356 results about "Failure data" patented technology

The invention relates to a power supply switching control optimization method and system, and the method comprises the steps: constructing a multi-dimensional data fusion monitoring system, integrating a real-time voltage ripple coefficient detection module of a main power supply, a load transient response analysis module and an environment temperature and humidity sensor array, and generating a dynamic health degree index; the method comprises the following steps: establishing a fault prediction model based on transfer learning, pre-training through multi-scene failure data in a historical database, receiving a health degree index in real time, outputting a reliability attenuation curve, and activating a dual verification mechanism when the slope of the attenuation curve exceeds a preset threshold; extracting characteristic energy distribution of a frequency band of 0.1-10 kHz; executing a transient load impact test, and measuring dynamic response parameters of voltage recovery time and overshoot amplitude; and generating a switching confidence score based on cross processing of the dual verification result and the attenuation curve, and triggering a hierarchical switching instruction. The accuracy of the switching time of the main power supply and the standby power supply can be improved.

The invention relates to the technical field of strength detection, and discloses radiator strength detection equipment and a detection method. The method comprises the following steps: carrying out adaptive wavelet denoising and structural association mapping on collected multi-scale structural feature data and multi-environment coupling material feature data of a radiator to be detected, and carrying out multi-dimensional feature extraction, space-time alignment and interactive mapping on an obtained fusion data set to obtain a multi-physical field coupling data matrix; carrying out physical field control equation set separation iteration solution and heat dissipation cyclic load and environmental action calculation on the coupling data matrix to obtain a damage evolution state; and performing multi-dimensional damage sensitive area identification based on the damage evolution state, generating a risk distribution map, performing failure comparison and life calculation based on historical failure data of the same type of radiators, and generating an intensity comprehensive detection result. Through multi-physical field coupling analysis and multi-dimensional damage evaluation, accurate detection of the strength state of the radiator is realized, and the accuracy of structural reliability evaluation is improved.

The invention belongs to the technical field of bearing fault diagnosis, and particularly relates to a transmission casing typical small bearing fault weak vibration signal extraction method, which comprises the following steps: reconstructing a bearing vibration signal to obtain a modulation impact signal, and calculating a weighted kurtosis value based on the modulation impact signal to improve a Protugram algorithm; noise interference is added to a bearing fault signal, and extraction of a small bearing fault weak impact signal is realized by optimizing a filtering frequency band based on a weighted kurtosis value improved Protugram algorithm; bearing impact characteristics can be enhanced based on a simulation sensor resonance enhancement algorithm (SPM); the impact quantification of the bearing can eliminate the influence of different rotating speeds and load working conditions, and realizes the quantitative diagnosis of bearing faults. And finally, the fault data of the rolling body, the inner ring and the outer ring of the flight attachment case small bearing part test bed are verified, and effective extraction and enhancement of bearing fault weak signals can be realized.

The invention discloses a property maintenance prediction system based on big data, and the system comprises the steps: monitoring equipment and environment in real time through an intelligent sensor, transmitting the data to a cloud or local storage through a wireless network, and carrying out the historical data processing and cleaning through a distributed architecture; predicting the equipment fault risk through machine learning, establishing a customized prediction model for different equipment, analyzing the life cycle and predicting the service life of the equipment; for old equipment, constructing a life extension model by using historical fault data; the system generates intelligent decision suggestions, pushes equipment health states to property management personnel, automatically formulates maintenance plans and priorities, and dispatches maintenance personnel; during fault prediction, early warning is given out in real time, related personnel are reminded through a short message or an APP, meanwhile, maintenance operation is recorded, historical data are analyzed, and future maintenance plans and budget are optimized; the system dynamically adjusts the life prediction model of the old equipment by accumulating historical data, and ensures the accuracy and efficiency of equipment health management.

The invention discloses an automatic driving test scene set optimization method, apparatus and device, and a storage medium. The method comprises the steps of generating a simulation scene file in an OpenSCENARIO format through preprocessing data; analyzing the simulation scene file through a teacher model, outputting a risk description text, receiving a scene feature vector and the risk description text through a student model, and outputting a failure probability; determining a comprehensive value index according to the failure probability, dynamically updating a scene library of automatic driving test scenes, collecting failure data in an AUT test, performing incremental fine tuning on the student model, and obtaining an optimized target test scene set; the problem of gradient estimation variance explosion caused by sparseness disasters can be effectively solved, the stability of model training is improved, the recognition capability of a rare long-tail failure scene is enhanced, the accuracy of failure probability prediction is improved, the judgment accuracy of the model to a boundary scene is improved, and the accuracy and comprehensiveness of scene value quantification are improved.

