320 results about "Anomaly detection algorithm" patented technology

The invention, which belongs to the technical field of intelligent breeding equipment, discloses an underwater bionic robotic fish school cooperative breeding environment dynamic monitoring and regulation system comprising a bionic robotic fish body core module, a group cooperative control core module and an intelligent regulation core module. The bionic robotic fish module simulates a real fish swimming mode and carries various water quality sensors to autonomously cruise, so that interference to cultured fishes is reduced; the group cooperation module utilizes a cluster intelligent algorithm and an underwater acoustic communication technology to realize coordinated movement of multiple robotic fishes and full coverage of a monitoring area; the intelligent adjusting module is based on an abnormal detection algorithm, and integrates a miniature oxygenation device, a pH adjusting device and the like to achieve precise regulation and control of the local environment. By the adoption of the system, dynamic sensing and active regulation and control of the culture environment are achieved, a bionic monitoring regulation and control solution is provided for modern aquaculture through intelligent cooperation of the robot fish school and the management system, and the system has the important value of improving the culture efficiency and improving the growth environment.

The invention relates to a standardized project management and intelligent professional scheduling system, and belongs to the technical field of project management and human resource intelligent scheduling. The project standardization module is disassembled into standardization task steps through a business process association rule mining algorithm, and project complexity and resource tensity are adapted by using a task attribute dynamic weight algorithm; the archive matching module constructs professional archives, extracts feature vectors through a multi-criterion decision-weighted bipartite graph matching algorithm, and generates optimal matching pairs in combination with adaptive weight adjustment and an integer linear programming model; the dynamic scheduling module plans a task execution scheme according to a resource constraint scheduling mechanism, and realizes efficient scheduling in combination with a skill supply and demand prediction algorithm and calendar integration; and the quality optimization module adopts a deliverable anomaly detection algorithm to monitor compliance, feeds back an iterative matching and scheduling strategy through a time sequence prediction optimization algorithm, and perfects a skill map based on a map increment updating algorithm. The system realizes a project full-process closed loop.

The invention discloses an ecological environment detection method and system based on multispectral remote sensing fusion, and relates to remote sensing image processing. The method comprises the following steps: collecting multispectral remote sensing image data of a target area; performing multiband joint atmospheric correction processing on the multispectral remote sensing image according to the scattering coefficient, the atmospheric light value and the transmissivity of each band; performing foreground and background analysis on the corrected multispectral remote sensing image; fusing the vegetation area and the non-vegetation area of each wave band image by adopting different weight strategies to generate a multispectral fusion image; and extracting spectral features of the multispectral fusion image, constructing a standard vegetation spectral feature library, and identifying regions deviating from a standard vegetation spectrum through an anomaly detection algorithm according to the extracted spectral features to obtain various vegetation coverage rates. In view of low vegetation identification precision caused by direct foreground and background division of a multispectral remote sensing image under an atmospheric interference condition, vegetation division is performed after a clear image is obtained, so that the detection precision is improved.

The invention discloses a health management service system and method based on AI optimization. The method comprises the following steps: S1, collecting multi-source health data of a user and constructing a structured health data set; s2, constructing a health state graph containing node attributes, edge connection weights and time indexes; s3, inputting the health state atlas into a linear graph neural network to generate a health state embedded vector; s4, collecting context information of a user, and fusing to generate personalized state perception representation and a health target embedding vector; s5, inputting the improved general value layering model to generate a structured intervention action candidate set; s6, screening and outputting a personalized health intervention scheme based on the matching score; s7, constructing a state evolution sequence, and inputting the state evolution sequence into a semi-supervised anomaly detection algorithm for monitoring and recognition; and S8, when the risk threshold is exceeded, triggering early warning and dynamically adjusting the intervention scheme. According to the method, collaborative optimization of personalized modeling and intelligent intervention strategies is realized, and the method is suitable for health management service scenes.

