547 results about "Space mapping" patented technology

The invention discloses a mine disaster prediction method based on multi-source data, and relates to the technical field of mine safety, and the method comprises the following steps: collecting original data from different types of sensors in a mine, the data types comprising gas concentration, temperature, humidity, wind speed, ground pressure, water level and vibration information; each type of data is provided with a corresponding timestamp and a spatial position identifier. According to the method, time resampling and space mapping standardization of multi-source data are realized, so that the time-space consistency of data fusion is remarkably improved, and the accuracy of disaster prediction model input is ensured. Meanwhile, a dynamic feature matrix is constructed and a high-precision position weight mechanism is introduced, so that the sensitivity of the model to key areas and key parameters is enhanced, the real-time performance and accuracy of mine disaster prediction are effectively improved, and the risk of missing report and false report of an early warning system is remarkably reduced.

The invention discloses a multi-sensor fusion heat pump full life cycle AI maintenance early warning system, and relates to the technical field of new energy utilization, and the early warning system comprises a data collection module which obtains operation parameters in a heat pump full life cycle based on a sensor array, and constructs a data set after preprocessing the parameters; the operation parameters comprise temperature, pressure, flow and micro vibration; the data fusion module is used for extracting trend correlation characteristics and parameter coupling characteristics from temperature, pressure and flow parameters by adopting a dynamic sliding window adaptive to a working condition, and preserving core nonlinear information through KPCA dimension reduction; the micro-vibration signal extraction comprises frequency domain and time domain features. According to the method, features are extracted through a working condition adaptive dynamic sliding window, then through cross-space mapping and a life cycle-working condition double-attention mechanism, the analysis and early warning module depends on a core feature mapping library and a two-dimensional dynamic baseline, through instantaneous and accumulated deviation double judgment, abnormal accurate recognition and stage division are achieved, and early warning perspectiveness is high.

The invention discloses a weld quality intelligent diagnosis system based on deep learning, and relates to the technical field of welding quality detection, and the weld quality intelligent diagnosis system comprises an image quality evaluation module, a feature alignment module, a deviation detection module, a path reconstruction module, a prior enhancement module and a defect identification module, identifying an area of which the signal-to-noise ratio is lower than a preset threshold value, and constructing a noise interference distribution diagram; and the feature alignment module executes a deformable convolution feature alignment operation with a confidence factor adjustment mechanism based on the noise interference distribution diagram to generate an initial space mapping result. Through mechanisms such as image quality perception, robust alignment, deviation detection, self-adaptive reconstruction and prior enhancement, a closed-loop weld joint intelligent diagnosis process is constructed, false alignment errors are effectively inhibited, the multi-modal fusion stability and the defect recognition precision are improved, and the reliability and the intelligent level of the system under complex working conditions are enhanced.

The invention discloses a power distribution equipment on-line monitoring system based on multi-modal data fusion, and relates to the field of power distribution equipment management, and the system comprises a sensing module which is used for collecting an electrical signal, a thermal infrared signal, a mechanical vibration signal and an acoustic signal generated in the operation process of power distribution equipment through a multi-type sensing channel, converting the collected various signals into processable equipment state original data to form an equipment operation state original data set; according to the invention, through accurate acquisition of multi-dimensional signals, combination of space mapping and time delay compensation, data quality is optimized, the reliability of original information is ensured, key features are extracted according to working conditions, subtle state changes are captured by means of temperature field analysis and a variable-resolution spectrum technology, the comprehensiveness, accuracy and response timeliness of equipment operation monitoring are effectively improved, and the real-time performance of equipment operation monitoring is improved. Misjudgment and missed judgment are reduced, and fault risks are avoided in advance.

