indicator, generating an indicator weight according to the first weight and the second weight;

[0019] Based on the indicator weights and indicator degradation information of all preset health status assessment indicators, the local health status of the attached substructure is detected.

[0020] In one embodiment, predicting the remaining life of a generator based on the equipment operation data, the local health status of a plurality of attached substructures, and the overall health status of the generator includes:

[0021] Determining an equipment fault evolution model of the generator, wherein the equipment fault evolution model is obtained based on historical operating data of a first equipment under a frequent start-stop condition of the generator;

[0022] Obtain the initial degradation state of the generator in the previous time period, and process the initial degradation state through the equipment fault evolution model to generate the current degradation state of the generator in the current time period;

[0023] Predict the remaining life of the generator based on the current degradation state, equipment operating data, the local health status of multiple attached substructures, and the overall health status of the generator.

[0024] In one embodiment, after evaluating the fault repair priority of the generator based on the remaining life of the equipment, equipment fault information, local health status of multiple attached substructures, and the overall health status of the generator, the method includes:

[0025] Obtain the available resources and planned maintenance duration for the target object, where the target object is the object of generator fault repair;

[0026] Based on the available resources for maintenance and the planned maintenance duration, determine the number of repairable devices for the target object and the range of fault repair priorities corresponding to the number of repairable devices;

[0027] When the fault repair priority is outside the fault repair priority range, an equipment repair result of the generator is generated, wherein the equipment repair result indicates that the fault repair of the generator is not performed.

[0028] In one embodiment, before performing fault detection on the generator based on the equipment operation data to obtain equipment fault information, the method further includes:

[0029] Acquire historical operation data of a second device of the generator and historical device fault information corresponding to the historical operation data of the second device;

[0030] Performing data expansion on the historical operation data of the second device to obtain the historical operation data of the target device;

[0031] Extract multiple historical equipment operation features corresponding to the historical operation data of the target equipment, and fuse the multiple historical equipment operation features to obtain historical fusion features;

[0032] Train equipment fault detection models based on historical fusion features and historical equipment fault information;

[0033] Based on the equipment operation data, the generator fault detection is carried out to obtain equipment fault information, including:

[0034] Extracting multiple device operation features from the device operation data, and fusing the multiple device operation features to obtain fused features;

[0035] The fused features are processed through the trained equipment fault detection model to obtain the equipment fault information of the generator.

[0036] In one embodiment, the fused features are processed by the trained equipment fault detection model to obtain equipment fault information of the generator, including:

[0037] Processing the fused features through the trained equipment fault detection model to obtain initial equipment fault information of the generator, wherein the initial equipment fault information includes a first equipment fault type;

[0038] Obtain a second device fault type of the generator, where the second device fault type is a theoretical device fault type of the generator uploaded by the target terminal;

[0039] Generate a target device failure type according to the first device failure type and the second device failure type;

[0040] Based on the target device fault type, the initial device fault information is updated to obtain the device fault information of the generator.

[0041] In a second aspect, the present application also provides an equipment failure repair device, comprising:

[0042] A data acquisition module is used to acquire the equipment operation data of the generator, wherein the equipment operation data includes an equipment operation status indicator value;

[0043] A fault detection module is used to detect faults on the generator based on the equipment operation data and obtain equipment fault information when the equipment operation status indicator value is outside the preset safety threshold range;

[0044] A health status assessment module is used to detect the local health status of multiple auxiliary substructures in the generator based on the equipment operation data, and to detect the overall health status of the generator based on the local health status of the multiple auxiliary substructures;

[0045] The remaining life prediction module is used to predict the remaining life of the generator equipment based on the equipment operation data, the local health status of multiple attached substructures and the overall health status of the generator;

[0046] A priority determination module is used to evaluate the fault repair priority of the generator based on the remaining life of the equipment, equipment fault information, the local health status of multiple attached substructures and the overall health status of the generator;

[0047] The maintenance control module is used to perform fault maintenance on the generator according to the equipment fault information and fault maintenance priority.

[0048] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0049] Acquiring equipment operating data of the generator, wherein the equipment operating data includes an equipment operating status indicator value;

[0050] When the equipment operating status indicator value is outside the preset safety threshold range, the generator fault detection is performed based on the equipment operating data to obtain equipment fault information;

[0051] Based on the equipment operation data, the local health status of multiple auxiliary substructures in the generator is detected, and based on the local health status of the multiple auxiliary substructures, the overall health status of the generator is detected;

[0052] Predict the remaining life of the generator based on equipment operating data, the local health status of multiple attached substructures, and the overall health status of the generator;

[0053] Evaluate the generator's fault repair priority based on the remaining life of the equipment, equipment fault information, the local health status of multiple attached substructures, and the overall health status of the generator;

[0054] Perform fault repairs on the generator based on equipment fault information and fault repair priority.

[0055] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0056] Acquiring equipment operating data of the generator, wherein the equipment operating data includes an equipment operating status indicator value;

[0057] When the equipment operating status indicator value is outside the preset safety threshold range, the generator fault detection is performed based on the equipment operating data to obtain equipment fault information;

[0058] Based on the equipment operation data, the local health status of multiple auxiliary substructures in the generator is detected, and based on the local health status of the multiple auxiliary substructures, the overall health status of the generator is detected;

[0059] Predict the remaining life of the generator based on equipment operating data, the local health status of multiple attached substructures, and the overall health status of the generator;

[0060] Evaluate the generator's fault repair priority based on the remaining life of the equipment, equipment fault information, the local health status of multiple attached substructures, and the overall health status of the generator;

[0061] Perform fault repairs on the generator based on equipment fault information and fault repair priority.

[0062] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0063] Acquiring equipment operating data of the generator, wherein the equipment operating data includes an equipment operating status indicator value;

[0064] When the equipment operating status indicator value is outside the preset safety threshold range, the generator fault detection is performed based on the equipment operating data to obtain equipment fault information;

[0065] Based on the equipment operation data, the local health status of multiple auxiliary substructures in the generator is detected, and based on the local health status of the multiple auxiliary substructures, the overall health status of the generator is detected;

[0066] Predict the remaining life of the generator based on equipment operating data, the local health status of multiple attached substructures, and the overall health status of the generator;

[0067] Evaluate the generator's fault repair priority based on the remaining life of the equipment, equipment fault information, the local health status of multiple attached substructures, and the overall health status of the generator;

[0068] Perform fault repairs on the generator based on equipment fault information and fault repair priority.

[0069] The above-mentioned equipment fault repair method, apparatus, computer device, computer-readable storage medium, and computer program product obtain equipment operating data of a generator, wherein the equipment operating data includes an equipment operating status indicator value; if the equipment operating status indicator value is outside a preset safety threshold range, perform fault detection on the generator based on the equipment operating data to obtain equipment fault information; detect the local health status of multiple auxiliary substructures in the generator based on the equipment operating data, and detect the overall health status of the generator based on the local health status of the multiple auxiliary substructures; predict the remaining life of the generator based on the equipment operating data, the local health status of the multiple auxiliary substructures, and the overall health status of the generator; assess the fault repair priority of the generator based on the remaining life, equipment fault information, the local health status of the multiple auxiliary substructures, and the overall health status of the generator; and perform fault repair on the generator based on the equipment fault information and the fault repair priority. Throughout the entire process, the generator fault repair priority is accurately generated by comprehensively considering multiple perspectives, including the remaining life, equipment fault information, the local health status of the multiple auxiliary substructures, and the overall health status of the generator. Based on the equipment fault information and the fault repair priority, timely and accurate repair of the generator requiring priority repair can be performed, thereby improving the accuracy of the generator repair process. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0071] Figure 1 A diagram showing an application environment of a method for repairing equipment failures in one embodiment;

[0072] Figure 2 A schematic flow chart of a method for repairing equipment failure in one embodiment;

[0073] Figure 3 A schematic flow chart of a method for repairing equipment failure in another embodiment;

[0074] Figure 4 A flowchart of a method for repairing equipment failure in a specific application embodiment;

[0075] Figure 5 A structural block diagram of an equipment failure repair device in one embodiment;

[0076] Figure 6 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0077] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are used to explain this application and are not intended to limit this application.