The invention discloses a dynamic network access control method and system based on a zero-trust architecture, and the method comprises the steps: integrating equipment health degree evaluation through the triple dynamic binding of biological feature dynamic binding, equipment fingerprint salt value hash verification and environmental state perception, and constructing a real-time trust basis; dynamic risk quantification is realized based on a multi-source heterogeneous data fusion machine learning model, real-time upgrading and degrading self-adaptive adjustment of authority is realized through an AI driving strategy generation module according to a real-time risk score, a zero-trust sandbox limitation sensitive operation is triggered for high-risk access, and a minimum authority channel is started for low-risk access; performing fine-grained access control and intercepting an unauthorized request in real time by adopting an agent-free API gateway technology, and monitoring an operation behavior in combination with a block chain non-tampering storage access log and an anomaly detection algorithm; finally, a continuous self-adaptive evolutionary cycle is formed through a risk assessment-policy execution-abnormal feedback closed loop mechanism, and the static lag problem of traditional network access control is systematically solved.

The invention relates to the technical field of cross-border logistics, and discloses a cross-border e-commerce commodity traceability method and system based on a block chain, and the method comprises the following steps: generating a unique traceability code for each cross-border e-commerce commodity, extracting an original information field of the cross-border e-commerce commodity, and calculating an anti-counterfeiting hash value; performing structured packaging on the original information field and the anti-counterfeiting hash value according to a preset template, confirming a transaction through a consensus mechanism, writing the transaction into a block chain, and generating a block hash value; operation information of all logistics nodes is collected, abnormity is recognized through an abnormity detection algorithm, and data fingerprints obtained after abnormity recognition are encrypted and stored; the encrypted and signed operation information is submitted to a block chain network, and a block chain node verifies the validity of a digital signature through a public key of a producer and verifies the legality through a consensus mechanism; and proving an aggregation verification result by using zero knowledge. According to the invention, a complete traceability path is displayed, and full-chain transparent management from production to consumption is realized.

The invention relates to the technical field of safety monitoring, and discloses a multistage safety early warning and linkage processing system for an electricity-hydrogen complementary energy station, and the system comprises a monitoring collection module which is used for deploying a multi-source sensor network in the electricity-hydrogen complementary energy station, collecting key parameter data in real time, and generating a real-time data set comprising a timestamp, a sensor ID and a parameter value; the anomaly detection module is used for performing multi-dimensional analysis by adopting an anomaly detection algorithm according to the real-time data set, identifying an anomaly mode and outputting a graded early warning signal; the evaluation grading module is used for carrying out dynamic risk evaluation in combination with the early warning signal and the system state, and outputting a quantitative risk grade and a disposal suggestion; the linkage triggering module is used for automatically matching and triggering a corresponding plan according to the risk level and outputting a specific linkage control instruction; an execution feedback module; according to the invention, the emergency response speed is improved, the false alarm rate caused by normal working condition fluctuation is reduced, and fundamental conversion from passive response to active early warning is realized.

The invention provides a cross-system fault diagnosis method and system combined with multi-dimensional anomaly detection, and the method comprises the steps: firstly obtaining an original data set which is generated by cross-system call and comprises fragmented call data, peripheral business transaction data and log data, and carrying out the breakpoint supplement processing to generate a complete data set; then, transverse anomaly diagnosis is carried out on the complete data set, a decision tree algorithm is used for positioning a fault propagation path and an abnormal object, then longitudinal anomaly detection is carried out on the abnormal object, operation logs, error stacks and infrastructure indexes are analyzed based on a multi-dimensional anomaly detection algorithm, and an anomaly reason analysis result is generated; and finally, determining a fault source, a fault object and an influence range according to an analysis result, and outputting a cross-system fault diagnosis report, thereby comprehensively utilizing multi-source data, realizing accurate positioning and deep analysis of cross-system faults, improving fault diagnosis efficiency and accuracy, and ensuring stable operation of the system.

