4503 results about "Long short term memory" patented technology

The invention relates to the field of power equipment data processing, in particular to a biomass power generation combustion parameter deep learning method and system, and aims to solve the problem of phase mismatch caused by sampling frequency difference and clock reference offset of multi-source heterogeneous time sequence data. A parallel multi-scale convolution and bidirectional long-short-term memory network hybrid model is constructed, transient fluctuation and long-period trend features are extracted, and combustion stage feature weights are dynamically distributed through a gating attention mechanism. The optimization control module generates a multi-target constraint condition, an operation instruction is output in combination with a fuzzy inference engine, and a digital twin platform simulates an extreme working condition to enhance model robustness. A closed-loop feedback mechanism dynamically adjusts model parameters through combustion efficiency monitoring data and simulation results, and a two-stage fault-tolerant strategy realizes sensor abnormity compensation and historical control strategy backtracking. The problem of asynchronous data stream feature misalignment is effectively solved, and the combustion efficiency prediction precision and the control decision reliability are improved.

The invention relates to the technical field of electronic manufacturing equipment, in particular to an adaptive thermal compensation system and method for a high-precision mounting head of a chip mounter, and the system comprises a temperature-deformation sensing unit, a thermal-mechanical coupling analysis unit, a dynamic compensation control unit, and a closed-loop execution unit. The temperature-deformation sensing unit collects temperature and deformation data of multiple parts of the mounting head in real time, the thermal-mechanical coupling analysis unit reconstructs a three-dimensional temperature field based on a finite element algorithm, the thermal expansion distribution quantity is dynamically calculated, the problem of rough model of traditional single-point temperature measurement is solved, and the measurement precision is improved. The dynamic compensation control unit predicts the thermal drift amount in the future 5 ms through online parameter identification and a long-short-term memory network model, a compensation strategy is adaptively adjusted in combination with the motion working condition, the closed-loop execution unit decomposes the compensation amount into displacement and torsion correction instructions, accurate offset of thermal deformation is achieved, and a whole-process thermal compensation closed loop is constructed. The precision stability of the mounting head in a complex thermal environment is improved, and the production efficiency is improved.

The invention relates to the technical field of fan blade fatigue damage prediction. The invention provides a fan blade fatigue damage prediction method and system. The method comprises the following steps: constructing a coupling finite element model based on blade anisotropy parameters; blade surface three-dimensional strain field data, blade vibration acceleration signals, environment temperature and humidity and wind speed and direction data are obtained in real time, and a multi-dimensional monitoring data set is constructed; based on the multi-dimensional monitoring data set, nonlinear coupling features of all the load components are extracted, a multi-dimensional feature tensor is obtained, and a reference stress field matched with the current working condition is generated; inputting the multi-dimensional feature tensor and the reference stress field into a bidirectional long-short-term memory network, and establishing a data-physics combined driven damage evolution model; and positioning a damage area based on a damage probability distribution diagram output by the damage evolution model. The problems that in the prior art, prediction errors are obvious, sensitivity to early damage is insufficient, the false alarm rate is high, and accurate positioning of the damage position and quantitative prediction of the residual life are difficult to achieve are solved.

The invention provides a supply chain multistage warehousing intelligent scheduling and collaboration method and system, and relates to the technical field of intelligent warehousing, and the method comprises the steps: carrying out the hierarchical modeling prediction of the short-term and long-term demands of each stage of warehouse through employing a long-short-term memory network based on historical order data, and obtaining the inventory demand; setting each warehouse node as an independent intelligent agent, and generating an inventory allocation strategy through strategy iteration optimization among the intelligent agents; converting the inventory allocation strategy into a scheduling instruction containing an allocation object, an allocation quantity and allocation time based on a rule knowledge base; and in the edge computing unit of each warehouse node, receiving a scheduling instruction and performing sorting execution, when inventory abnormity or resource conflict is detected, initiating an emergency cooperation request to an adjacent warehouse node, returning response information by the warehouse node receiving the emergency cooperation request according to the own resource state, and performing scheduling according to the response information. And determining an emergency processing scheme through local negotiation between the nodes.