[0078] The equipment failure repair method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal 102 communicates with the generator 104 and the maintenance equipment 106 via a network. The maintenance equipment 106 is used to repair the generator 104 that needs to be repaired under the control of the terminal 102.

[0079] The user triggers the equipment fault maintenance control of the generator 104 on the equipment fault maintenance interface of the terminal 102, and the terminal 102 responds to the trigger request for the equipment fault maintenance and obtains the equipment operation data of the generator 104, wherein the equipment operation data includes the equipment operation status indicator value; when the equipment operation status indicator value is outside the preset safety threshold range, the generator is fault detected based on the equipment operation data to obtain equipment fault information; based on the equipment operation data, the local health status of multiple subsidiary substructures in the generator is detected, and based on the local health status of the multiple subsidiary substructures, the overall health status of the generator is detected; based on the equipment operation data, the local health status of the multiple subsidiary substructures and the overall health status of the generator, the remaining equipment life of the generator is predicted; according to the remaining equipment life, the equipment fault information, the local health status of the multiple subsidiary substructures and the overall health status of the generator, the fault maintenance priority of the generator is evaluated; according to the equipment fault information and the fault maintenance priority, the maintenance equipment 106 is controlled to perform fault maintenance on the generator 104.

[0080] The terminal 102 may be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices may include smart speakers, smart TVs, smart air conditioners, smart car devices, and projectors. Portable wearable devices may include smart watches, smart bracelets, and head-mounted devices. Head-mounted devices may include virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, and the like.

[0081] In an exemplary embodiment, Figure 2 As shown, a method for repairing equipment failure is provided, which is applied to Figure 1 The terminal 102 in FIG. 1 is used as an example for explanation.

[0082] S100: Acquire equipment operation data of a generator, wherein the equipment operation data includes an equipment operation status indicator value.

[0083] The generator in this application refers to a thermal power plant generator, which converts thermal energy into electrical energy to provide electricity services to users. Equipment operation data includes at least one equipment operation status indicator value, which includes but is not limited to existing equipment operation status indicators from the DCS (Distributed Control System) and equipment operation status indicators collected in real time by the online monitoring system. Furthermore, it also includes relevant operation status indicators such as the generator's full lifecycle test data and special condition data.

[0084] Specifically, the user triggers the equipment fault maintenance control of the generator on the equipment fault maintenance interface of the terminal, and the terminal responds to the trigger request of the equipment fault maintenance and obtains the equipment operation data of multiple generators. In the process of the terminal obtaining the equipment operation data of multiple generators, it can obtain the existing equipment operation status indicator value from the database, or it can detect the equipment operation status indicator value of the generator online in real time. In addition, it can also obtain equipment operation status indicator values such as the generator's full life cycle test data and the generator's special situation data.

[0085] In an exemplary embodiment, the existing equipment operating status indicator values of the DCS (Distributed Control System) include but are not limited to: thermal parameter values, electrical parameter values, mechanical parameter values and environmental parameter values; thermal parameter values include but are not limited to: stator winding coil temperature, winding bar interlayer temperature and core temperature, hydrogen purity, pressure, hydrogen leakage, slip ring temperature, stator cooling water temperature, pH value and slip ring temperature, etc.; electrical parameter values include but are not limited to: stator partial discharge, rotor interturn short circuit, shaft voltage, shaft current, main engine short circuit and discharge, etc.; mechanical parameter values include but are not limited to: generator rotor shaft system shaft vibration, seat vibration and stator end vibration, etc.; environmental parameter values include but are not limited to: humidity and leakage parameters, etc.

[0086] The equipment operating status indicator values collected by the online monitoring system include but are not limited to: rotor inter-turn short circuit, stator partial discharge, collector ring temperature, carbon brush current and temperature and other parameter values.

[0087] The generator's full life cycle test data includes but is not limited to: factory test data, handover test data, pre-test test data, technical specifications and test cycles, important component replacement time and cycle, etc.

[0088] Special condition data of the generator include but are not limited to: start and stop parameters, etc.

[0089] In an exemplary embodiment, after obtaining the equipment operation data of the generator, it is necessary to perform data preprocessing on the obtained equipment operation data of the generator, including data screening, data cleaning, online monitoring system data correction, data standardization processing, normalization processing, multi-dimensional data fusion, data dynamic threshold setting, etc.

[0090] Among them, data screening refers to screening out suitable data for analysis based on system functions, data classification, time series, etc.

[0091] Data cleaning involves using various methods, such as multivariate heterogeneous data fusion diagnosis, data time-frequency separation diagnosis, wavelet analysis, and curve fitting, to eliminate the effects of abnormal data, loads, and environmental conditions on parameters. For example, multivariate heterogeneous data fusion diagnosis involves selecting a few parameters and simultaneously displaying their corresponding time domain waveforms, performing polarity and time difference analysis on the graphs to facilitate diagnosis. Data time-frequency separation diagnosis involves using time-frequency separation to diagnose high-frequency signals with significant noise. Furthermore, issues such as missing values, out-of-bounds values, and duplicate data can be addressed from the perspectives of data accuracy, completeness, consistency, uniqueness, timeliness, and validity.

[0092] Online monitoring system data correction refers to the regular comparison of the same items in the online monitoring system data and the pre-test data. The pre-test data is used as a sample, and the change trend of the online monitoring data should match it. For large deviations, sensor calibration or numerical correction should be performed.

[0093] Data standardization, normalization processing and multi-dimensional data fusion refer to extracting the characteristic vectors of the data for standardization, normalizing the data, and then fusing the multi-dimensional data required for the same function into a vector matrix.

[0094] The setting of dynamic data thresholds refers to setting the dynamic thresholds of generator voltage, current, temperature, vibration, hydrogen, cooling water, sealing oil and other related parameters under different loads in accordance with national industry regulations such as DL / T 596-2021 "Preventive Tests for Power Equipment", GB50150-2016 "Electrical Equipment Acceptance Test Standards for Electrical Installation Engineering", manufacturer's manuals, and on-site personnel operation and maintenance experience requirements, and determining the abnormal classification alarm thresholds.

[0095] S200 , when the device operating status indicator value is outside a preset safety threshold range, a fault detection is performed on the generator based on the device operating data to obtain device fault information.

[0096] Specifically, based on the study of the operating parameter characteristics of generators under deep peak-shaving conditions, a generator equipment operating status indicator value deviation early warning method is established. When the equipment operating status indicator value is outside the preset safety threshold range, the generator is considered to have failed. At this time, the terminal promptly generates an early warning signal for the generator.

[0097] When a generator fails, fault detection is performed based on the device's operating data to obtain device fault information, which typically includes the fault type and location. Fault detection is typically achieved through machine learning. This involves processing device operating data using machine learning methods to detect the generator's faults and obtain fault information.

[0098] In an exemplary embodiment, the equipment operation data of the generator is collected by setting multiple measuring points on the generator. When individual data produces unexpected deviations due to problems with the measuring points, intelligent analysis can be performed to eliminate the erroneous impact of the deviation data on the calculation, and timely warnings can be given for abnormal parameter deviations, thereby ensuring the accuracy of the operation data collection equipment in collecting equipment operation data and timely warnings.

[0099] In an exemplary embodiment, the preset safety threshold range can be determined by similar abnormal data or change trends that appear in historical equipment operating status indicator values when the operating data acquisition device collects historical equipment operating status indicator values of generators of the same or similar model and capacity level within a historical time period.