The invention relates to the technical field of PVD coating, in particular to a PVD coating design quality control method based on machine learning. The method comprises the following steps: obtaining a coating substrate object, carrying out PVD coating design, and carrying out substrate PVD coating simulation based on the design to obtain related simulation data; analyzing mechanical property parameters by using the PVD coating simulation data of the base material, evaluating the mechanical property of the coating material in combination with the mechanical data, and generating dynamic stress response data of the coating layer through a stability test; evaluating based on the chemical stability of the base material coating film, and predicting the failure condition of the coating film layer by combining the dynamic stress response data of the coating film layer; and through comprehensive analysis of dynamic stress response data and failure data of the coating layer, the coating design is further optimized. According to the invention, through the quality control of PVD coating, the quality of the PVD coating is more perfect.

The invention discloses a life prediction method and device for an accelerated test product and electronic equipment, and the method carries out typical failure mechanism analysis based on fault data, so as to determine a key device and sensitive stress of a target product. And carrying out clustering analysis on the fault data related to the key device, and determining the failure contribution rate of each sensitive stress to the target product. And according to the failure contribution rate corresponding to each sensitive stress, carrying out a multi-stress accelerated life test of the target product. And training the neural network model by adopting test data of the multi-stress accelerated life test. The accelerated test life of a target product under the target stress condition is predicted by using the trained neural network model, life prediction of the accelerated test product is realized, the neural network model is trained by using a multi-stress accelerated life test based on fault physics, the influence of the coupling effect between environmental stresses on the product is emphasized, and the service life prediction of the accelerated test product is realized. Therefore, the model better conforms to the actual application process of the product, and the life prediction precision is improved.

A hotspot data elimination and maintenance method in a distributed cache system comprises the steps that the access frequency, the time locality score, the service weight and the real-time data popularity value of data are acquired, and the dynamic priority score of the data is generated; determining failure data based on the dynamic priority score; dividing the data into hot spot area data, temperature spot area data and cold spot area data, when the cache space is insufficient, preferentially eliminating the data with the lowest dynamic priority score and the retention benefit value lower than a preset threshold value in the cold spot area data, and if the cache space is still insufficient, performing data elimination on the temperature spot area data according to the latest access time of the temperature spot area data; packaging the cold point area elimination data, the warm point area elimination data and the failure data into a final transaction message; and broadcasting the final transaction message through a preset protocol, and performing distributed consistency maintenance operation on the final transaction message through the receiving node. According to the invention, intelligent and efficient cache data elimination and consistency maintenance management can be realized.

The invention is suitable for the field of equipment fault prediction, and provides an underground equipment multi-source fault prediction method and system based on a large model, and the method comprises the following steps: synchronously collecting an acoustic signal, a visual image, a vibration signal, environment data and working condition parameters of underground equipment in real time, and integrating the acoustic signal, the visual image, the vibration signal, the environment data and the working condition parameters into a multi-modal time sequence data stream; performing time alignment and cleaning on the multi-modal time sequence data stream, extracting deep features of each modal, and generating a multi-source fusion feature vector; inputting the multi-source fusion feature vector into a preset prediction model to obtain prediction information of the underground equipment, and generating fault traceability information; continuously optimizing parameters of the prediction model based on fault data collected in real time; and dynamically displaying prediction and traceability results, and automatically generating a maintenance strategy and an equipment scheduling scheme based on the results. According to the invention, panoramic perception of the operation state of the equipment and accurate capture of early weak fault features are realized, and the comprehensiveness and accuracy of fault detection are greatly improved.

The invention discloses a dynamic correction method and system for equipment fault mode occurrence degree, and belongs to the technical field of power plant electrical overhaul and maintenance, and the method comprises the steps: collecting operation parameters, maintenance records and historical fault data of target electrical equipment, and recognizing a potential fault mode set of the target electrical equipment; failure mode and influence analysis is carried out on each failure mode, and the initial occurrence degree, severity and detectability of each failure mode are determined; calculating an initial risk priority number of each fault mode, and determining an initial risk according to a preset risk grade division rule; dynamically adjusting the initial occurrence degree, recalculating the updated risk priority number, and generating a corrected risk level; and constructing a dynamic maintenance decision model, and outputting a maintenance priority and a maintenance time window for each fault mode. According to the method, the defects of real-time performance, comprehensiveness and flexibility of electrical operation and maintenance of the power plant are overcome, and efficient and accurate data support is provided for operation and maintenance of electrical equipment.