The invention relates to a multi-model space-time combination flood peak prediction method fusing physical constraints, which comprises the following steps of: acquiring static space data, dynamic time sequence data and boundary data of a research drainage basin, converting the static space data of the digital elevation model into a grid matrix, and calculating the dynamic time sequence data of the research drainage basin according to the grid matrix and the dynamic time sequence data of the research drainage basin; representing an elevation value of each geographic position, extracting gradient features by using gradient calculation according to the elevation values, and performing normalization processing on the gradient features to obtain normalized gradient features; a CNN-Bi-LSTM-Transform prediction model is constructed, the prediction model comprises a spatio-temporal data alignment module, a CNN convolutional network, a spatio-temporal feature splicing module, a bidirectional long and short term memory network and a Transform module, the prediction model is trained, and a loss function during training adopts a physical constraint loss function composed of mean square error loss and water conservation constraint terms and is used for flood peak prediction.

The invention relates to the technical field of structural health monitoring and intelligent diagnosis, in particular to a structural fatigue damage identification method based on acoustic emission and deep learning, and the method comprises the steps: collecting a structural response signal under a fatigue load through an acoustic emission sensor array, inputting the structural response signal to a CNN-BiLSTM-Attention mixed deep learning model, and carrying out the recognition of the structural fatigue damage through the CNN-BiLSTM-Attention mixed deep learning model; the model extracts local time domain features through a dynamic adaptive convolution kernel, captures long time sequence dependence by using a bidirectional long-short-term memory network, focuses key damage features through a bimodal space-time attention mechanism, divides damage stages based on a nonlinear dynamic threshold algorithm of fracture opening amount, constructs a training data set of physical-data fusion, and performs dynamic time domain feature extraction. The learning rate is optimized by adopting a gradient sensitive cosine annealing algorithm, and the robustness of the model is improved in combination with an anti-noise and anti-loss function. The method integrates physical characteristics and an intelligent algorithm, and has the advantages of adaptive noise suppression, strong cross-domain generalization ability, high real-time performance and the like.

The invention relates to an intelligent power distribution room operation and maintenance method and system based on multi-source data fusion, and the method comprises the steps: collecting the environment and equipment data of a power distribution room through multi-source sensing equipment, and obtaining original multi-source data; performing standardization processing on the original multi-source data to obtain standardized heterogeneous data; performing space-time correlation modeling on the equipment vibration characteristics and the equipment power parameters through a space-time diagram convolutional network, and fusing the cross-modal depth characteristics extracted by the standardized heterogeneous data to obtain panoramic perception characteristics; performing anomaly detection through an isolated forest-long and short-term memory hybrid model according to the panoramic perception features, and deploying a causal inference engine to analyze a causal relationship among multiple variables to obtain a health state assessment result of the power distribution room; and performing operation and maintenance decision according to the health state evaluation result of the power distribution room, and performing operation and maintenance on the power distribution room. According to the invention, deep fusion of multi-source data is realized, equipment abnormity can be accurately detected, root causes can be analyzed, and the operation and maintenance efficiency and reliability of a power distribution room are improved.

The invention relates to a computing power network resource scheduling method. The method comprises the following steps: acquiring floating point operation performance parameters and operation states of node processors and energy index data of data centers where the node processors are located, and calculating to generate a node list; constructing a global resource pool based on the list, and generating a resource distribution table containing the total calculation power of the region; obtaining calculation requirements and time delay constraints of the task queue, extracting feature vectors in combination with the resource distribution table, and generating a resource utilization rate table; obtaining network link flow data, predicting a link congestion probability through a long short-term memory network, and generating a flow control strategy table; and finally, updating resource pool network constraints according to the resource utilization rate table and the flow control strategy table, and remapping tasks by taking node effective computing power as a weight to generate a scheduling execution scheme. According to the method, accurate quantitative evaluation of the computing power resources is realized, the matching precision of tasks and the computing power resources is effectively improved, the resource utilization rate of the computing power network can be remarkably improved, and the overall scheduling efficiency and stability of the system are enhanced.