[0100] In an exemplary embodiment, there is more than one generator. After obtaining the equipment operation data of multiple generators, a clustering method can be used to analyze the safety parameters of different generators, different service years, and different peak-shaving depths. For example, key safety parameter data such as the temperature of key components of the generator, end vibration, inter-turn current, and hydrogen leakage can be analyzed. Through the transfer learning method, knowledge sharing and migration between different units can be achieved, and the preset safety threshold range corresponding to the equipment operation status indicator value can be obtained to provide support for strategies such as safe peak-shaving, maintenance and technical transformation of different types of units.

[0101] S300 , detecting local health status of multiple subsidiary substructures in the generator based on equipment operation data, and detecting the overall health status of the generator based on the local health status of the multiple subsidiary substructures.

[0102] Specifically, pre-commissioning information, operating information, maintenance and test information, and other information throughout the equipment's life cycle can be obtained from the equipment's operating data, and a health status assessment can be performed on the electrical and non-electrical status quantities contained in the information. The generator's health status assessment should include a local health status assessment of multiple subsidiary substructures and an overall health status assessment of the generator. Subsidiary substructures include stators, rotors, and subsidiary systems such as hydrogen, cooling water, and sealing oil. The local health status of multiple subsidiary substructures is divided into normal, cautionary, abnormal, and severe states according to the degree of degradation. The specific degradation standard table for subsidiary substructures can be found in GB∕T 43188-2023, "Guidelines for Generator Equipment Condition Assessment." The overall health status should be determined based on the local health status of multiple subsidiary substructures and the overall performance of the generator.

[0103] In other words, when detecting the local health status of multiple auxiliary substructures within a generator based on equipment operating data, the local equipment operating data for each auxiliary substructure can be extracted from the equipment operating data. Based on this local equipment operating data, the local health status of each auxiliary substructure can be assessed. Furthermore, the assessed local health status of the multiple auxiliary substructures can be comprehensively analyzed to generate the overall health status of the generator.

[0104] S400 , predicting the remaining life of the generator based on the equipment operation data, the local health status of the plurality of attached substructures, and the overall health status of the generator.

[0105] Specifically, based on the equipment operation data, the local health status of multiple subsidiary substructures and the overall health status of the generator, the operating status of the generator can be judged, and based on the operating status of the generator, the remaining life of the generator can be predicted. For example, if the values of multiple equipment operation status indicators are all outside the preset safety threshold range, and the local health status of multiple subsidiary substructures and the overall health status of the generator both indicate abnormal health status, then the remaining life of the generator equipment is short; if fewer equipment operation status indicator values are outside the preset safety threshold range, fewer local health status of subsidiary substructures indicate abnormal health status, and the overall health status of the generator indicates normal health status, then the remaining life of the generator equipment is long.

[0106] In an exemplary embodiment, a generator degradation model characterized by "stress load-electromagnetic characteristics-fatigue damage" is established using dynamic and electromagnetic equations, and the generator degradation model is used to process equipment operation data to obtain the degradation state of the generator. Based on the degradation state of the generator, the local health status of multiple auxiliary substructures and the overall health status of the generator, the remaining equipment life of the generator is predicted.

[0107] S500 , evaluating the fault repair priority of the generator based on the remaining life of the device, the device fault information, the local health status of multiple subsidiary substructures, and the overall health status of the generator.

[0108] Specifically, a fault repair priority of a generator is generated when repairing the generator by comprehensively considering the remaining life of the device, device fault information, the local health status of multiple subsidiary substructures, and the overall health status of the generator.

[0109] Among them, fault repair priority can be considered from two levels. The first level is to evaluate the fault repair priority of each subsidiary substructure in a certain generator in the factory. The second level is to evaluate the fault repair priority of multiple generators under the jurisdiction of multiple factories or provincial companies, so as to diagnose and analyze the health status of each generator, coordinate the overall management, coordinate the health level and power generation of generators in the region, and achieve a steady improvement in equipment health and power generation in the region.

[0110] In other words, fault repair priority can refer to which generator is repaired first. The higher the fault repair priority of a generator among multiple generators, the higher the priority of the generator, the higher the priority of repair. Conversely, the lower the fault repair priority of a generator among multiple generators, the later it will be repaired. Fault repair priority can also refer to which part of a generator is repaired first. The higher the fault repair priority of a generator's auxiliary substructure, the higher the priority of repair. Conversely, the lower the fault repair priority of a generator's auxiliary substructure, the later it will be repaired. In addition, decision-making recommendations for the location and number of additional measuring points can be provided. For example, more measuring points can be added at locations with higher repair priorities, or more measuring points can be added at fault locations that require special attention.

[0111] In practical applications, when evaluating the fault repair priority of each of multiple generators, the remaining equipment life of the generator is directly proportional to the fault repair priority; the equipment fault information includes the equipment fault level, which is inversely proportional to the fault repair priority; the local health status of multiple subsidiary substructures includes the local health status level, which is inversely proportional to the fault repair priority; the overall health status of the generator includes the overall health status level, which is inversely proportional to the fault repair priority; that is, generators with shorter remaining equipment life, higher equipment fault level, local health status level, and overall health status level can be repaired first.

[0112] In evaluating the fault repair priorities of multiple subsidiary substructures in each generator, the equipment fault information includes the subsidiary substructure fault level, which is inversely proportional to the fault repair priority; the local health status of multiple subsidiary substructures includes the local health status level, which is inversely proportional to the fault repair priority; that is, subsidiary substructures in the generator with higher subsidiary substructure fault levels and local health status levels can be repaired first.

[0113] In an exemplary embodiment, the generator fault repair priority is generated from multiple perspectives, including the remaining life of the equipment, equipment fault information, the local health status of multiple subsidiary substructures, and the overall health status of the generator. This can be accomplished by obtaining maintenance weight coefficients corresponding to the remaining life of the equipment, equipment fault information, the local health status of multiple subsidiary substructures, and the overall health status of the generator, and performing weighted processing on the parameters based on the maintenance weight coefficients corresponding to the above parameters to generate the generator fault repair priority.

[0114] In an exemplary embodiment, the preset safety threshold range corresponding to the equipment operation status index value can also be adjusted based on the remaining life of the equipment, equipment fault information, the local health status of multiple subsidiary substructures and the overall health status of the generator. For example, when the fault repair priority obtained based on the remaining life of the equipment, equipment fault information, the local health status of multiple subsidiary substructures and the overall health status of the generator is low, the preset safety threshold range corresponding to the equipment operation status index value can be adjusted based on the equipment operation data of the generator so that the equipment operation data of the equipment is within the preset safety threshold range.

[0115] S600: Perform fault repair on the generator according to the equipment fault information and the fault repair priority.

[0116] Specifically, based on the fault repair priority, multiple generators are repaired in sequence. During the repair process of each generator, the target repair strategy information corresponding to the equipment fault information of the generator can be queried, and the generator is repaired according to the target repair strategy information.

[0117] In an exemplary embodiment, repairing the generator according to the equipment fault information and the fault repair priority further includes:

[0118] Obtain a fault maintenance mapping relationship, where the fault maintenance mapping relationship is a mapping relationship between the fault information of the generator and the maintenance strategy information; based on the fault maintenance mapping relationship, query the target maintenance strategy information corresponding to the equipment fault information; and repair the generator based on the fault maintenance priority and the target maintenance strategy information.

[0119] Specifically, a mapping relationship is formed between the equipment failure information and maintenance strategy information in historical failure cases and added to the failure information database. The maintenance strategy information includes accident handling measures and expert maintenance suggestions.

[0120] After obtaining the equipment fault information, the target maintenance strategy information corresponding to the equipment fault information is queried based on the fault maintenance mapping relationship. In the process of repairing multiple generators in sequence based on the fault maintenance priority, the target maintenance strategy information is used to guide on-site production management personnel to repair the generators.