The invention relates to the technical field of battery monitoring, and discloses a battery global temperature and pressure distribution monitoring system and method, and the system comprises a distributed sensing unit, a signal processing unit, a data analysis unit, an intelligent decision-making unit, a wireless communication module, and a man-machine interaction terminal. Battery temperature and pressure data are acquired through a high-density flexible sensor array, fine sensing is realized by combining multi-dimensional feature extraction and an anomaly detection algorithm, and a response strategy is generated by using a rule engine; the battery state can be comprehensively monitored, local abnormity can be captured in time, safety and reliability are improved, the method is suitable for the field of power batteries and energy storage systems, the problem that in the prior art, due to insufficient sensor arrangement density, monitoring of the internal state of the battery is not comprehensive is solved, and safety is improved. And the real-time monitoring of the global temperature and pressure distribution of the battery and the intelligent response of the abnormal state are realized.

The invention discloses an asset vulnerability detection method and device, electronic equipment and a storage medium, and relates to the technical field of network security, and the method comprises the steps: actively sending a multi-protocol detection packet to scan a target network segment, and obtaining a first asset set; passively monitoring network traffic to extract asset feature information, and obtaining a second asset set to generate an asset list; port scanning tasks of all assets are dispatched to a plurality of scanning nodes in a distributed and parallel mode, dynamic port scanning is carried out according to a descending order of a plurality of key elements in combination with a port scanning optimization model based on risk prediction, and a full-amount port risk map is constructed; the static layer is matched with known vulnerabilities; the dynamic layer identifies suspicious behaviors deviating from a normal behavior baseline through an anomaly detection algorithm, and obtains an asset vulnerability detection result in combination with a cross validation method; according to the invention, the detection requirements of asset full coverage and early threat discovery in a complex network environment are met.

The invention discloses a power load prediction method, and the method comprises the steps: firstly solving an extreme event data sparsity problem through a generative adversarial network, and constructing an event time sequence library through a time sequence anomaly detection algorithm; then analyzing the causal relationship between the event and the load by applying a causal discovery algorithm, and converting prediction output into probability distribution by adopting a Bayesian neural network to quantify uncertainty; constructing a prediction model triggered by an event, and generating a multi-time scale probability prediction interval; and finally, generating a multi-scene prediction result through Monte Carlo simulation, quantifying the system recovery capability in combination with a toughness index, and integrating the system recovery capability to a decision support system to generate a risk response scheme. According to the method, the accuracy and robustness of load prediction under the extreme climate are remarkably improved, full-chain risk insight from early warning to recovery is realized, and prospective decision support is provided for safe operation of a power system.

The invention discloses an air conditioner energy consumption adaptive management system and method based on dynamic feature selection, and relates to the technical field of air conditioner energy consumption management, and the method comprises the steps: building a system energy consumption digital twin model as a theoretical optimal energy consumption baseline; operating parameters are collected in real time, and key parameter subsets are screened through working condition recognition and feature importance dynamic evaluation; inputting the key parameters into the model to obtain theoretical energy consumption, and comparing the theoretical energy consumption with a measured value to generate an energy consumption deviation rate; smooth processing is carried out on the deviation ratio sequence, recognition is carried out in combination with a dynamic threshold value and various anomaly detection algorithms, and grading early warning is triggered; the system comprises four core modules, namely a digital twinborn model construction module, a key parameter dynamic screening module, an energy consumption deviation calculation module and a grading early warning triggering module. The method can adapt to different working conditions, accurately capture energy consumption abnormities, and effectively improve the intelligent level and accuracy of energy efficiency management of the air conditioning system.

The invention relates to the field of charging pile use, and discloses a charging pile use frequency analysis method based on charging frequency data, which comprises the following steps: acquiring charging frequency original data of a plurality of charging piles at different time periods and different dates, performing structured processing in combination with a time label, and constructing a charging behavior data set containing a time dimension label; introducing a sliding window algorithm and a periodic spectrum analysis method, and performing frequency feature extraction on the charging behavior data set; using a density peak clustering algorithm to classify the use frequency matrixes of different charging piles under the same time scale, and identifying use preferences and active features of the charging piles under various typical time scenes; based on a lightweight time sequence neural network structure, a frequency fluctuation sensing mechanism is introduced; and predicting the use condition of the charging pile in a future time period based on the trained frequency model in combination with regression prediction and an anomaly detection algorithm. The method has the advantage of improving operation scheduling.