The invention relates to the technical field of medical image diagnosis, and discloses a medical image automatic diagnosis method and system based on deep learning. According to the method, multi-modal medical image data of a target object is acquired and standardized, a two-channel convolutional neural network is utilized to extract features, the features are processed through cross-modal feature fusion, adaptive attention weight distribution and other technologies, a cascaded two-way long-short-term memory network is adopted for modeling, abnormity is detected based on a probabilistic graph model, and the target object is identified. And the nidus is segmented by a multi-scale context information enhancement module, and finally a diagnosis suggestion is generated by a diagnosis inference engine driven by a knowledge graph. The system comprises a multi-modal image acquisition interface module, a distributed feature calculation cluster, a visual interaction terminal and a security audit module. According to the method, the accuracy and efficiency of medical image diagnosis can be improved, comprehensive diagnosis reference is provided for doctors, and meanwhile data safety and privacy are guaranteed.

The invention discloses a ship arrival and departure prediction method based on a multi-modal neural network and an adaptive LSTM, and the method comprises the following steps: S1, collecting data, including ship position data, environment data and historical arrival and departure data; s2, data preprocessing: cleaning and standardizing the data; s3, feature extraction optimization is carried out on various types of data; s4, performing multi-modal data fusion, including data splicing fusion, data weighted fusion and data deep fusion; s5, training and constructing a self-adaptive LSTM model, wherein the self-adaptive LSTM model predicts the port entering and leaving behaviors and time of the ship; s6, evaluating and optimizing a self-adaptive LSTM (Long Short Term Memory) model; s7, performing real-time prediction and scheduling; and S8, dynamically predicting and providing scheduling suggestions. Based on the ship position data, the environment data and the historical port entering and leaving data, the multi-mode neural network is used for data fusion, the port entering and leaving prediction model based on the self-adaptive LSTM is constructed, and the prediction accuracy and real-time performance are improved.

The invention relates to the technical field of bearing fault detection, and discloses a bearing fault detection method and system based on a health state index. The method comprises the following steps: collecting multi-source sensing signals at least comprising a vibration signal, a temperature signal and an acoustic signal during bearing operation; respectively performing time domain feature extraction and frequency domain feature extraction on the multi-source sensing signals, and performing normalized fusion on the extracted time domain features and frequency domain features to generate a multi-dimensional health state index sequence; constructing a long-short-term memory network model based on an attention mechanism, inputting the multi-dimensional health state index sequence into the model for training, and outputting a bearing health state prediction sequence; and calculating a dynamic early warning threshold according to the historical health state prediction sequence, comparing the current prediction value with the dynamic early warning threshold in real time, and generating a fault early warning signal. The method can improve the accuracy of bearing health state evaluation and fault early warning, and is suitable for complex operation conditions.

The invention provides a brushless motor attitude detection and balance control method and system based on a gyroscope, and relates to the technical field of control, and the method comprises the steps: collecting angular velocity and acceleration data through a six-axis gyroscope, carrying out the noise reduction through wavelet transform, and carrying out the attitude calculation through the combination of an extended Kalman filter and a quaternion algorithm. A rotor position signal is obtained through a magnetic encoder, nonlinear compensation is carried out, and rotating speed data are calculated. A motor state is modeled by adopting a long-short-term memory network, a double-layer adaptive fuzzy neural network controller is constructed, and attitude error compensation and rotation speed fluctuation suppression are realized. A controller model is optimized through particle swarm optimization and a genetic algorithm, a compensation current vector is corrected in real time, and the control precision and stability of the brushless motor are improved. According to the method, the operation efficiency and the dynamic response capability of the brushless motor are effectively improved.

The invention relates to the technical field of intelligent power grids, in particular to a power distribution network battery digital dynamic management system based on digital twinning. Comprising a data acquisition unit; the digital twinborn modeling unit is used for constructing a battery-power grid-environment multi-dimensional dynamic twinborn body and realizing virtual-real bidirectional mapping and adaptive updating by combining a multi-physics field coupling model and a long-short-term memory network time sequence prediction algorithm; a dynamic optimization unit; and executing the feedback unit. Through a distributed heterogeneous sensing network of a data acquisition unit, multi-dimensional operation data of a battery pack and a key node of a power distribution network are acquired, and a high-fidelity data set containing four-dimensional labels of a battery state, a power grid parameter, time and a position is generated in combination with a spatial-temporal feature extraction technology; the deep fusion of the full life cycle state of the battery and the global operation data of the power distribution network is realized, and the comprehensive data support covering the global is provided for the optimization decision.