The invention relates to a PCB production line process parameter intelligent matching method based on feature space mapping, and the method comprises the steps: collecting and fusing the material attributes, structure parameters, sizes and historical process records of a plurality of batches of PCB products, carrying out the normalization preprocessing, removing abnormal data, and constructing a high-quality feature matrix; after multi-dimensional feature expression is realized by utilizing a multi-scale embedded network, a mapping relation between features and process parameters is dynamically learned on the basis of an adaptive space mapping network in combination with a soft constraint multi-objective loss function, and gradient cutting, step length adjustment and a disturbance elasticity pool mechanism are introduced in a parameter recommendation process to guarantee convergence and stability. After the parameters are implemented, feedback data are collected in real time, periodic iteration distillation optimization and loss function recalibration are carried out, and finally a fine-tuning parameter recommendation scheme marked with conflict indexes, confidence intervals and weight suggestions is output for an engineer. According to the scheme, the intelligence, the traceability and the field adaptation capability of parameter recommendation are improved.

The invention discloses a fusion semantic vector space mapping-based translation ambiguity term accurate matching method, which comprises the following steps of: S1, obtaining source language ambiguity terms, context texts and a target language candidate translation list, and extracting domain tags and term matching features to form a multi-modal data set; s2, using improved XLM-R model coding to generate term-level, sentence-level and translation-level semantic vectors; s3, training a dynamic mapping matrix based on a bilingual parallel corpus, and aligning source side vectors to a shared semantic space; s4, fusing the source-side basic vector and the multi-dimensional features through a double-channel attention fusion network, and generating source-side and translation-side comprehensive semantic vectors; s5, introducing term-context attention weight to correct cosine similarity; and S6, outputting an optimal translation through normalized sorting and part-of-speech secondary judgment. According to the method, multi-field ambiguous term accurate matching is realized, the term translation precision and efficiency in professional fields are improved, and the requirements of high reliability of term translation in the fields of medicine, machinery, computers and the like are met.

The invention relates to a multi-dimensional high-resolution turbine blade heat-flow coupling field synchronous measurement method and system, and the method comprises the steps: obtaining a target spot image of known space distribution in a measurement space, and building a three-dimensional space mapping function; carrying out particle matching on the velocity field measurement original image based on a three-dimensional space mapping function, and reconstructing a three-dimensional particle trajectory by adopting a self-adaptive grid algorithm combined with machine learning to obtain a high-resolution three-dimensional velocity vector field; acquiring a temperature field measurement original image, performing three-dimensional space registration according to a three-dimensional space mapping function, further calculating voxel-level temperature distribution, and inverting a three-dimensional temperature field; and carrying out coupling analysis on the three-dimensional velocity vector field and the three-dimensional temperature field, and outputting a three-dimensional heat-flow coupling cloud picture. According to the invention, multi-dimensional and high-precision synchronous measurement of the temperature field and the velocity field in the complex flow field of the turbine blade is realized, the spatial resolution and dynamic response capabilities are remarkably improved, and a reliable experimental basis is provided for analysis and optimization of the cooling performance of the blade.

The invention belongs to the technical field of virtual power plants, and particularly relates to a virtual power plant resource aggregation method for a dynamic peak regulation demand of a power grid, which comprises the following steps of: acquiring multi-source data, preprocessing the multi-source data, and then verifying the data quality; aiming at different resource types including temperature control load, energy storage and charging piles, respectively constructing refined models, setting constraint conditions of the refined models, and solving a resource operation feasible region by applying multi-dimensional space mapping and linear programming; establishing a target function and a constraint condition by taking the lowest cost and the minimum energy abandoning as targets; solving a target function by using a dung beetle optimization algorithm, and screening an optimal resource aggregation scheme by using an entropy weight method; and based on the optimal resource aggregation scheme, dividing peak, valley and normal periods, constructing a four-dimensional peak regulation index, determining a weight by using an analytic hierarchy process, and screening an optimal resource combination in each period to execute scheduling. The method can guarantee the accuracy and high efficiency of the peak regulation demand response of the power grid, and assists in improving the stability of the power system and the renewable energy consumption level.