[0121] In the above-mentioned equipment fault repair method, equipment operating data of a generator is obtained, wherein the equipment operating data includes an equipment operating status indicator value; if the equipment operating status indicator value is outside a preset safety threshold range, a fault detection is performed on the generator based on the equipment operating data to obtain equipment fault information; the local health status of multiple auxiliary substructures in the generator is detected based on the equipment operating data, and the overall health status of the generator is detected based on the local health status of the multiple auxiliary substructures; the remaining equipment life of the generator is predicted based on the equipment operating data, the local health status of the multiple auxiliary substructures, and the overall health status of the generator; the fault repair priority of the generator is assessed based on the remaining equipment life, equipment fault information, the local health status of the multiple auxiliary substructures, and the overall health status of the generator; and the generator is repaired based on the equipment fault information and the fault repair priority. Throughout the entire process, the generator fault repair priority is accurately generated by comprehensively considering multiple perspectives, including the remaining equipment life, equipment fault information, the local health status of the multiple auxiliary substructures, and the overall health status of the generator. Based on the equipment fault information and the fault repair priority, timely and accurate repair of the generator requiring priority repair can be performed, thereby improving the accuracy of the generator repair process.

[0122] In an exemplary embodiment, the factory test and acceptance test data in the full life cycle test data should enter the relevant test item data in accordance with the requirements of GB50150-2016 "Electrical Equipment Acceptance Test Standard for Electrical Installation Engineering", including the insulation resistance and absorption ratio or polarization index, DC resistance, DC withstand voltage test and leakage current, AC withstand voltage test of the generator stator winding, the insulation resistance, DC resistance, AC withstand voltage of the rotor winding, the insulation resistance and AC withstand voltage test of the excitation circuit, the AC impedance and power loss of the rotor winding, the no-load characteristic curve, the demagnetization time constant and rotor overvoltage multiple at the no-load rated voltage of the generator, the generator stator residual voltage, phase sequence, shaft voltage, dynamic characteristics test of the stator winding end, DC voltage measurement of the stator winding end insulation, rotor ventilation test, water flow test, etc. Pre-test items are entered in accordance with the requirements of DL / T596-2021, the 25 countermeasures, the group's regular work, and the group's technical supervision implementation rules. The pre-test cycle generally follows the principle of "starting from the new and stricter" in accordance with the above regulations. When conflicting procedures and standards from different sources occur during implementation, the group standard takes precedence over industry standards and then over national standards. In special circumstances, the system has the authority to modify the pre-test cycle and can adjust it based on on-site conditions and expert advice. The replacement time and cycle of important components are set according to actual on-site conditions, manufacturer's manual requirements, the replacement cycle of the same or similar models from the manufacturer, and the experience of operation and maintenance personnel.

[0123] In an exemplary embodiment, detecting local health status of multiple auxiliary substructures in a generator based on equipment operation data includes:

[0124] The local operating data of multiple subsidiary substructures in the generator are extracted from the equipment operating data. For each subsidiary substructure, the indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators is generated based on the local operating data. Based on all the indicator degradation information, the local health status of the subsidiary substructure is detected.

[0125] Specifically, in order to realize the health status assessment function, first of all, a generator health status assessment system can be established based on GB∕T 43188-2023 "Guidelines for Generator Equipment Condition Assessment" and expert experience.

[0126] Secondly, the local operating data of multiple subsidiary substructures in the generator are extracted from the equipment operation data. For each subsidiary substructure, the indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators is generated based on the local operating data. The indicator degradation message is used to characterize the degree of degradation of the preset health status assessment indicator. According to the degree of impact on the generator from light to heavy, the indicator degradation messages can be divided into: Level I, Level II and Level III.

[0127] More specifically, (a) when the indicator degradation information exceeds the standard limit requirements and has a relatively small impact on the performance and safe operation of the generator, the indicator degradation information is classified as Level I; (b) when the indicator degradation information exceeds the standard limit requirements and has a relatively large impact on the performance and safe operation of the generator but does not directly affect the safety of the equipment, the indicator degradation information is classified as Level II; (c) when the indicator degradation information exceeds the standard limit requirements and directly affects the safe operation of the generator, the generator indicator degradation information is classified as Level III.

[0128] Finally, based on the degradation information of all indicators, the local health status of the attached substructure is detected, which includes normal state, caution state, abnormal state and severe state.

[0129] More specifically, for each subsidiary substructure, if (a) all indicator degradation information is stable and within the standard limit, the local health status of the subsidiary substructure is normal; (b) there is one indicator degradation information of level I, the local health status of the subsidiary substructure is caution; (c) there is at least one indicator degradation information of level II or multiple indicator degradation information of level I, the local health status of the subsidiary substructure is abnormal; (d) there is at least one indicator degradation information of level III, the local health status of the subsidiary substructure is serious.

[0130] In an exemplary embodiment, based on local operating data, indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators is generated, including: processing the uncertainty of multiple preset health status assessment indicators of the subsidiary substructure through a Gaussian cloud model to obtain the indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators.

[0131] In the above embodiment, indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators is generated according to local operation data, and based on all indicator degradation information, the local health status of the subsidiary substructure can be accurately generated.

[0132] In an exemplary embodiment, detecting the local health status of the attached substructure based on all indicator degradation information includes:

[0133] Obtain first weights and second weights of multiple preset health status assessment indicators; generate an indicator weight for each preset health status assessment indicator based on the first weight and the second weight; and detect the local health status of the attached substructure based on the indicator weights and indicator degradation information of all preset health status assessment indicators.

[0134] The first weight represents the theoretical importance of the preset health status assessment indicator for evaluating the local health status of the accessory substructure, and the second weight represents the actual importance of the preset health status assessment indicator for evaluating the local health status of the accessory substructure.

[0135] Specifically, the indicator weights of multiple preset health status assessment indicators are obtained, and the indicator degradation information of the multiple preset health status assessment indicators is processed according to the indicator weights of the multiple preset health status assessment indicators corresponding to the subsidiary substructure to detect the local health status of the subsidiary substructure.

[0136] The indicator weights are determined using a subjective and objective comprehensive judgment method, namely, using the analytic hierarchy process (AHP) and association rule method that consider expert opinions to obtain the indicator weights. The AHP method that considers expert opinions can be used to obtain the first weights of multiple preset health status assessment indicators. In other words, the first weights are obtained based on expert opinions and represent the theoretical importance of the preset health status assessment indicators for assessing the local health status of the attached substructures. Furthermore, the association rule method can be used to obtain the second weights of multiple preset health status assessment indicators. The second weights reflect the actual importance of the preset health status assessment indicators for assessing the local health status of the attached substructures. The first weights may or may not be consistent with the second weights.

[0137] Furthermore, an indicator weight is generated based on the first weight and the second weight. Generally, the first weight and the second weight of each preset health status assessment indicator can be assigned their own weights, and the first weight and the second weight of the preset health status assessment indicator can be weighted to generate the indicator weight of each preset health status assessment indicator. Finally, based on the indicator weight of each preset health status assessment indicator, the multiple indicator degradation information is weighted to generate the local health status.

[0138] In the above embodiment, by obtaining the first weight and the second weight of multiple preset health status assessment indicators, and the first weight represents the theoretical importance of the preset health status assessment indicator for assessing the local health status of the subsidiary substructure, and the second weight represents the actual importance of the preset health status assessment indicator for assessing the local health status of the subsidiary substructure, the indicator weight of each preset health status assessment indicator can be accurately generated, and then the local health status of each subsidiary substructure can be accurately assessed.

[0139] In one exemplary embodiment, the overall health of a generator is detected based on the local health status of multiple subsidiary substructures. This includes utilizing DSMT (Dynamic Stochastic Multi-criteria Theory) to fuse the local health status of each subsidiary substructure to determine the overall health status of the generator. Overall health status includes, but is not limited to, normal, caution, abnormal, and critical. When the local health status of a key subsidiary substructure or the overall health status of the generator reaches caution, an alert is issued, along with recommendations for addressing the alert.