The invention discloses a nondestructive testing device and a testing method for heterogeneous parts of engineering machinery. The method comprises the following steps: acquiring material attributes and historical failure data of a heterogeneous component, extracting defect form and position feature parameters, constructing a statistical distribution model and generating a typical feature set; according to the statistical distribution model and the material attributes, generating a virtual defect data set containing form and position features by adopting numerical simulation; performing finite element analysis on the virtual defect data set, calculating a stress concentration effect and an expansion trend, and outputting a defect harmfulness quantitative index; and if the harmfulness index exceeds a preset threshold value, marking the defect as a high-risk defect, and generating a preferential simulation target set and the like. According to the method, all links of nondestructive testing are covered from data acquisition to standard construction, a set of complete working process is formed, and the nondestructive testing problem of the heterogeneous parts of the engineering machinery can be systematically solved.

The invention discloses a data transmission method and computer equipment, and relates to the technical field of data transmission. The method comprises the following steps: acquiring a sending work request containing a local sending queue pair number and a work queue element; querying target sending information by taking a local sending queue number as an index; storing the working queue elements and the target sending information as operation contexts into an internal memory of a network card of a request end, and recording corresponding indexes; assembling a target data packet based on the target sending information and the memory index; receiving a feedback data packet of the target data packet and extracting a memory index and a serial number carried by the feedback data packet; if the feedback data packet is a correct confirmation data packet, judging that the sending of the work request is completed and releasing the context; and if the feedback data packet is a confirmation failure data packet, calculating a retransmission data initial address and organizing retransmission. The technical problem that the effective connection number of network cards is limited by hardware resources in the prior art is solved, and the technical effect of supporting large-scale network parallel communication is achieved.

The invention discloses a high-speed rail vehicle body aluminum alloy material corrosion fatigue performance simulation method and system, and belongs to the technical field of material corrosion fatigue testing. The method comprises the following steps: acquiring multi-source data of a high-speed rail body aluminum alloy material on an actual service route; establishing a material corrosion conversion coefficient database; obtaining an equivalent mapping relation based on the service environment and the load input set; generating a laboratory equivalent accelerated environment spectrum; and executing the laboratory equivalent accelerated environment spectrum, obtaining material failure data, and if the comparison difference is not less than a preset threshold value, correcting the calculation parameters of the equivalent mapping relation and regenerating the spectrum system. Through synchronous acquisition and fusion analysis of the multi-source environmental data and the mechanical load and in combination with the corrosion conversion coefficient database construction and closed-loop verification optimization method, the problem of test result distortion caused by separation of a laboratory simulation environment and an actual service condition in the prior art is solved.

The invention relates to the technical field of building construction, in particular to a prefabricated building construction progress automatic monitoring method. The method comprises the following steps: obtaining a building construction design drawing; based on the building construction design, prefabricated building component design data is extracted; building a building construction three-dimensional model based on the prefabricated building component design data; building component transportation simulation is conducted according to the building construction three-dimensional model, and building component transportation data are obtained; identifying a component cracking degree based on the building component transportation data; performing connection failure analysis based on the component cracking degree to obtain component connection failure data; predicting a reinforcement corrosion risk based on the component connection failure data; evaluating assembly durability based on the reinforcement corrosion risk; and extracting hoisting design data of the building construction design drawing. Based on the building construction technology, the efficiency and accuracy of prefabricated building construction progress monitoring are improved, and the accuracy of construction plan execution and the resource scheduling efficiency are improved.

The invention discloses an insulation residual life prediction algorithm and device based on real-time temperature acquisition and a medium, and relates to the technical field of new energy automobiles, and the method mainly comprises the steps: constructing an offline insulation failure database through an accelerated thermal aging experiment of a twisted pair sample, and recording the PDIV of a target motor at a brand new state room temperature as an initial PDIV0; based on the off-line insulation failure database, obtaining the aging slope of the PDIV decaying along with the aging time at each motor winding temperature and a corresponding aging curve; according to the mapping relation between the PDIV and the motor winding temperature, obtaining an insulation threshold voltage corresponding to the current motor winding temperature under the correction of a temperature correction coefficient; and based on the current PDIV0 and the insulation threshold voltage, substituting the aging curve corresponding to the current motor winding temperature to obtain and output the difference value of the corresponding insulation failure time of the two as the insulation residual life. The method can accurately capture the initial characteristics of insulation aging, the early warning time is advanced, and the timeliness is improved.