The invention discloses a generative odor real-time synthesis method and system based on cross-modal potential space mapping, and belongs to the technical field of artificial intelligence and olfaction calculation. The method comprises the following steps: acquiring a multi-modal input stream of a current scene, and extracting an emotion semantic feature vector by using a deep neural network; mapping the semantic features into target odor chemical feature vectors by using nonlinear projection through a pre-constructed vision-smell joint embedding space; constructing a convex optimization model based on olfactory perception, and calculating a basic liquid optimal mixing proportionality coefficient matrix capable of fitting the target vector; the matrix is converted into a micro-fluidic driving signal, and the target smell is synthesized in situ in the micro-fluidic chip. The invention further discloses a self-adaptive cleaning logic and olfactory fatigue compensation mechanism based on scene mutation detection. The method solves the problems that in the prior art, label matching is dependent, new smell cannot be synthesized, and dynamic transition is lacked, and olfactory replicating and real-time generation of abstract semantic scenes are achieved.

The invention discloses an adaptive multi-angle elbow airtightness detection method, which comprises the following steps: arranging a flexible pressure sensor array and a micro inertial measurement unit on the inner wall of an elbow, collecting multi-point sealing pressure and attitude signals in real time, fusing geometric parameters and a spatial mapping relation, and completing denoising, normalization and multi-curvature region mapping of pressure data; a distributed pressure-attitude database is combined, a function model associated with sealing pressure and pipeline geometric parameters is established, inflation, exhaust and deformation of each node are dynamically adjusted through an intelligent closed-loop regulation and control algorithm, collaborative optimization of multi-node pressure and a fitting state is achieved, and the sealing performance of the pipeline is improved. According to the method, the accuracy and the self-adaptive adjusting capacity of sealing pressure distribution under a complex bent pipe structure are effectively improved, the tiny leakage real-time positioning and self-learning optimization capacity is achieved, and the method is suitable for sealing quality guarantee of various types of curved surface pipelines.

The invention relates to the field of mine risk prediction, in particular to a mine dynamic risk prediction method and system based on multi-modal parameter fusion analysis. The method comprises the following steps: obtaining a multi-modal mine omnibearing data stream based on a multi-modal sensor cluster, carrying out adaptive filtering noise reduction and time-space synchronization space mapping, and constructing a cross-modal mine data stream model; performing mine geological micro-fluctuation behavior deep analysis according to the cross-modal mine data flow model, performing dynamic geological risk situation evolution, and constructing a multi-modal geological risk situation map; and obtaining a mine historical risk event log, performing risk event extraction, performing adaptive risk chain deep learning on the multi-modal geological risk situation map, and constructing a mine risk state dynamic evolution model. According to the invention, intelligent prediction and dynamic early warning of mine risks are realized, and the mine intrinsic safety level and the emergency response efficiency are improved.

The invention relates to a construction method of a power grid equipment defect training sample set and a defect detection method thereof in the field of computer vision and power inspection. The construction method comprises the following steps: potential space mapping; injecting noise and conditions; performing condition denoising; defects are directionally generated. The defect-free background and structure information is reserved under the guidance of a diffusion model, and defect features consistent with text vector description are generated in a specified area, so that a piece of defect-free image is converted into an image with a known fault state for forming a power grid equipment defect training sample set. According to the method, the controllably generated diffusion model is introduced to enhance scarce defect category samples, the sample diversity is improved in combination with a traditional and adversarial generation method, and the problems of incomplete data, scarce defect samples, unbalanced category distribution, insufficient fine granularity detection precision and the like generally existing in existing power grid equipment defect detection are solved.