[0140] In an exemplary embodiment, predicting the remaining life of a generator based on the equipment operation data, the local health status of a plurality of attached substructures, and the overall health status of the generator includes:

[0141] Determine the equipment failure evolution model of the generator; obtain the initial degradation state of the generator in the previous time period, and process the initial degradation state through the equipment failure evolution model to generate the current degradation state of the generator in the current time period; predict the remaining equipment life of the generator based on the current degradation state, equipment operation data, the local health status of multiple attached substructures, and the overall health status of the generator.

[0142] The equipment fault evolution model is obtained based on the historical operating data of the first equipment under the condition of frequent start and stop of the generator.

[0143] Specifically, before predicting the remaining life of a generator, the system first obtains historical operating data of the generator under frequent start-stop conditions. Based on this data, a multi-field coupling method is used to refine the failure mechanism model under these conditions. This data is then continuously updated and modulated into the performance degradation process using a particle filter, taking into account the dynamic evolution of the generator's degradation state under frequent starts and stops. Noise is then added to the state transition submodel and observation submodel of the equipment fault evolution model to obtain a target equipment fault evolution model. This model simulates the uncertainty of system operating conditions and external loads, enabling comprehensive monitoring, data acquisition, and intelligent diagnosis.

[0144] Furthermore, in the process of predicting the remaining life of the generator, the initial degradation state of the generator in the previous time period is obtained. Through the Bayesian filtering method, the current degradation state at the current moment is recursively deduced from the initial degradation state using the target equipment fault evolution model. Then, the equipment operation data in the current time period, the local health status of multiple subsidiary substructures and the overall health status of the generator are used to complete the Bayesian dynamic update of the current degradation state, obtain the optimal posterior estimate of the degradation state, and predict the remaining life of the generator.

[0145] In this embodiment, by obtaining the equipment fault evolution model of the generator, the current degradation state can be accurately recursively derived from the initial degradation state of the generator in the previous time period. Then, based on the current degradation state, equipment operation data, the local health status of multiple subsidiary substructures and the overall health status of the generator, the remaining equipment life of the generator can be accurately predicted.

[0146] In an exemplary embodiment, after evaluating the fault repair priority of the generator based on the remaining life of the equipment, equipment fault information, local health status of multiple attached substructures, and the overall health status of the generator, the method includes:

[0147] Obtain the available resources for maintenance and the planned maintenance duration of the target object; based on the available resources for maintenance and the planned maintenance duration, determine the number of repairable devices of the target object, and determine the fault repair priority range corresponding to the number of repairable devices; when the fault repair priority is outside the fault repair priority range, generate an equipment maintenance result for the generator, wherein the equipment maintenance result indicates that the fault repair will not be performed on the generator.

[0148] The resources available for maintenance refer to the target object's budget, the planned maintenance duration may be the duration of the target object's plan to perform an overall maintenance on all generators in the area, and the target object is the object of generator fault maintenance.

[0149] Specifically, the budget for maintenance of the target object and the planned maintenance time are limited. The available resources for maintenance and the planned maintenance time of the target object are obtained. Based on the available resources for maintenance and the planned maintenance time, the number of repairable equipment can be determined. The number of repairable equipment can be the number of all equipment that can be repaired when all available resources for maintenance are used up within the planned maintenance time, or it can be the number of all equipment that can be repaired when the planned maintenance time ends but the available resources for maintenance are not used up.

[0150] Determine the fault repair priority range corresponding to the total number of repairable devices. When the fault repair priority is outside the fault repair priority range, the generator will not be repaired. For example, when the total number of repairable devices is 100, the corresponding fault repair priority range is 0-100. If the fault repair priority of the generator evaluated at this time is 120, there are no available maintenance resources to repair it, or the planned maintenance time has exceeded. At this time, the generator will not be repaired.

[0151] When the fault repair priority is within the fault repair priority range, the generator is repaired according to the fault repair priority and the equipment fault information.

[0152] In the above embodiment, by obtaining the available resources for maintenance and the planned maintenance duration of the target object, the number of repairable equipment of the target object can be determined, and it can be judged whether the fault repair priority is within the fault repair priority range corresponding to the number of repairable equipment, so as to accurately determine whether the generator can be repaired.

[0153] In an exemplary embodiment, Figure 3 As shown, before S200, the method further includes:

[0154] S120: Acquire historical operating data of a second device of the generator and historical device fault information corresponding to the historical operating data of the second device.

[0155] S140: Expand the historical operation data of the second device to obtain historical operation data of the target device.

[0156] S160 , extracting multiple historical device operation features corresponding to the historical operation data of the target device, and performing feature fusion on the multiple historical device operation features to obtain historical fusion features.

[0157] S180: Training a device fault detection model based on the historical fusion features and historical device fault information.

[0158] Specifically, the historical operating data of the generator's second device and the historical equipment fault information corresponding to the second device's historical operating data are obtained and entered into the equipment fault database. This historical equipment fault information includes whether the equipment failed, the time of the failure, online monitoring system data before and after the failure, pre-test data before and after the failure, generator equipment parameters, fault waveform, protection action status, accident handling measures, generator model and manufacturer, and other data. Equipment failure status includes whether the equipment failed, the specific fault location (such as the stator, rotor, casing, excitation system, hydrogen cooler, stator cooling water system, hydrogen system, and oil system), and the type of equipment failure (electrical failure, mechanical failure, thermal failure, and system failure). If the cause of the failure is unclear or is caused by multiple factors, the probability of the possible causes or the weighted proportions of the multiple factors should be entered.

[0159] Next, an adversarial network is used to perform data expansion on the historical equipment operation characteristics to obtain the historical operation data of the target equipment, and multiple historical equipment operation characteristics corresponding to the historical operation data of the target equipment are extracted. The specific feature extraction method can be to perform bandpass filtering on the historical operation data of the target equipment to obtain a resonance band, and then perform envelope demodulation analysis on the resonance band to obtain the historical equipment operation characteristics, and then study the fault characteristics of the generator from the historical equipment operation characteristics.

[0160] Furthermore, the convolution kernel and fully connected layer weights of the equipment fault detection model to be trained are adjusted through the cross-entropy loss function, and the stochastic gradient descent method and error back propagation are used to update the parameters of the equipment fault detection model to be trained.

[0161] A self-attention mechanism is introduced into the equipment fault detection model to fuse multiple historical equipment operation features to obtain historical fusion features. This allows the equipment fault detection model to be trained to pay more attention to key features at each scale after repeated iterations, thereby improving the model's feature learning ability. Finally, the equipment fault detection model is trained based on the historical fusion features and historical equipment fault information. The trained equipment fault detection model is used for generator fault diagnosis.

[0162] S200, including:

[0163] S220 , when the device operation status indicator value is outside a preset safety threshold range, extract multiple device operation features from the device operation data, and fuse the multiple device operation features to obtain a fused feature.

[0164] S240 , processing the fused features through the trained equipment fault detection model to obtain equipment fault information of the generator.

[0165] Specifically, when the equipment operation status indicator value is outside the preset safety threshold range, the above-mentioned band-pass filtering method is used to extract multiple equipment operation features of the equipment operation data, which will not be repeated here. The multiple equipment operation features are fused through the attention mechanism to obtain fused features. Finally, the trained equipment fault detection model is used to perform fault detection on the fused features to obtain the equipment fault information of the generator.

[0166] For example, the equipment operation data related to the temperature rise of key parts of the generator, partial discharge and end vibration of the stator winding, short circuit and grounding detection between rotor turns, collector ring system failure, hydrogen system purity and hydrogen leakage, operating parameter deviation problems, etc. can be tested to obtain equipment failure information of the generator.

[0167] In the above embodiment, the equipment fault detection model is trained through methods such as data expansion, feature extraction, and feature fusion, so that the trained equipment fault detection model can be used more accurately for equipment fault detection. Then, the equipment operation data is processed through the trained equipment fault detection model to obtain accurate equipment fault information of the generator.