The invention discloses a monorail crane inspection robot intelligent test method based on data analysis, and relates to the technical field of intelligent detection, and the method comprises the following steps: synchronously collecting track images, point cloud and attitude data, carrying out time alignment and preprocessing, and outputting a standardized data packet; detecting an inspection target by using a YOLO detection network, and outputting a multi-scale feature vector in combination with a point cloud feature hierarchy extraction network and a time sequence convolutional network; predicting a fault development trend in combination with an improved A-star algorithm and a long short-term memory network, optimizing an inspection path through reinforcement learning, and outputting a maintenance decision scheme; the maintenance decision scheme is converted into a control instruction, the robot is driven to execute an inspection task and feed back the operation state in real time, incremental learning and point cloud reconstruction are combined, and a visual diagnosis report is output. According to the method, the dynamic threshold algorithm is adopted for self-adaptive analysis, and the key geometric indexes are calculated in combination with the cross-modal attention mechanism, so that the recognition capability of structural anomalies is improved.

The invention provides a micro-grid energy management method and system based on deep reinforcement learning, and relates to the technical field of power grids, and the method comprises the steps: constructing a dual-time scale deep reinforcement learning model which comprises a day-ahead scheduling sub-network and a real-time scheduling sub-network; the day-ahead scheduling sub-network predicts micro-grid operation strategies at a plurality of time points in the future based on the long short-term memory network; the real-time scheduling sub-network is based on a depth deterministic strategy gradient algorithm, real-time state data and a day-ahead scheduling prediction result are fused to construct an evaluation function, an instant reward value is calculated, and renewable energy power generation, energy storage charging and discharging and an external power grid electricity purchasing and selling power adjustment instruction are optimized online. According to the invention, through dual-time-scale collaborative optimization, the energy management efficiency and economic benefits of the micro-grid are improved, and the operation stability of the micro-grid is enhanced.

The invention discloses a real-time surrounding rock deformation monitoring and data acquisition method and system, which is applied to long-distance weak surrounding rock tunnel construction, and comprises the following steps: determining the dynamic change trend of underground water seepage rate and ground stress distribution gradient by adopting a time sequence analysis method; based on the trend, carrying out risk partitioning on the tunnel construction section by adopting a K-means clustering algorithm, determining a deformation sensitive area, and optimizing the spatial distribution of the monitoring points according to the deformation sensitive area; monitoring data are acquired in real time, and when the data fluctuation period exceeds a threshold value, the data acquisition frequency of the corresponding monitoring point is automatically improved; processing high-frequency acquired data by adopting a long-short-term memory network to obtain a real-time surrounding rock deformation prediction result; the prediction result and the multi-source real-time geological parameters are fused, a Bayesian updating method is adopted for processing, a quantitative surrounding rock stability evaluation result is obtained, closed-loop self-adaptive optimization of a monitoring scheme and accurate risk prediction are achieved, and the safety early warning capacity of tunnel construction and the utilization efficiency of monitoring resources are remarkably improved.

The invention provides a crop monitoring system and method based on multispectral remote sensing and deep learning, and the system comprises a data preprocessing module which is used for carrying out the data preprocessing of a multispectral remote sensing image, and generating a standard reflectivity data set; the feature extraction module is used for extracting a high-dimensional spectral feature vector from the standard reflectivity data set through a multi-scale convolutional neural network; the time sequence dynamic analysis module is used for performing time sequence correlation analysis on the high-dimensional spectral feature vector through a long short-term memory network to generate a weighted time sequence feature vector; the physiological parameter quantification module is used for mapping the weighted time sequence feature vectors into quantitative indexes of crop physiological parameters; and the monitoring result generation module is used for performing dynamic deduction according to the quantitative index and generating dynamic trend prediction data of the crop growth state. The system can dynamically sense the growth stage characteristics of crops and adaptively adjust the spectral feature extraction strategy, thereby improving the crop monitoring precision in a complex agricultural environment.