The invention relates to the technical field of geological disaster monitoring and early warning, and discloses a geological disaster early warning system based on multi-source data fusion. The system comprises a data acquisition and preprocessing module, a dynamic coupling modeling module, a space mapping and feature recognition module and a risk analysis and early warning generation module. Firstly, transient disturbance and long-term steady-state components in monitoring data are separated; a dynamic coupling model containing bidirectional geomechanical feedback is constructed, the component fusion proportion is automatically adjusted according to the feedback intensity, and physically consistent fusion data is generated; spatial mapping is carried out by using an adaptive grid, and an effective abnormal feature cluster is identified through parallel scanning and prior geological knowledge constraint; and constructing a causal graph based on the abnormal clusters to carry out risk assessment and early warning. According to the system, physical driving of data fusion and knowledge guidance of anomaly recognition are realized, and the accuracy and reliability of early warning are improved.

The invention discloses a spraying control method of a waveform guardrail spraying manipulator, which relates to the technical field of industrial automation and robot control, and comprises the following steps: acquiring relative position change data of the manipulator and the surface of a waveform guardrail in a rotation process through a pre-established manipulator rotation track model; and for each time node, recording the spraying angle change and distance dynamic adjustment information, establishing a dynamic space mapping relationship under a manipulator rotation state for spraying control, and triggering a secondary spraying supplementing control instruction according to the coating quality consistency feedback, so that the reliability and stability of the spraying supplementing effect can be further ensured, and the spraying efficiency is improved. Therefore, the equipment utilization rate of the manipulator and the automation and intelligence level of spraying operation are remarkably improved while the uniformity of the coating and the consistency of spraying quality are improved.

The invention discloses an immersive interactive system and method integrating pet robot perception and AR rendering, and the system carries out the visual-inertial combined mapping and anchor point generation, the coding and network transmission of visual frames and spatial data, the synchronous decoding and coordinate system alignment of a terminal side, virtual object anchoring, and spatial mapping and AR rendering through a robot side. User intention analysis, control issuing and robot control, and low-delay synchronization, timeline alignment and closed-loop consistent maintenance are carried out; according to the method, the first visual angle image collected by the front camera of the pet robot can be utilized, the augmented reality technology is combined, and the user is substituted into the visual angle of the robot, so that the user can'personally feel 'the observation and action process of the pet robot, and the immersion and immediacy of human-pet interaction are remarkably enhanced; and human-machine-virtual object ternary space interaction is realized.

The invention discloses a deep reinforcement learning optimization method for injection molding process parameters, and belongs to the technical field of intelligent manufacturing. The method comprises the following steps: constructing a dynamic causal graph network through information entropy flow analysis and transfer entropy calculation, and revealing a causal relationship and time delay characteristics among process parameters; manifold learning is adopted to map a high-dimensional parameter space to a low-dimensional manifold, and Riemannian metric guide optimization search is constructed based on the quality gradient; generating enhanced state representation fusing causal association and manifold geometric information; identifying a production element state and selecting a corresponding optimization strategy; a geodesic line is planned in a manifold space to obtain an optimal parameter adjustment path; historical experience is utilized through memory retrieval and case adaptation; cross-task knowledge migration is realized; adopting a depth deterministic strategy gradient algorithm to optimize the decision; and online learning is realized through elastic weight consolidation. According to the method, the problems of black box decision, slow convergence, difficulty in knowledge reuse and the like in the prior art are solved, the optimization efficiency and the interpretability are improved, and the method has the capability of quickly adapting to new tasks.

The invention discloses a water body health assessment system based on big data, relates to the technical field of water body health assessment, and realizes time synchronization, space mapping and quality assessment and restoration of multi-source data, calculation of water quality semantic indexes, data fusion, health level output and sampling work order generation. According to the method, deep fusion and standardized processing of multi-source heterogeneous monitoring data are realized, data islands are broken through, through high-precision space-time alignment and intelligent quality control, data reliability is ensured, human activity events are creatively fused for causal analysis, result interpretation is enhanced, multi-scale dynamic evaluation and adaptive threshold determination are supported, accuracy is improved, and the method is suitable for popularization and application. A closed-loop mechanism of sampling verification is evaluated, the unmanned ship is used for actively checking a suspicious area, secondary judgment is triggered, the response speed and result reliability of emergencies are remarkably improved, the whole-process compliance recording and standardized interface guarantee process is traceable, results are easy to share, and key technical support is provided for fine management of a water body.