The invention provides a low-power-consumption embedded data recording system and algorithm for on-line monitoring of SF density. The system is provided with an MCU main control unit, an external independent watchdog circuit, a LoRa wireless communication module, an RS485 interface, an SF density and temperature measurement sensor, a battery power supply and voltage acquisition module and a Flash data storage and real-time clock module. Algorithms of the system comprise an intelligent data recording strategy, a dynamic sensor activation and sleep mechanism, a power failure data protection and breakpoint continuation mechanism, a multi-parameter compression coding algorithm, a dynamic sampling frequency adjustment strategy, a real-time clock synchronization and timestamp calibration mechanism, a multi-sensor fusion sensing mechanism and an ultra-low power consumption system architecture mechanism. The low-power-consumption embedded data recording system and algorithm for on-line monitoring of the SF density have the functions of multi-data change trend intelligent judgment, multi-data fusion perception judgment, ultra-low power consumption and flexible communication, and are suitable for long-term field deployment.

The invention discloses an intelligent switching method and system based on multi-platform overseas payment, and the method comprises the following steps: accessing a plurality of overseas third-party payment platforms, and building a unified order processing flow through a transaction service; a monitoring point is embedded in the payment interaction process of the application program and the transaction service, payment success data and payment failure data are collected, and corresponding counters are maintained respectively; regularly pulling the counter in a parallel task mode, and counting the failure rate of each payment mode in the first time window T1; when it is detected that the failure rates of a certain payment mode in continuous N first time windows T1 all exceed a threshold theta, triggering exception handling, including marking the payment mode as abnormal, and automatically switching to a standby payment mode; and when it is detected that the failure rate of the off-shelf payment mode in the second time window T2 is recovered to be below the threshold value, automatically executing recovery operation, including re-shelving the payment mode.

The invention discloses a safety valve service life evaluation method based on pressure history statistics and dynamic damage accumulation. According to the method, only safety valve inlet pressure time sequence data is needed, a load cycle is extracted through an improved rain flow counting method, and the nonlinear damage increment is calculated according to an S-N curve model dynamically degraded along with the accumulated damage degree. According to the method, a self-adaptive failure probability threshold is established based on historical failure data, and probabilistic prediction of the residual life is realized. The model is clear in physical significance, only needs a single data source, significantly reduces the implementation complexity and cost, greatly improves the precision of life prediction by considering the degradation of material performance, and is especially suitable for predicting the accelerated failure risk of the safety valve in the later period of long-term service.

The invention relates to the technical field of data mining, in particular to a mining analysis method and system for vehicle wire harness fault data, and the method comprises the following steps: extracting a maintenance timestamp and a path segment index to construct a two-dimensional coding matrix, carrying out the segment division of a vehicle operation mileage, generating a total number of segment faults and a probability sequence, and carrying out the mining analysis. And executing logic comparison according to the serial numbers of the node major loop and the branch line, configuring a path weight parameter, executing tensor operation on the probability sequence and the weight parameter, obtaining an aggregation degree parameter, and constructing a mining analysis result. According to the method, a data multi-dimensional space topological structure is remodeled, dynamic recession mapping is calculated to capture a deep derivation relation, a probability sequence and a weight parameter are input into a sensing network to execute feature extraction, an aggregation degree index is obtained, and according to feature collection operation, space associated conduction paths of wire harness state degradation at different positions are deeply mined. And a fault data mining analysis system with a network deep topology feature analysis capability is formed.

The invention discloses a ship mechanical and electrical product accelerated life test stress profile generation method and device, a medium and a product, and relates to the field of ship reliability engineering and test.The method comprises the steps that working condition profile data and marine environment stress data of a target ship mechanical and electrical product under a typical navigation task are obtained, and a working condition-environment mapping model is established; according to the designed life target of the target ship mechanical and electrical product, the whole life cycle total damage to the target ship mechanical and electrical product is determined by using the working condition-environment mapping model and the failure physical damage model, an accelerated stress profile is generated by strengthening the stress level with damage equivalence as the principle, an accelerated factor is calculated, and an accelerated life test is performed; and failure data of the failure event is recorded, parameters of the failure physical damage model are calibrated, and the accelerated stress profile is iteratively optimized by using the calibrated failure physical damage model. Accurate and dynamic generation of the accelerated life test stress profile of the ship mechanical and electrical product is realized.