The invention discloses a cable defect detection method and system based on artificial intelligence. The method comprises the following steps: acquiring cable detection data from a multi-modal data acquisition device; performing standardization processing and time dimension alignment operation on the cable detection data to obtain a normalized multi-modal data set; carrying out feature extraction on the multi-modal data set by adopting a convolutional neural network, and generating a preliminary feature set by alternately executing convolution operation and pooling operation for three times; obtaining a multi-modal feature vector according to the preliminary feature set, and generating dynamic weight distribution of each modal feature through an attention weight calculation module; performing weighted fusion on the multi-modal feature vectors based on dynamic weight distribution to obtain a fused feature set; and judging whether the fusion feature set contains a working condition parameter feature vector or not, and if the fusion feature set contains the working condition parameter feature vector, inputting the fusion feature set into the long-short-term memory network for time sequence modeling. The safety and reliability of the power system are effectively improved.

The invention provides a BIM intelligent management platform oriented to a whole life cycle of project construction. A building information model, Internet of Things sensing data and a block chain evidence storage mechanism are integrated through a multi-source data fusion technology, and a whole-process data chain of association planning, design, construction, operation and maintenance is associated. The platform adopts space optimization Huffman coding to realize model lightweight, combines a constraint genetic algorithm to optimize a construction path, and applies a bidirectional long-short-term memory network to analyze an equipment state. A three-chain block chain system is reconstructed on the architecture, intelligent association of engineering quantity and payment nodes is realized through cooperation of a main chain, a calculation quantity side chain and an auditing side chain, and mobile terminal offline interaction is supported based on a digital-analog separation technology. The platform covers an intelligent design management unit, a block chain investment management unit, a dynamic correction management unit, a quality safety responsibility tracing unit, an NLP risk management unit and a digital twin operation and maintenance unit. The units achieve cross-system cooperation through a unified data bus, and a closed-loop management architecture covering the whole life cycle of project construction is formed.

The invention relates to the technical field of machining precision control, in particular to a laser carving precision control method and system.The laser carving precision control system comprises a feature collecting unit, a model building and analyzing unit, a dynamic threshold value adjusting unit and an online incremental learning unit, and the feature collecting unit collects vibration, current and temperature data through a multi-source sensor array; the model construction analysis unit realizes dynamic prediction of processing parameters by combining a bidirectional long-short-term memory network with an attention mechanism, and the dynamic threshold adjustment unit dynamically updates parameters of a numerical control system based on a material hardness real-time detection and thermal coupling model. The online incremental learning unit automatically generates training samples through error data, continuously optimizes model parameters and constructs a'data acquisition-intelligent modeling-dynamic compensation-model evolution 'closed loop, so that accurate prediction and adaptive adjustment of machining parameters are realized, and the adaptability of the manufacturing process to multi-variety and small-batch working conditions is remarkably improved.

A forest point cloud reconstruction method and system based on combination of a handheld radar and an unmanned aerial vehicle, relating to the technical field of remote sensing. The method comprises: on the basis of an unmanned aerial vehicle and a handheld radar device, using a multi-band radar sensor to perform forest terrain and vegetation information acquisition to generate original radar data. The use of the multi-band radar sensor improves the data acquisition accuracy and the coverage range, so that forest terrain and vegetation information is recorded more comprehensively and accurately; a generative adversarial network improves the data quality, and reduces the noise, thereby ensuring the accuracy of subsequent analysis; a convolutional neural network and a random forest algorithm are combined to improve the accuracy of vegetation classification and terrain feature recognition; the generation of comprehensive environmental feature data integrates multidimensional environmental parameters, providing richer information for in-depth analysis; a three-dimensional point cloud model visually displays a forest structure; and time series analysis and a long short-term memory network improve the accuracy of forest change trend prediction, and provide decision support for forest management strategies.