[0168] In an exemplary embodiment, the process of fault detection for a generator based on device operating data can be divided into five layers: the structure layer, the data layer, the technical support layer, the software assurance layer, and the functional application layer. Taking a steam turbine generator as an example, the specific contents of each layer are as follows:

[0169] The structural layer includes but is not limited to the auxiliary substructures of the turbine generator's rotor, stator and other auxiliary systems; the data layer includes but is not limited to the turbine generator information parameter library, fault data set, diagnostic model library and fault knowledge base; the technical support layer includes but is not limited to data-driven intelligent fault diagnosis technology, digital twin mapping model construction technology and expert intelligent diagnosis system; the software assurance layer includes but is not limited to Unity 3D engine, Solid Works, Python, Matlab, 3dsMax and other tools; the functional application layer includes but is not limited to signal display, fault diagnosis, fault warning and treatment measures and other functions.

[0170] In an exemplary embodiment, the fusion features are processed by the trained equipment fault detection model to obtain equipment fault information of the generator, including:

[0171] The fusion features are processed through the trained equipment fault detection model to obtain the initial equipment fault information of the generator, wherein the initial equipment fault information includes the first equipment fault type, and the second equipment fault type of the generator is obtained. The second equipment fault type is the theoretical equipment fault type of the generator uploaded by the target terminal; according to the first equipment fault type and the second equipment fault type, the target equipment fault type is generated; based on the target equipment fault type, the initial equipment fault information is updated to obtain the equipment fault information of the generator.

[0172] Specifically, the trained equipment fault detection model is first used to process the fused features to obtain the generator's initial equipment fault information. From this initial equipment fault information, the generator's first equipment fault type is determined. If the first equipment fault type is unclear or there are multiple first equipment fault types, the first equipment fault types are ranked by probability or impact weight, either from high to low or from low to high.

[0173] Secondly, the first device fault type is pushed to the target terminal. The target terminal can be a terminal held by an expert in this field or a terminal held by other staff members. The holder of the target terminal uses the first device fault type as a reference and performs on-site fault diagnosis on the generator to obtain the second device fault type, and uploads the second device fault type to the local terminal. Similarly, if the second device fault type is not clear or there are multiple second device fault types, the second device fault types can be sorted in order of probability or impact weight. The sorting method can be from high to low or from low to high.

[0174] Finally, the local terminal generates a target device failure type based on the first device failure type and the second device failure type. For example, if the first device failure type includes: failure type A - 80% probability, failure type B - 20% probability, and the second device failure type includes: failure type B - 60% probability, failure type A - 40% probability, then at this time, since the second device failure type is uploaded by the target terminal and the first device failure type is automatically detected by the model, and the automatic detection of the model may have certain errors, the target device failure type can be determined to be failure type B. For another example, if the first device failure type includes: failure type A - 70% probability, failure type B - 30% probability, and the second device failure type includes: failure type A - 60% probability, failure type B - 40% probability, then the target device failure type is determined to be failure type A.

[0175] Furthermore, based on the first device failure type and the second device failure type, the target device failure type is generated. It can also be determined based on the probability of the first device failure type and the probability of the second device failure type. For example, if the first device failure type includes: failure type A - 70% probability, failure type B - 30% probability, and the second device failure type includes: failure type A - 60% probability, failure type B - 40% probability, then the maximum failure type probability is 70%, and the fault type of the generator is considered to be fault type A.

[0176] In addition, when the fault type of the first device differs greatly from the fault type of the second device, a third party organization may be requested to perform fault diagnosis to obtain a more accurate fault type of the target device.

[0177] Furthermore, the target device fault type and the corresponding device operating data can be used as new training data for the device fault detection model to iteratively train the model, resulting in a more accurate device fault detection model. Furthermore, device fault information and corresponding treatment measures can be generated and entered into the system fault information database, supplementing and improving the system fault diagnosis rules and knowledge base.

[0178] Finally, the first device failure type in the initial device failure information is updated to the target device failure type to obtain updated device failure information.

[0179] In the above embodiment, by combining the second equipment fault type of the generator uploaded by the target terminal with the first equipment fault type of the generator automatically diagnosed by the equipment fault detection model, the equipment fault type of the generator is accurately evaluated from both automatic diagnosis and manual diagnosis aspects to obtain the accurate target equipment fault type of the generator.

[0180] In an exemplary embodiment, Figure 4 As shown, the following will take the steam turbine generator as an example to explain in detail the fault repair method of the steam turbine generator set, including:

[0181] 1. Failure mechanism analysis.

[0182] The main function of this module is to monitor the operating status index values of the existing generator set in DCS (thermal parameter values: stator winding coil temperature, winding bar interlayer temperature and core temperature, hydrogen purity, pressure, hydrogen leakage, collector ring temperature, stator cooling water temperature, pH value, collector ring temperature, etc.; electrical parameter values: stator partial discharge, rotor interturn short circuit, shaft voltage, shaft current, host short circuit and discharge, etc.; mechanical parameter values: generator rotor shaft vibration, seat vibration and stator end vibration, etc.; environmental parameter values: humidity, leakage detection, etc.) and online monitoring system. The operating status indicator values (rotor inter-turn short circuit, stator partial discharge, collector ring temperature, carbon brush current and temperature, etc.) collected in real time by the system, full life cycle test data (factory test data, handover test data, pre-test test data, technical specifications and test cycle, important component replacement time and cycle, etc.), fault cases and special situation data (start-up and shutdown parameters, parameters during fault anomalies), and other operating status indicator values are all used as steam turbine generator set operating data, so as to analyze the fault mechanism of the steam turbine generator based on the steam turbine generator set operating data.

[0183] 2. Model establishment.

[0184] Establishment of the fault evolution model: Based on dynamic and electromagnetic equations, an initial fault evolution model of the steam turbine generator set is established, characterized by "stress load-electromagnetic characteristics-fatigue damage". Specifically, the historical operating data of the first generator set of the steam turbine generator set under frequent start-stop conditions is obtained, and multi-field coupling is performed under frequent start-stop conditions to improve the failure mechanism model. The historical operating data of the first generator set is continuously updated and modulated into the performance degradation process through particle filtering, considering the dynamic evolution of the degradation state of the steam turbine generator set under frequent start-stop conditions. By adding noise to the state transition equation and observation equation of the steam turbine generator set performance degradation model to simulate the uncertainty of system operating conditions and external loads, the fault evolution model of the target steam turbine generator set is obtained, realizing comprehensive monitoring, data acquisition and intelligent diagnosis.

[0185] Fault Detection Model Establishment: Historical operating data for the second generator set of a steam turbine generator, such as the time of previous faults, online monitoring system data before and after the faults, pre-test data before and after the faults, generator set parameters, fault waveforms, protection action status, and fault diagnosis causes, is entered into a fault database and correlated with the diagnosed fault types and phenomena at the time of the generator fault. Based on this data, an adversarial network is used to augment the missing fault data. The augmented data is then bandpass filtered, and envelope demodulation analysis is performed on the resonance bands to obtain historical generator set operating characteristics. The convolution kernel and fully connected layer weights of the convolutional neural network are adjusted using a cross-entropy loss function. Stochastic gradient descent and error backpropagation are used to update the parameters of the fault detection model. A self-attention mechanism is introduced to ensure that the model focuses on critical features at all scales after repeated iterations, thereby improving the feature learning capability of the fault detection model and enabling intelligent fault diagnosis of steam turbine generator sets.

[0186] 3. Fault warning.

[0187] Based on the research on the operating parameter characteristics of steam turbine generator sets under deep peak regulation state, an operating parameter deviation early warning method for steam turbine generator sets is established. When the generator operating parameters exceed the preset threshold value, an alarm can be issued in time. In addition, when individual parameters produce unexpected deviations due to measurement point problems, intelligent analysis is performed to eliminate the erroneous impact of deviation data on calculations, and timely warnings are given for abnormal parameter deviations to ensure the accuracy of generator operating parameters.

[0188] 4. Fault diagnosis.

[0189] 4.1 Fault information detection.