The invention relates to the technical field of building disease detection, in particular to an urban building disease detection method and device, electronic equipment and a storage medium. Multi-modal image data formed by original visible light and thermal infrared image data is obtained, and an original thermal infrared image is subjected to geometric correction; calculating a mapping relation with an original visible light image so as to complete pixel-level registration, obtaining target multi-modal image data, inputting the target multi-modal image data into a hierarchical deep learning recognition model, recognizing building disease information, then performing three-dimensional space mapping, generating a building three-dimensional mesh model containing disease three-dimensional space setting coordinates, and finally performing three-dimensional mesh modeling. And then calculating a relationship between a model surface grid vertex and a disease point cloud density, generating a disease distribution thermodynamic diagram, analyzing disease aggregation characteristics in multiple dimensions according to the thermodynamic diagram, and quantitatively analyzing spatial correlation between the disease and a building construction node in combination with building component information. According to the invention, the urban building disease detection efficiency and precision are improved.

The invention relates to the technical field of digital twinning, in particular to a high-performance digital twinning system rendering method based on dual-grid and neural network mapping, and the method comprises the steps: constructing a high-precision calculation grid and a low-precision rendering grid, and building a space mapping relation between the two grids; executing multi-working-condition numerical simulation to obtain a high-dimensional simulation result snapshot matrix; the dimension reduction projection operator is used for reducing the dimension of a high-dimensional simulation result of any working condition into an r-dimensional feature vector; generating low-dimensional reference field data by using the space mapping relation; training the neural network until a mapping network parameter is obtained through convergence; obtaining current working condition parameters of the equipment in real time, and obtaining a current r-dimensional feature vector by using the dimensionality reduction projection operator; and quickly generating physical field prediction distribution on the low-precision rendering grid and performing pseudo-color rendering. According to the scheme, the problem that a traditional method is low in rendering efficiency is solved, and the method has the advantages of reducing computing resource consumption and keeping visualization precision.

The invention discloses an assembly process knowledge graph construction and updating method based on graph attention network mapping, and the method mainly comprises the following steps: standardization system design construction, data collection, data classification, multi-modal data preprocessing, multi-source information extraction and graph node feature initialization. Semantic modeling and vector space mapping driven by a graph attention network, knowledge storage, construction of an assembly process knowledge graph by using stored knowledge data, and semantic complementation and dynamic updating of a process information model. Through the unique self-attention mechanism of the GAT, the weight can be dynamically distributed for the relationship among different information nodes, so that knowledge reasoning and expression are more accurately carried out, and the problems of knowledge fragmentation, management rigidity, semantic understanding deficiency, low knowledge utilization efficiency and the like in the prior art are systematically solved.

The invention provides an efficient task scheduling strategy recommendation method and system for a distributed network surveying and mapping scene, and relates to the crossing field of network space surveying and mapping, network security, distributed scheduling and intelligent optimization recommendation algorithms. Analyzing the task request to extract a task type, a task target network address set, a regional distribution feature, a task execution timeliness requirement and calculation and bandwidth resources required by estimation, and forming a task feature vector; and performing state acquisition and capability evaluation on the distributed surveying and mapping nodes based on task resource requirements defined by the task feature vectors, and performing normalization processing to form node capability vectors. Through multi-source information vectorization and closed-loop optimization, adaptive matching of task scheduling and resource dynamics is realized.