The invention relates to the technical field of power equipment monitoring, and discloses a high-voltage circuit breaker state detection and fault diagnosis method. The method comprises the following steps: acquiring real-time state data of the high-voltage circuit breaker and preprocessing to obtain standardized data; extracting time domain, frequency domain and time frequency features from the standardized data to form feature vectors; analyzing the feature vector by using an anomaly detection algorithm, and identifying an abnormal state mode; based on the abnormal state mode, combined with historical fault records, a causal reasoning method is applied to mine a causal relationship between an abnormal state and a fault, and a key fault index is determined; establishing a health reference model according to the key fault indexes, and calculating the deviation degree between the current state data and the health reference model; evaluating the deviation degree to obtain a fault risk level; and estimating the fault probability of the high-voltage circuit breaker in the future time period by using the prediction model according to the fault risk level and the key fault index, and generating a state detection report.

The invention discloses a distributed photovoltaic output prediction method and system in changeable weather, and the method comprises the steps: firstly obtaining historical photovoltaic station operation data, and carrying out the processing through a photovoltaic data preprocessing model: carrying out the feature engineering through timestamp feature extraction and multi-scale wavelet transformation, and combining a clustering algorithm with an anomaly detection algorithm, thereby achieving the prediction of the distributed photovoltaic output. Dividing and purifying the data into a plurality of weather type data sets; then, for each weather type data set, an independent photovoltaic output prediction model is trained, and the model is composed of an LSTM layer, a multi-head attention layer and a full connection layer; during prediction, the weather type attribution of the real-time data is firstly judged, and then the corresponding pre-training model is called to complete prediction. Weight calculation of multiple attention layers is dynamically guided by using frequency domain features extracted by wavelet transform, adaptive modeling of global features and local details is realized, and the accuracy and robustness of distributed photovoltaic output prediction are improved.

The invention belongs to the technical field of wind power generation, and particularly relates to a fault diagnosis and early warning method and system for a wind power plant, and the method comprises the steps: collecting the state data of a plurality of physical fields through a plurality of types of sensor arrays disposed at key parts of a fan; performing preprocessing and feature extraction on the data to obtain high-dimensional feature data; driving the field-level digital twinborn model to perform real-time simulation based on the high-dimensional feature data, and determining an initial fault recognition result and severity by comparing a simulation predicted value with actual sensing data and combining with an unsupervised anomaly detection algorithm; a diagnosis result is calibrated through multi-modal high-dimensional feature data fusion; predicting the remaining useful life of the key component based on the fault severity and the evolution trend thereof, and generating a collaborative operation and maintenance decision by using a multi-objective optimization algorithm; and an operation and maintenance decision is automatically executed through a closed-loop feedback execution mechanism. According to the invention, the accuracy of fault diagnosis and early warning of the wind power plant is improved.

The invention relates to the technical field of large model cheating detection, in particular to an online interview large model cheating detection system and method based on eye movement tracking, and the detection method comprises the steps: S1, system initialization and environment adaptation; s2, eye movement tracking calibration and baseline establishment; s3, collecting and preprocessing real-time eye movement data; s4, executing an anomaly detection algorithm; and S5, carrying out risk assessment and outputting a result. According to the scheme, based on the WebGazer.js technology, the mapping relation between the fixation point and the screen coordinates is established through nine-point initialization calibration, and the personalized coefficient k is generated in combination with the pupil diameter and the facial features of the candidate, so that the eye movement tracking error is greatly reduced, and the large model cheating detection precision requirement can be met without special hardware. The design not only reduces the deployment cost, but also breaks through the hardware limitation, so that the technology can adapt to mainstream notebook equipment, is convenient for large-scale popularization and application, and solves the contradiction that high precision and low cost cannot be achieved at the same time in the traditional scheme.