[0190] Extracting turbine generator set operating characteristics from turbine generator set operating data and fusing the turbine generator set operating characteristics; analyzing the fused turbine generator set operating characteristics using a trained fault detection model to obtain initial turbine generator set fault information and a first turbine generator set fault type in the initial turbine generator set fault information; if there is more than one first turbine generator set fault type, the first turbine generator set fault types can be ranked according to the probability corresponding to each fault type to provide a reference for an expert, so that the expert can determine a second turbine generator set fault type for the turbine generator set. The first turbine generator set fault type output by the fault detection model is then compared with the second turbine generator set fault type determined by the expert to generate a target turbine generator set fault type. Based on the target turbine generator set fault type, the initial turbine generator set fault information is updated to obtain turbine generator set fault information.

[0191] 4.2. Health status assessment.

[0192] (a) The Gaussian cloud model is used to address the uncertainty of the pre-set health status assessment indicators, obtaining indicator degradation information for each pre-set health status assessment indicator. (b) A subjective and objective comprehensive judgment method is used to determine the indicator weights for each pre-set health status assessment indicator. Specifically, the analytic hierarchy process (AHP) and association rule method, which take into account expert opinions, are used to determine the weight of each indicator's impact on the generator's condition. (c) Local operating data of multiple auxiliary substructures within the turbine generator set, monitored online by sensors, is substituted into the above method to calculate the local health status of each auxiliary substructure. (d) The DSMT theory is used to integrate the local health status of each auxiliary substructure to obtain the overall health status of the turbine generator. When the local health status of important auxiliary substructures of the turbine generator set or the overall health status of the turbine generator set is in a cautionary state, an early warning is issued and corresponding treatment suggestions are given.

[0193] 4.3. Remaining life prediction.

[0194] Using the previously established degradation model and combining it with the Bayesian filtering method, the current degradation state of the generator set at the current moment is recursively inferred from the initial degradation state of the generator set in the previous time period. Then, the current degradation state of the generator set, the operating data of the steam turbine generator set, the local health status of multiple subsidiary substructures and the overall health status of the steam turbine generator set are used to complete the Bayesian dynamic update, obtain the optimal posterior estimate of the degradation state, and realize the remaining life prediction of the steam turbine generator set over a long period of time.

[0195] 5. Maintenance.

[0196] By using information such as the remaining life of the steam turbine generator set, equipment fault information, the local health status of multiple subsidiary substructures, and the overall health status of the steam turbine generator set, the power plant is automatically guided to perform fault diagnosis and operation and maintenance technical transformation decisions, and the maintenance priority of the steam turbine generator set is sorted to evaluate the fault repair priority of the steam turbine generator set. In combination with the budget and planned maintenance time of the target object performing the maintenance operation, it is determined whether the target object has time or funds to repair the steam turbine generator set. If the target object has time and funds to repair the steam turbine generator set, then during the maintenance process of the steam turbine generator set, the target maintenance strategy information corresponding to the steam turbine generator set fault information can be queried based on the preset fault maintenance mapping relationship; based on the fault repair priority and the target maintenance strategy information, the steam turbine generator set is repaired. In addition, decision-making recommendations on the location and number of additional measuring points can also be given.

[0197] Based on the above analysis, the equipment failure repair method provided by this application has the following advantages in the field of equipment failure repair:

[0198] (1) The system is based on a model-driven approach to establish a three-layer intelligent condition monitoring and fault diagnosis system consisting of fault warning, diagnosis, and root cause analysis. It integrates an integrated intelligent platform for warning, diagnosis, health assessment, and operation and maintenance decision-making, and intelligently guides power plants in the selection of online monitoring systems, optimization of installation plans, dynamic maintenance, and intelligent management of key spare parts, thereby promoting generator condition maintenance and advanced management.

[0199] (2) The system integrates multi-source online monitoring data and multi-dimensional fault diagnosis technology, integrates electrical, chemical and mechanical characteristic data, and constructs a mutually reinforcing evaluation model. Through mutual verification between data and technical complementarity, it deeply analyzes the equipment operating status and achieves a simultaneous improvement in the diagnostic efficiency and accuracy of generator faults.

[0200] (3) Integrate online real-time data, online non-real-time data, and maintenance test data of the generator, and use big data analysis and intelligent diagnosis technology to achieve comprehensive and integrated perception of equipment status and accurate assessment of equipment health. Through deep data mining and artificial intelligence algorithms, accurate prediction of health trends can be achieved, providing a scientific basis for the transition from power equipment health management to predictive maintenance.

[0201] (4) The system can also use a unified technical architecture to establish a unified remote management platform to achieve function integration, data integration, multi-intelligent terminal access, and multi-authority management, ensuring the convenience and security of operation and maintenance personnel.

[0202] (5) This system has the functions of self-adaptation, self-learning, and self-correction. By continuously collecting generator equipment data, health assessment results, fault diagnosis results, and accident handling plans, as the database sample capacity expands, the relationship between fault causes and equipment data, and health status and equipment data becomes increasingly clear. The accuracy of the data analysis model is improved through big data analysis and machine learning technology.

[0203] This system will combine big data analysis, machine learning technology and on-site personnel experience to form a multi-functional, multi-authority, multi-dimensional data analysis intelligent remote management platform for generators, providing a comprehensive reference for generator equipment production management.

[0204] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0205] Based on the same inventive concept, embodiments of the present application also provide an equipment failure repair device for implementing the aforementioned equipment failure repair method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the equipment failure repair device provided below can be found in the above-described limitations of the equipment failure repair method and will not be further elaborated here.

[0206] In an exemplary embodiment, Figure 5 As shown, a device for repairing equipment failure is provided, comprising: a data acquisition module 100, a fault detection module 200, a health status assessment module 300, a remaining life prediction module 400, a priority determination module 500 and a repair control module 600, wherein:

[0207] The data acquisition module 100 is used to acquire the equipment operation data of the generator, wherein the equipment operation data includes the equipment operation status indicator value;

[0208] The fault detection module 200 is used to detect the fault of the generator based on the equipment operation data and obtain equipment fault information when the equipment operation status indicator value is outside the preset safety threshold range;

[0209] A health status assessment module 300 is configured to detect the local health status of multiple auxiliary substructures in the generator based on the equipment operation data, and detect the overall health status of the generator based on the local health status of the multiple auxiliary substructures;

[0210] The remaining life prediction module 400 is used to predict the remaining life of the generator based on the equipment operation data, the local health status of multiple auxiliary substructures and the overall health status of the generator;

[0211] A priority determination module 500 is configured to evaluate the fault repair priority of the generator based on the remaining life of the device, the device fault information, the local health status of multiple attached substructures, and the overall health status of the generator;

[0212] The maintenance control module 600 is used to perform fault maintenance on the generator according to the equipment fault information and the fault maintenance priority.

[0213] In one embodiment, the health status assessment module 300 is also used to extract local operating data of multiple subsidiary substructures in the generator from the equipment operating data; for each subsidiary substructure, based on the local operating data, generate indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators, and based on all indicator degradation information, detect the local health status of the subsidiary substructure.

[0214] In one embodiment, the health status assessment module 300 is further used to obtain first weights and second weights of multiple preset health status assessment indicators, wherein the first weight represents the theoretical importance of the preset health status assessment indicator for assessing the local health status of the accessory substructure, and the second weight represents the actual importance of the preset health status assessment indicator for assessing the local health status of the accessory substructure; for each preset health status assessment indicator, an indicator weight is generated according to the first weight and the second weight; and based on the indicator weights and indicator degradation information of all preset health status assessment indicators, the local health status of the accessory substructure is detected.

[0215] In one embodiment, the remaining life prediction module 400 is also used to determine an equipment failure evolution model of the generator, wherein the equipment failure evolution model is obtained based on the historical operating data of the first device under the frequent start-stop condition of the generator; the initial degradation state of the generator in the previous time period is obtained, and the initial degradation state is processed through the equipment failure evolution model to generate the current degradation state of the generator in the current time period; based on the current degradation state, equipment operation data, the local health status of multiple subsidiary substructures and the overall health status of the generator, the remaining equipment life of the generator is predicted.