The invention belongs to the technical field of electric power automation, and discloses a DTS intelligent interaction method and system fusing a large language model and a vector database, and the method comprises the steps: obtaining the text data of an electric power regulation and dispatcher training simulation system DTS field, and constructing a semantic vector knowledge base; based on the constructed semantic vector space mapping and knowledge base, performing problem reconstruction representation on the proposed problem, and then converting the problem reconstruction representation into a vector through an embedded network; retrieving in the constructed semantic vector knowledge base based on the converted vector, calculating the comprehensive similarity between the problem vector and the vector in the knowledge base, and retrieving the most relevant knowledge fragment; and driving a large language model to generate a final intelligent question and answer response by taking the most relevant knowledge fragment and a dialogue context provided by a memory mechanism as input. Efficient, accurate and intelligent questioning and answering of complex dispatching business problems are achieved, and a new artificial intelligence interaction tool is provided for training of a digital intelligent novel power system dispatcher.

The invention relates to the technical field of sensor networks and Internet of Things, in particular to a data center air quality intelligent early warning method and system based on a sensor network. The method comprises the following steps: constructing a self-organizing cooperative network by deploying multiple types of sensor nodes in a key area of a data center, and collecting and correcting multi-dimensional air data in real time; the data reliability is improved through inter-node dynamic weight fusion and local anomaly recognition; establishing an air parameter and machine room structure correlation model by utilizing space mapping and time sequence analysis, and identifying a micro-scale diffusion trend; a self-adaptive dynamic threshold mechanism is constructed, and threshold rolling optimization is realized in combination with historical statistics and real-time feedback; and generating a graded alarm strategy based on multi-stage early warning judgment and triggering conditions, and continuously self-optimizing early warning precision and response efficiency through closed-loop feedback. According to the invention, all-around, intelligent and high-reliability early warning and regulation and control of the air quality of the data center are realized.

The invention provides an intraoperative real-time navigation system based on digital twinning, and relates to the technical field of digital twinning, and the system comprises a data collection module which is used for obtaining the three-dimensional form data of a liver and a biliary tract through an intraoperative ultrasonic imaging device, and collecting the elastic modulus distribution parameters of liver parenchyma and a bile duct wall; the construction registration module is used for constructing a real-time digital twin in the liver and gall area based on the three-dimensional form data and the elastic modulus distribution parameters, and carrying out multi-mode non-rigid registration on the digital twin and the liver surface topography and the vascular structure which are subjected to optical coherence tomography in the operation; therefore, a dynamic space mapping relation between the organ structure and the image data is established. According to the method, real-time navigation in the operation is realized through digital twinning, the obstacle avoidance path is updated, and the safety and efficiency in the operation are improved.

The invention discloses a power grid scattered resource aggregation scheduling method and device based on block chain data, terminal equipment and a storage medium, and the method comprises the steps: allocating a scheduling weight for each block chain node, and converting the scheduling weight into a probability space; and then a random entropy value is generated through a verifiable random function, a historical packaging node private key and a block chain hash value so as to ensure the randomness when a packaging node is selected through the random entropy value subsequently, so that the packaging weight of the node and the qualification weight of the node participating in power dispatching in a block chain consensus process can be bound; according to the method and the system, the probability of nodes with high scheduling weights is higher through a weight and probability space mapping mechanism instead of hardware computing power, and the monopoly of packaging weights is avoided by combining generation of random entropy values, so that each block chain node is possible to obtain the packaging weights, and a final scheduling decision can realize global optimization scheduling.

The invention relates to the technical field of forestry intelligent monitoring, and particularly discloses an intelligent monitoring system for the growth condition of afforestation and greening nursery stocks based on deep learning, which is characterized in that multi-modal growth data of the nursery stocks are acquired through a multi-spectral imaging sensor, a three-dimensional laser scanning sensor and an environment monitoring sensor, and a standardized data set is formed through space-time alignment processing; extracting morphological structure and spectral response features through spatial domain and frequency domain parallel analysis, and generating multi-dimensional feature representation through cross-modal fusion; converting the parameters into growth state parameters by utilizing a feature recombination and space mapping technology; identifying an abnormal growth mode through time sequence dynamic analysis; and finally, adaptively adjusting the working parameters of the sensor according to the abnormal type to form closed-loop monitoring. The problem that an existing monitoring system cannot autonomously optimize a monitoring strategy according to the abnormal state is solved, and accurate monitoring and intelligent regulation and control of the nursery stock growth condition are achieved.