The invention belongs to the field of computer vision, provides a video anomaly detection algorithm based on infrared visible light image feature fusion, and aims to solve the problem of video anomaly detection in an insufficient illumination condition or a low-visibility environment. A convolutional neural network encoder is utilized to extract visible light features of a video frame, a multi-stage memory encoder is utilized to perform enhancement and hierarchical memory on significant features in an infrared image to obtain enhanced infrared features, and finally a result is output through an upper sampling layer of another convolutional neural network. Deep fusion is carried out on the infrared and visible light features to obtain fusion features, and finally, a decoder corresponding to the convolutional neural network encoder is guided to reconstruct a target frame according to the fusion features and the enhanced infrared features. And detecting and judging whether an abnormal behavior exists or not through the target frame. The method is mainly applied to video anomaly detection system design and manufacturing occasions.

The invention discloses a test method and system for a motor driver of a vehicle-mounted motor of a new energy automobile, and the method comprises the steps: S1, collecting dynamic environment and working condition parameters: collecting the performance parameters of the motor driver under a preset dynamic environment condition and a simulated working condition; the dynamic environment conditions comprise the combination of the temperature of-40 DEG C to 85 DEG C, the humidity of 10%-95% and the vibration frequency of 10-2000 Hz; the simulation working conditions comprise an NEDC circulation working condition, a WLTC circulation working condition and a rapid acceleration / deceleration dynamic working condition; s2, multi-modal anomaly detection: on the basis of the parameters acquired in the step S1, through a multi-modal anomaly detection algorithm fusing a random forest, an isolated forest and an auto-encoder, and in combination with a Bayesian optimization dynamic threshold, identifying an abnormal mode of the motor driver; s3, performing adaptive parameter optimization and performance evaluation; and S4, generating a test result. The method has the advantages that the technical scheme covers the temperature of-40 DEG C to 85 DEG C, the vibration of 10 Hz to 2000 Hz and the dynamic working condition, the reliability verification scene is expanded to the full life cycle from the static state, and the defect that the environment and working condition coverage is incomplete in the original scheme is overcome.

The invention relates to the technical field of data processing and analysis, and discloses a visual anomaly detection method and system for a rail transit operation and maintenance scene, and the method comprises the steps: collecting rail transit operation and maintenance image data, carrying out the preprocessing of the image data, and selectively transmitting the image data to a constructed anomaly detector through a target detector, and performing feature extraction and aggregation, feature distribution fitting and anomaly score calculation by using an anomaly detector to obtain an anomaly detection result of the anomaly detector for the input image, and finally performing rail transit operation and maintenance anomaly alarm according to the anomaly detection result. Therefore, by adopting the optimization design of network structures such as a feature breathing mechanism, an asymmetric codec structure and a self-attention decoder layer, the model performance of an anomaly detection algorithm is remarkably improved, anomaly detection of multiple types of parts or equipment can be realized in one model, and meanwhile, the detection accuracy is improved. The method can improve the missing detection risk of an existing detection system, and improves the robustness of the system and the adaptability to diversified anomalies.

The invention relates to the field of Internet of Things data processing, in particular to a hydropower station IoT real-time data acquisition and uplink method based on edge computing, the core technology comprises data acquisition and preliminary screening processing of received data, and the screening processing method comprises a threshold detection algorithm and a moving average anomaly detection algorithm. Cleaning the data through a quartile distance judgment algorithm; judging and calculating the state of the equipment, and calculating the credibility of the data vector to obtain the running state of the equipment; data packaging and uploading: generating a data uploading block chain collected by a digital fingerprint combination through dual hash; the oracle machine establishes a contract rule to monitor the uploaded data in real time, and pushes abnormal information to the early warning platform when the data is abnormal, so that real-time early warning is realized; through the three-level architecture of edge calculation, the block chain and the oracle machine, on the premise of ensuring data credibility, the decision-making precision and the response speed are improved, and the three pain points of uncredibility, high delay and many false alarms in the industrial monitoring field are solved.