The invention belongs to the technical field of reliability analysis, and discloses an electric energy meter adaptive residual life prediction method and system based on random degradation modeling, and the method comprises the steps: obtaining degradation data of an electric energy meter through an accelerated degradation test, removing failure data, and calculating a drift coefficient and a diffusion coefficient through a maximum likelihood estimation method; establishing a multi-source data fusion model of the drift coefficient and the temperature and humidity stress in combination with historical data, and fitting model parameters; expanding the model to a normal-temperature and normal-humidity condition, and calculating a corresponding drift coefficient; and based on the improved random degradation model, deriving a probability density function, an accumulative distribution function and a reliability function, and predicting the residual life of the electric energy meter at normal temperature and normal humidity. The system comprises a temperature and humidity test box, a power source table body and an upper computer, and is used for automatically collecting and analyzing data. Through multi-source data fusion and random degradation modeling, the prediction accuracy is improved, the test cost and time are reduced, and the method is suitable for electric energy meter reliability evaluation under complex working conditions.

The invention relates to the technical field of data synchronization, in particular to a financial data synchronization method and system, and the method comprises the steps: checking factor packaging: extracting original financial data, adding checking factors to generate a checking packaging unit, and enabling the checking factors to comprise data structure fingerprints, articulation relation codes and environment dependence parameters; dynamic integrity verification: inputting the verification packaging unit into a dynamic verification engine, loading a corresponding dynamic verification matrix to carry out data integrity verification, and outputting a verification passing unit which passes the verification; adaptive feedback regulation and control: performing attribution analysis and strategy feedback on an abnormal mode in a verification failure unit to realize system internal control closed-loop optimization; and exception handling writing: executing an exception handling tree-based handling operation on verification failure data, writing a verification passing unit into a target financial system, and generating an audit tracking chain. According to the method, each piece of write-in data is ensured to have traceability and compliance, and the intelligence, self-healing property and audit credibility of the financial data synchronization process are enhanced.

The invention provides a communication method which is used for clearing invalid data in time. The communication device receives the first data packet and the second data packet, and discards the first data packet according to the second data packet. In the application, the communication device can clear the failure data according to the received data packet, and does not need to clear the failure data after the timer corresponding to the failure data is overtime. And the problem that the transmission of the valid data increases the time delay of the transmission of the valid data is avoided.

The invention discloses a film capacitor life intelligent prediction method fusing multi-dimensional data, and relates to the technical field of capacitor reliability. According to the method, a multi-dimensional data fusion model is constructed, and three types of core data are creatively integrated: 1, process parameters of the whole manufacturing process of the thin-film capacitor; 2, operation condition data is monitored in real time; and 3, a large-batch historical failure database based on acceleration factors. And dynamically associating the multi-source heterogeneous data by adopting a machine learning algorithm, establishing a nonlinear life attenuation model, and realizing early life early warning by quantifying the contribution degree of process defects to electrical performance degradation. Compared with a traditional method, the method has the advantages that the prediction precision is improved by more than 40%, progressive failure caused by metal layer micro-scratch, self-healing effect accumulation and metal spraying contact degradation hidden factors can be identified, and an accurate basis is provided for preventive maintenance of power electronic equipment.

The invention discloses a metro vehicle supervision manufacturing multi-station data traceability control system, and relates to the technical field of industrial data traceability, and the system comprises a data collection module which obtains the original manufacturing data of each station of metro vehicle manufacturing; the feature extraction module is used for processing the original manufacturing data and generating behavior fingerprint vectors and station features; the identity analysis module is used for carrying out feature matching calculation and outputting a candidate component identity set and matching similarity; the logic verification module is used for carrying out logic consistency verification and determining the identity of the target component; and the traceability control module establishes a mapping relation with the original manufacturing data based on the identity of the target component, generates a unique traceability node, and writes the traceability node into a multi-station traceability data chain. According to the method, the problems of identification failure, data link breakage and affiliation errors caused by station switching or equipment sharing are solved, supervision data continuity and identity recognition accuracy are improved, and reliable support is provided for vehicle full-life-cycle data tracing.