[0216] In one embodiment, the equipment fault repair device also includes an equipment maintenance evaluation module, which is used to obtain the maintenance resources available and the planned maintenance time of the target object, wherein the target object is an object for performing fault repair on the generator; based on the maintenance resources available and the planned maintenance time, the number of repairable equipment of the target object is determined, and the fault repair priority range corresponding to the number of repairable equipment is determined; when the fault repair priority is outside the fault repair priority range, an equipment maintenance result of the generator is generated, wherein the equipment maintenance result indicates that the fault repair is not performed on the generator.

[0217] In one embodiment, the equipment fault maintenance device also includes a fault detection model training module, which is used to obtain the historical operation data of the second device of the generator and the historical equipment fault information corresponding to the second device historical operation data; perform data expansion on the historical operation data of the second device to obtain the historical operation data of the target device; extract multiple historical equipment operation features corresponding to the historical operation data of the target device, and fuse the multiple historical equipment operation features to obtain historical fusion features; train the equipment fault detection model based on the historical fusion features and the historical equipment fault information; the fault detection module 200 is also used to extract multiple equipment operation features of the equipment operation data, and fuse the multiple equipment operation features to obtain fusion features; process the fusion features through the trained equipment fault detection model to obtain the equipment fault information of the generator.

[0218] In one embodiment, the fault detection module 200 is also used to process the fusion features through the trained equipment fault detection model to obtain the initial equipment fault information of the generator, wherein the initial equipment fault information includes the first equipment fault type; obtain the second equipment fault type of the generator, the second equipment fault type is the theoretical equipment fault type of the generator uploaded by the target terminal; generate the target equipment fault type according to the first equipment fault type and the second equipment fault type; based on the target equipment fault type, update the initial equipment fault information to obtain the equipment fault information of the generator.

[0219] Each module in the aforementioned equipment fault repair apparatus may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor within a computer device in the form of hardware, or may be stored in a memory within the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0220] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 6As shown. The computer device includes a processor, memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals via wired or wireless means, and the wireless means can be implemented via Wi-Fi, a mobile cellular network, near-field communication (NFC), or other technologies. When executed by the processor, the computer program implements a method for repairing equipment faults. The display unit of the computer device is used to form a visually visible image, and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0221] Those skilled in the art will understand that Figure 6 The structure shown in the figure is a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0222] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0223] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0224] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0225] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0226] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0227] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for repairing equipment failure, characterized in that: The method comprises: Acquiring equipment operating data of the generator, wherein the equipment operating data includes an equipment operating status indicator value; When the device operating status indicator value is outside a preset safety threshold range, performing fault detection on the generator based on the device operating data to obtain device fault information; detecting local health status of a plurality of subsidiary substructures in the generator based on the equipment operation data, and detecting an overall health status of the generator based on the local health status of the plurality of subsidiary substructures; predicting the remaining life of the generator based on the equipment operation data, the local health status of the plurality of attached substructures, and the overall health status of the generator; evaluating a fault repair priority of the generator based on the remaining life of the device, the device fault information, the local health status of the plurality of subsidiary substructures, and the overall health status of the generator; Perform fault repair on the generator according to the equipment fault information and the fault repair priority.

2. The method according to claim 1, characterized in that The detecting, based on the equipment operation data, the local health status of the plurality of subsidiary substructures in the generator comprises: extracting local operation data of a plurality of subsidiary substructures in the generator from the equipment operation data; For each of the subsidiary substructures, indicator degradation information of the subsidiary substructure under multiple preset health status assessment indicators is generated according to the local operation data, and the local health status of the subsidiary substructure is detected based on all the indicator degradation information.

3. The method according to claim 2, characterized in that The detecting the local health status of the auxiliary substructure based on all the indicator degradation information includes: Obtaining first and second weights of the plurality of preset health status assessment indicators, wherein the first weight represents a theoretical importance of the preset health status assessment indicator for assessing the local health status of the accessory substructure, and the second weight represents an actual importance of the preset health status assessment indicator for assessing the local health status of the accessory substructure; For each of the preset health status assessment indicators, generating an indicator weight according to the first weight and the second weight; Based on the indicator weights and indicator degradation information of all the preset health status assessment indicators, the local health status of the auxiliary substructure is detected.

4. The method according to claim 1, wherein The predicting of the remaining life of the generator based on the equipment operation data, the local health status of the plurality of auxiliary substructures, and the overall health status of the generator includes: Determining an equipment fault evolution model of the generator, wherein the equipment fault evolution model is obtained based on historical operating data of a first equipment under a frequent start-stop condition of the generator; Acquiring an initial degradation state of the generator in a previous time period, and processing the initial degradation state using the equipment fault evolution model to generate a current degradation state of the generator in a current time period; The remaining life of the generator is predicted based on the current degradation state, the equipment operation data, the local health states of the plurality of subsidiary substructures, and the overall health state of the generator.

5. The method according to claim 1, wherein After evaluating the fault repair priority of the generator based on the remaining life of the device, the device fault information, the local health status of the plurality of subsidiary substructures, and the overall health status of the generator, the method includes: Obtaining available resources and planned maintenance duration for a target object, wherein the target object is an object for which fault repair of the generator is to be performed; Determine the number of repairable devices of the target object based on the available resources for repair and the planned maintenance duration, and determine a range of fault repair priorities corresponding to the number of repairable devices; When the fault repair priority is outside the fault repair priority range, an equipment repair result of the generator is generated, wherein the equipment repair result indicates that the fault repair of the generator is not performed.

6. The method according to claim 1, characterized in that Before performing fault detection on the generator based on the device operation data to obtain device fault information, the method further includes: Acquire historical operation data of a second device of the generator and historical device fault information corresponding to the historical operation data of the second device; Performing data expansion on the historical operation data of the second device to obtain historical operation data of the target device; Extracting multiple historical device operation features corresponding to the historical operation data of the target device, and fusing the multiple historical device operation features to obtain a historical fusion feature; Training a device fault detection model based on the historical fusion features and the historical device fault information; The performing fault detection on the generator based on the equipment operation data to obtain equipment fault information includes: extracting multiple device operation features from the device operation data, and fusing the multiple device operation features to obtain a fused feature; The fusion features are processed by the trained equipment fault detection model to obtain equipment fault information of the generator.

7. The method according to claim 6, characterized in that The trained equipment fault detection model processes the fusion features to obtain equipment fault information of the generator, including: Processing the fused features using a trained equipment fault detection model to obtain initial equipment fault information of the generator, wherein the initial equipment fault information includes a first equipment fault type; Acquire a second device fault type of the generator, where the second device fault type is a theoretical device fault type of the generator uploaded by a target terminal; generating a target device failure type according to the first device failure type and the second device failure type; Based on the target device failure type, the initial device failure information is updated to obtain device failure information of the generator.

8. An equipment failure repair device, characterized in that: The device comprises: A data acquisition module, configured to acquire the device operation data of the generator, wherein the device operation data includes a device operation status indicator value; a fault detection module, configured to perform fault detection on the generator based on the equipment operation data to obtain equipment fault information when the equipment operation status indicator value is outside a preset safety threshold range; a health status assessment module, configured to detect the local health status of a plurality of subsidiary substructures in the generator based on the equipment operation data, and detect the overall health status of the generator based on the local health status of the plurality of subsidiary substructures; a remaining life prediction module, configured to predict the remaining life of the generator based on the equipment operation data, the local health status of the plurality of subsidiary substructures, and the overall health status of the generator; a priority determination module, configured to evaluate a fault repair priority of the generator based on the remaining life of the device, the device fault information, the local health status of the plurality of subsidiary substructures, and the overall health status of the generator; A maintenance control module is used to perform fault maintenance on the generator according to the equipment fault information and the fault maintenance priority.

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 any one of claims 1 to 7 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 any one of claims 1 to 7 are implemented.

Images