The invention relates to the technical field of building construction intellectualization, in particular to a building construction intelligent monitoring system based on multi-source sensor data fusion. Comprising a data acquisition and storage unit; a data fusion processing unit; the construction state analysis unit uses an improved anomaly detection algorithm; a risk assessment early warning unit; and a visualization and decision-making unit. Through the dynamic adaptation module, the structural displacement feature weight is adjusted based on the construction stage parameters to preferentially extract the construction key stage structural features, and the structural material elastic deformation limit is corrected in combination with the material linear expansion coefficient and the temperature difference value; meanwhile, the dynamic critical generation module generates a risk critical judgment basis changing along with the construction process in combination with structural bearing capacity parameters and equipment mechanical stress limits in the construction stage, so that the problem that abnormal detection and risk judgment are lack of dynamic adaptability is solved; and the accuracy of personnel border crossing, equipment abnormity and structure abnormity identification and the adaptability of risk assessment are effectively improved.

The invention discloses a safety monitoring system and monitoring method for a hydrogen storage device, and relates to the technical field of hydrogen storage, and the system comprises an energy supply module which supplies power to a data collection device through an intelligent power management system; the data acquisition module is used for collecting real-time operation data from a sensor of the hydrogen storage device; the intelligent analysis module is used for performing real-time analysis on the collected data based on a machine learning algorithm, establishing a safety risk assessment model and obtaining a risk development trend; the early warning control module is used for judging whether abnormity occurs or not based on the risk development trend; and the remote monitoring module communicates with the subsystems through a network and provides remote data monitoring. According to the method, key data of the hydrogen storage device are collected in real time, intelligent analysis is conducted through machine learning, environment changes are monitored in real time, safety risks are recognized in advance, leakage and pressure anomalies are recognized through deep learning and an anomaly detection algorithm, the risk development trend is predicted, and safe operation of equipment is guaranteed.

The invention discloses a self-powered multi-parameter farmland microenvironment monitoring system, and belongs to the technical field of smart agriculture and Internet of Things. The system comprises a distributed self-powered wireless sensing node array, gateway equipment and a cloud data platform. The sensing node adopts a photovoltaic and soil temperature difference synergistic hybrid energy supply module, integrates temperature, humidity, pH and other multi-parameter sensing probes, and realizes long-distance low-power-consumption transmission through an improved LoRa communication module. And the gateway is used for receiving and fusing the data and uploading the data to the cloud platform through 4G or Wi-Fi. The cloud platform can store and visualize data, spatial modeling is realized by using improved Kriging interpolation, a soil moisture content prediction model and an anomaly detection algorithm are introduced, and intelligent irrigation suggestions are generated in combination with a crop water demand model. The system has the advantages of stable energy supply, high sensing precision, strong communication anti-interference capability, intelligent data processing and the like, can operate unattended for a long time, is widely suitable for scenes of field planting, greenhouse agriculture, scientific research experiments and the like, and provides reliable support for precise and intelligent management of agricultural production.

The invention discloses a distributed new energy monitoring method and system based on machine vision, and the method comprises the steps: obtaining real-time image data through an image collection device, combining an abnormality detection algorithm and a posture detection algorithm, positioning equipment abnormality, evaluating the energy conversion efficiency, and fusing a light and shadow distribution analysis result to judge a potential abnormal region. And for an abnormal region, extracting operation data from the edge calculation unit, integrating image analysis and operation parameters by adopting a multi-source data fusion algorithm, generating a global state evaluation index, and triggering an early warning and maintenance process based on the global state evaluation index. Through multi-dimensional data fusion and intelligent algorithm integration, the operation reliability and maintenance efficiency of renewable energy facilities are remarkably improved, and important technical support is provided for sustainable development of green energy.

The invention relates to a power operation data processing method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring power data of each data source in a power system; analyzing data attributes of the power data, and converting the data attributes into a predetermined format to obtain target data; based on the target data and the corresponding data attributes, establishing a knowledge graph; wherein the knowledge graph comprises nodes and association relationships between the nodes; when it is monitored that the electric power data of any data source has a change condition, updating the knowledge graph based on the change condition; and screening out key nodes from the knowledge graph, and carrying out real-time monitoring on the key nodes according to an anomaly detection algorithm to obtain an anomaly detection result. Therefore, the knowledge graph is established based on the target data and the data attributes, the dispersed data sources and data entities in the power system are presented in a node form, and the association between the data sources and the data entities is reflected by using the association relationship, so that workers can intuitively understand the internal relationship among the elements of the power system.