The invention relates to the technical field of industrial equipment control, and discloses an industrial equipment autonomous decision-making control method based on reinforcement learning, and the method comprises the steps: obtaining the time sequence data of a vibration sensor, the continuous data of a temperature transmitter and the discrete data of a pressure instrument of industrial equipment through a multi-mode sensing module; multi-modal data alignment is achieved through a feature space mapping algorithm, the thermal drift compensation amount is calculated in combination with a thermodynamic state model, self-adaptive threshold segmentation processing is conducted on pressure data, the processed data are input into a reinforcement learning decision model to generate a fusion decision result, exploration-utilization balance parameters are adjusted through a strategy updating mechanism, and the fusion decision result is obtained. And outputting the autonomous decision control scheme. The method solves the problems of poor adaptability, low decision-making efficiency and the like of a traditional control method, can improve the autonomous decision-making capability, the control precision and the operation stability of the equipment, reduces manual intervention, reduces the cost, and is suitable for intelligent control of the industrial equipment.

The embodiment of the invention provides a method and device for recognizing a rock mass structural plane. The method comprises the following steps: in response to an obtained photo control point geodetic coordinate and an aerial survey photo about a target area, generating a three-dimensional entity model about the target area through an image data processing rule based on the photo control point geodetic coordinate and the aerial survey photo; in response to a two-dimensional image map extracted from the aerial survey photo, inputting the two-dimensional image map into a pre-trained neural network, and identifying the two-dimensional image map by the pre-trained neural network to obtain a corresponding structural surface two-dimensional contour; associating the two-dimensional contour of the structural plane to a three-dimensional entity model through a space mapping algorithm, and generating a three-dimensional structural plane edge contour; and carrying out geometric fitting on the edge contour of the three-dimensional structural surface to obtain the structural surface size and occurrence parameters of the target area. By utilizing the method disclosed by the embodiment of the invention, the spatial orientation and size parameters of the structural plane can be accurately solved.

The invention provides a Flash wafer detection method and device based on deep learning, and relates to the technical field of semiconductor detection. The method comprises the following steps: firstly, receiving an initial electric signal response sequence which is acquired by a scanning probe and comprises reflection intensity and phase deviation; performing time-frequency domain joint transformation processing on the sequence to generate a frequency domain energy distribution map and a time domain attenuation characteristic curve; further constructing a three-dimensional space mapping model containing a frequency domain energy amplitude and a time domain attenuation time constant; a pre-trained condition generation network is called to reconstruct the model, and a reconstructed phase image set with the resolution consistent with that of a standard template is generated; and finally, performing pixel-by-pixel comparison on the reconstructed image and a standard template to generate a detection report containing defect space positioning coordinates and a contour boundary sequence. According to the method, deep mining and visual reconstruction of Flash wafer microscopic electric signal features are realized, and the accuracy and the automation degree of defect detection are remarkably improved.

The invention discloses an NPC inverter fault diagnosis method based on multi-modal feature fusion and dynamic sparse attention cooperation, and belongs to the technical field of inverter fault diagnosis. The method comprises the following steps of: firstly, respectively converting three-phase fault current signals into two-dimensional gray images through a Grubby angle difference field, and generating a color image containing three-phase characteristics by utilizing RGB color space mapping fusion; inputting the original current signal and the color image into a diagnosis model in parallel, and respectively extracting a time sequence feature and an image feature through a convolutional neural network; on this basis, a dynamic sparse attention mechanism is introduced to perform feature screening and enhancement, dynamic weight distribution and information fusion of time sequence and image features are realized through a cross-modal interaction module, and finally fault classification is completed. According to the method, the problems that a traditional method is complex in model, poor in noise immunity and weak in generalization ability are effectively solved, high precision and high robustness can still be kept under the conditions of strong noise and various variable working conditions, and the method has important engineering application value.