The invention relates to a network flow anomaly detection strategy generation method based on machine learning, and belongs to the technical field of machine learning. The method comprises the following steps: firstly, collecting network traffic data in a preset time window, and extracting feature vectors containing traffic, a time sequence and a protocol type; and inputting the feature vector into a long short-term memory auto-encoder model, and calculating a reconstruction error to judge whether the network flow is abnormal or not. Aiming at the abnormal feature vector, adopting a multi-agent depth deterministic strategy gradient algorithm to construct a plurality of cooperative agents, and independently generating a candidate abnormal detection strategy by each agent; through a cross-agent strategy evaluation mechanism, the difference between a joint strategy and a single-agent strategy in the aspect of anomaly detection accuracy is compared, cooperation gain is calculated, strategy exploration parameters of all agents are adjusted according to the cooperation gain, and a global optimal anomaly detection strategy is optimized and determined in real time. According to the method, high-precision and low-missing-report network traffic anomaly detection can be realized, and the method has good self-adaptability and real-time performance.

The invention discloses a safety monitoring system of a liquid cooling over-charging pile, and relates to the technical field of over-charging pile monitoring, the system comprises a data acquisition module, a data processing and analysis module, a dynamic model construction module, a temperature prediction module and a safety early warning module; according to the method, the dynamic model is constructed through the long short-term memory network LSTM, the complex nonlinear relation and the time sequence dependence of the multi-dimensional data are mined by using the gating mechanism of the dynamic model, the accurate characterization of the operation state of the liquid cooling over-charging pile is realized, the actual operation state of the equipment can be accurately described, the temperature data time sequence modeling is performed through the LSTM, and the accuracy of the temperature data time sequence modeling is improved. Parameters such as multi-source temperature and cooling liquid flow are fused, real-time prediction of the temperature change trend is achieved, the defect that a traditional algorithm is insufficient in temperature time sequence dependence capture is overcome, temperature abnormity can be recognized in advance, a safety threshold value is dynamically adjusted through a fuzzy logic algorithm, and self-adaptive threshold value adjustment is achieved in combination with parameters such as charging power. The problem that a traditional fixed threshold value is poor in adaptability is solved, and the early warning accuracy is improved.

The invention relates to the technical field of industrial production lines, and discloses an industrial production line multi-device dynamic collaborative scheduling method based on reinforcement learning, comprising the following steps: S1, modeling a three-dimensional state space; s2, hierarchical reinforcement learning architecture; and S3, edge-cloud cooperative execution. According to the industrial production line multi-device dynamic collaborative scheduling method and system based on reinforcement learning, device states, task constraints and resource occupation are integrated into a structured matrix through three-dimensional state space modeling, and a global decision-making layer captures production time sequence dependence by using a bidirectional long-short-term memory network; modeling equipment space association and process constraints through a graph attention network, and generating a global strategy including task allocation, capacity adjustment and resource pre-allocation; and after the edge layer detects the dynamic event, the cloud platform generates a candidate scheme through Monte Carlo tree search, and realizes dynamic event response and multi-target collaborative optimization by combining multiple targets such as global value network evaluation task completion time and equipment load balancing.

The invention relates to a substation equipment fault early warning method and system, and belongs to the technical field of intelligent substations, and the method comprises the steps: obtaining the target operation state data of substation equipment, the target operation state data comprising target sensor data and target image data obtained by an unmanned plane which carries out the field inspection of a substation; inputting the target operation state data into a completely trained fault diagnosis model, extracting spatial features of target image data by the completely trained fault diagnosis model through a long-short-term memory network, extracting time features of target sensor data through a convolutional neural network, and splicing and fusing the spatial features and the time features to obtain a fault diagnosis result; outputting a fault prediction result; and performing early warning based on the fault prediction result. According to the invention, the fault diagnosis model is combined with the spatial features of the image data and the time sequence features of the sensor data to carry out early warning on the substation equipment fault, and multi-modal data fusion and fault prediction of the substation equipment are realized.