The invention discloses a real-time monitoring method and system for deformation of surrounding rock and a supporting structure in a construction period. The method comprises the following steps: acquiring attitude angle change data of a radar platform in pitching and rolling directions in real time through a high-precision tilt angle sensor; calculating an attitude angle variation of the attitude angle change data, comparing the attitude angle variation with a preset threshold value, and if the attitude angle variation is smaller than the preset threshold value, determining that the radar platform is stable; performing time sequence analysis on the initial deformation data by using an LSTM neural network model, predicting a surrounding rock deformation trend, identifying an abnormal deformation mode by using an anomaly detection algorithm, calculating a surrounding rock risk level based on a random forest classifier, generating a displacement cloud picture, calculating a convergence rate and positioning a maximum displacement point; and calculating a convergence rate and positioning the maximum displacement according to the displacement cloud picture to generate a risk prediction report, and triggering a multi-stage intelligent alarm. Dependence on external manual rechecking is reduced, and personal errors and labor intensity are reduced.

The invention discloses an AI substitute inspection method and system based on multi-modal perception, and relates to the technical field of intelligent detection. According to the method, images, audios, environmental parameters and equipment state data of an inspection area are synchronously collected through a multi-modal sensor array, and after space-time alignment preprocessing, multi-modal features are extracted and input into a pre-trained fusion model to generate unified feature representation. The fusion model adopts a Transform architecture to realize cross-modal feature interaction, an anomaly detection algorithm is used to identify and classify anomalies, a confidence coefficient is calculated in combination with an evidence theory, and finally an inspection report containing anomaly positions, types and processing suggestions is generated. According to the invention, the information limitation of single-mode inspection is solved, the detection accuracy and robustness in a complex environment are improved, unmanned inspection in a high-risk scene is realized, the labor cost is reduced, the safety is guaranteed, and the method is suitable for automatic inspection in the fields of industrial equipment, public facilities and the like.

The invention discloses an automatic operation and maintenance method and system for a distributed system, and the method comprises the steps: collecting a multi-dimensional monitoring index, and carrying out the format conversion; acquiring historical data of the multi-dimensional monitoring index in the same time period within a preset time, calculating a dynamic baseline based on a 3-Sigma dynamic baseline anomaly detection algorithm, and generating a current dynamic upper limit and a current dynamic lower limit; if the current dynamic upper and lower limits exceed a dynamic baseline and the duration exceeds a threshold value, generating an original alarm event; inputting the original alarm event into a topology engine, checking whether a father node of an input node also gives an alarm, and if the father node does not give an alarm, marking the alarm as a root cause alarm; and based on the root cause alarm, matching a predefined processing strategy. The method can automatically adapt to periodic fluctuation of services, the false alarm rate is remarkably reduced, and the threshold does not need to be manually and frequently adjusted; according to the system, when a core component fails, troubleshooting interference is greatly reduced, and efficient alarm storm suppression is realized.

The invention discloses an unmanned aerial vehicle nest full life cycle management and control method, system, product and device, and relates to the field of unmanned aerial vehicle nests, and the method comprises the steps: employing a bidirectional access mechanism to carry out the bidirectional validity verification of an unmanned aerial vehicle and a nest, and generating an access safety certificate; integrating multi-dimensional data through edge computing nodes, and constructing a visual relation model and a dynamic behavior monitoring model; carrying out real-time risk identification and early warning by adopting a multi-dimensional feature fusion anomaly detection algorithm; the digital twinning technology is utilized to fuse aircraft nest operation data, and predictive maintenance is achieved through a life prediction model; and safely erasing the sensitive data of the equipment and cancelling the voucher to form a management closed loop. Through a bidirectional access mechanism, multi-dimensional anomaly detection, predictive maintenance and safe decommissioning processing, full-link active safety protection from access, operation, maintenance to decommissioning is realized, the safety and reliability of a machine nest system are improved, and the system can be applied to complex scenes with strict safety requirements such as electric power inspection.