The invention relates to an equipment fault rate calculation method and system for an offshore converter station, and the method comprises the following steps: obtaining fault data of offshore converter station equipment in a preset time, and dividing the offshore converter station equipment into components according to functions; performing correlation analysis on the fault time sequence by using an autocorrelation function to obtain a fault type; if the fault is a static fault, constructing a confidence interval by using a large number theorem and normal distribution to obtain an initial fault rate of each component; if the fault is a dynamic fault, obtaining an initial fault rate of each component by using a time sequence analysis model; constructing an environment correction coefficient model, and correcting the initial failure rate of each component to obtain a corrected component failure rate; analyzing each component by using an analytic hierarchy process to obtain a fault weight, and combining the fault weight and the corrected component fault rate to obtain the overall fault rate of the overall equipment of the offshore converter station; and obtaining an annual average fault time expected value according to the overall fault rate.

The invention discloses an intelligent factory data analysis and management system based on digital-intelligent integration, which belongs to the technical field of digital-intelligent integration and comprises a data acquisition and preprocessing module, a situation awareness module, an analog simulation model module, an intelligent analysis module, an intelligent prediction and diagnosis module, a dynamic decision module and a self-adaptive optimization scheduling module. According to the invention, data of the data acquisition and preprocessing module and the context awareness module are organically combined, the time-varying feature matrix is extracted, dynamic evaluation and update selection are carried out on features, and deep influence factors can be mined out when association of production data and environment data is analyzed, so that the production efficiency is improved. Through an abnormal score, a trend prediction result and a virtual simulation model state of an intelligent analysis module, and through calculating a risk factor and a composite risk score, when an equipment fault is predicted, real-time operation data and historical fault data of the equipment and performance simulation in virtual simulation are comprehensively considered, and a potential fault hidden danger is found in advance.

The invention discloses a reliability index distribution method and system suitable for a ship power system. The objective of the invention is to overcome the defects of a traditional ship power system reliability evaluation method in the aspects of dynamic performance, real-time performance and adaptability. According to the method, real-time operation data, including temperature, pressure, vibration frequency, abrasion loss and other key parameters, of all subsystems in a ship power system are collected, and dynamic reliability influence factors are constructed based on the data. By introducing a deep reinforcement learning model, the system can adjust the weight coefficient of each subsystem according to the current dynamic reliability influence factor. In addition, in combination with historical fault data, a fault propagation path is modeled through a graph neural network (GNN), and a fault rate correction factor is calculated, so that the accuracy of reliability distribution is improved. And finally, verifying the rationality of a reliability distribution result through a Monte Carlo simulation method, and dynamically adjusting system parameters according to a simulation result to ensure stable operation of the ship power system.

The invention relates to a method for establishing an SIL dynamic evaluation model based on Weibull distribution, and belongs to the technical field of safety protection systems. The method comprises the following steps: (1) establishing a dynamic cumulative failure probability function model based on Weibull distribution according to obtained failure data through a reliability test; (2) calculating Weibull distribution parameters by using a median rank method, a linear regression fitting method and a least square method; (3) drawing a function model image by using MATLAB software; and (4) in combination with a fault tree method, establishing an SIL dynamic evaluation model based on Weibull distribution. According to the SIL dynamic evaluation model for the final executive component of the safety instrument system based on Weibull distribution, the dynamic change trend of the final executive component PFDavg along with time obtained by the model basically accords with the reality, the accuracy of SIL evaluation is improved, and scientific method support is provided for design personnel of the safety instrument system.

The invention discloses an unmanned aerial vehicle inspection method and system for a wind turbine generator set, and belongs to the technical field of intelligent operation and maintenance of the wind turbine generator set. The method comprises the following steps: establishing a digital inspection task model based on the layout and historical operation fault data of the wind generating set, and generating an initial inspection path scheme; controlling the unmanned aerial vehicle to fly along the path, and collecting multi-source inspection data by using an airborne sensor; carrying out processing and fusion analysis on the data, realizing abnormity identification and fault diagnosis, and outputting a diagnosis result containing a fault type and a severity level; and generating a structured inspection report according to a diagnosis result, and dynamically adjusting a flight path, a hovering position and sensor parameters. According to the invention, intelligent optimization of the inspection path and adaptive control of data acquisition are realized, the inspection efficiency and the data quality are improved, real-time diagnosis and predictive maintenance of faults are supported, the labor cost and the safety risk are effectively reduced, and the method is suitable for intelligent operation and maintenance of a large-scale wind power plant.