The invention relates to a time sequence knowledge graph multi-hop reasoning method and system oriented to the legal field. The method comprises the following steps: establishing a dynamic mapping relationship through a semantic association technology; extracting life cycle states of the legal provisions, and introducing a legal conflict detection algorithm; setting a revocation influence factor; determining time embedding representation, and introducing clause conflict matrix elements to construct a time sequence knowledge graph; constructing a space-time coupled completion model based on the graph attention network and the long short-term memory network to perform path completion; determining three selected agents, performing reinforcement learning, designing a reward function for each agent, constructing a collaborative arbitration mechanism in combination with a dynamic priority strategy and a weight adaptive algorithm, and performing multi-hop reasoning. By collecting multi-source legal data, the authority, the real-time performance and the relevance of the data are ensured; according to the method, the path reasoning capability in a low-coverage scene can be improved, so that a complex legal knowledge multi-hop path reasoning task can be accurately and efficiently completed.

The invention discloses a power equipment surface defect image recognition and early warning system and related equipment, which comprehensively utilizes the technical means of multi-modal data acquisition, deep learning model recognition, risk quantitative evaluation, trend prediction and the like by constructing a multi-module collaborative system architecture. And comprehensive detection and intelligent management of the surface defects of the power equipment are realized. Multi-modal sensing data are acquired through an image and data acquisition module, and the defect identification precision and the adaptability to complex defect characteristics are remarkably improved by combining an improved ResNet-50 network and a defect identification and positioning module of a YOLOv5 target detection algorithm. And the defect risk assessment and trend prediction module adopts defect area ratio calculation and a long short-term memory (LSTM) network, so that quantitative analysis of defect risks and accurate prediction of an expansion trend are realized, and a reliable basis is provided for operation state assessment of power equipment.

The invention relates to the technical field of power transformer discharge fault diagnosis in a power supply system, in particular to a power transformer discharge fault diagnosis method based on a neural network, and the method comprises the steps: collecting a partial discharge signal, characteristic gas data in oil and power supply system operation parameters in real time through a sensor, and unifying timestamps; statistical aggregation is performed on the partial discharge signals, and standardization processing is performed on the electrical quantity parameters and the discharge characteristic parameters to generate fusion characteristic vectors; constructing an expansion mapping relation library containing laboratory simulation data and actual operation data; training a diagnosis model by adopting a mixed structure of a convolutional neural network and a long-short-term memory network; and deploying the model in an edge computing device, and dynamically verifying and optimizing the performance of the model through online learning. According to the method, the problems of low multi-source data fusion efficiency and insufficient generalization ability of a neural network model to a complex discharge mode are solved, and the operation stability of a power supply system is improved.

The invention discloses a large model data distributed management method and device, equipment and a storage medium, and relates to the technical field of data processing, and the method comprises the steps: segmenting large model training data to obtain a plurality of data blocks, and determining a predicted access frequency based on a long short-term memory network model and the historical access frequency of the data blocks; caching the large model training data corresponding to the data block to a corresponding data cache layer by utilizing the predicted access frequency; constructing a resource portrait by using the static attribute and the dynamic index of the GPU node, and allocating the large model training task to a target GPU node by using a preset hybrid strategy and the resource portrait based on the predicted access frequency and the storage position corresponding to the data block in the data cache layer; when it is monitored that the large model training task on the target GPU node is executed, periodic snapshot is conducted on the large model training task through a distributed snapshot algorithm, and the obtained complete data state is stored in a distributed storage center. And the resource vacancy rate is reduced.

The invention relates to the technical field of aluminum alloy cold stamping quality control, in particular to a multi-source data fusion and real-time optimization-based cold stamping quality control method, which comprises the following steps of: inputting a fusion feature space into a pre-trained LSTM (Long Short Term Memory) model, and associating a real-time stamping speed with a dwell time parameter through an attention mechanism; combined prediction of springback value deviation and defect probability is realized for the first time, a process parameter correction value is dynamically generated based on a prediction result, and a thermal expansion compensation algorithm and micron-sized mold compensation are combined, so that the quality regulation response speed reaches a millisecond level, the compensation precision is improved to + / -2 [mu] m, and the error is reduced by more than 50% compared with that of traditional PID (Proportion Integration Differentiation) control.