3861 results about "Feature based" patented technology

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical treatment and health and the like, and discloses a violation content auditing method, device and equipment based on multi-modal data and a medium. Inputting the visual semantic features and the composite audio features into a multi-modal model, generating fusion features through model alignment and fusion, and analyzing the fusion features based on a knowledge base to judge whether illegal content fragments exist in the multi-modal data, and when the illegal content fragments exist, positioning the illegal content fragments in the multi-modal data and generating an auditing report. According to the method, the visual semantic features and the composite audio features are fused, cross-modal compliance analysis is realized in combination with the knowledge base, frame-level or time-axis-level positioning is performed on the illegal content segments, and the auditing report containing the evidence is generated, so that the problems of insufficient single-modal detection accuracy and poor positioning capability are solved, and the auditing accuracy is improved.

The invention discloses a virtual historical character dialogue method and system with role knowledge and context awareness, and relates to the technical field of man-machine interaction, and the method comprises the steps: constructing a multi-level role depth model; when a question of a current user is received, identifying information of a virtual scene where the current user is located, analyzing micro-expressions of the face of the user and voice rhythm characteristics of speech of the user, analyzing an emotional state and an interaction intention of the user based on a multi-modal fusion algorithm, and generating a user state vector; executing a dynamic Prompt construction program, extracting related information from the multi-level role depth model and the user state vector, and generating a structured Prompt; and inputting the structured Prompt into a large language model, generating a reply text conforming to role features based on questions of the current user, and driving a virtual character model. The method solves the problem that in the prior art, virtual historical figures cannot provide real immersion and credible interactive experience with emotional connection.

The invention relates to the technical field of pipeline monitoring and artificial intelligence, in particular to a multi-source data fusion pipeline monitoring method and system. The method comprises the steps of performing field sorting, structure unification and risk segmentation processing by obtaining pipeline line data, historical operation archives and strategy update configuration records, and generating a session primary key configuration table; a multi-source acquisition time window is configured, an acquisition task is issued, time anchor point registration and field aperture unification are completed, and a multi-source session data packet set is generated; performing session and risk unit association, performing multi-modal feature extraction and cleaning aggregation based on artificial intelligence, and constructing a pipe network risk map structure by using a map structure data model; and calling a multi-task reasoning model and a rule component based on the atlas, and performing risk type reasoning and grade judgment to obtain a risk assessment result and a strategy updating record. According to the invention, intelligent fusion of multi-source data and closed-loop optimization based on machine learning can be realized, and intelligence and reliability of pipeline safety management are effectively enhanced.

The invention relates to the technical field of power equipment state detection, in particular to an ultrahigh-frequency partial discharge online detection system, method and device and a medium, and the system comprises the steps: collecting an initial discharge signal in real time through an ultrahigh-frequency sensor array, and carrying out the preprocessing of the initial discharge signal, so as to obtain an ultrahigh-frequency discharge signal; carrying out peak detection on the ultrahigh-frequency discharge signal, and triggering a high-speed analog-to-digital converter to collect an original waveform when the amplitude exceeds a preset threshold value; performing multi-dimensional feature extraction on the original waveform by using a digital signal processor to obtain multiple groups of dimensional features; identifying and classifying the multiple groups of dimension features based on a random forest algorithm, generating discharge type labels and confidence coefficients, and storing the discharge type labels and the confidence coefficients in a dynamic database; carrying out spatial position calculation on the ultrahigh-frequency discharge signal by adopting a time difference method, and determining a three-dimensional coordinate of a discharge source; the dynamic database and the three-dimensional coordinates of the discharge source are subjected to space-time correlation and multi-dimensional analysis, a defect analysis result is generated, and high-precision online detection of the partial discharge defect of the high-voltage equipment is achieved.

The invention relates to the technical field of bridge health monitoring, and discloses a method for judging rigidity change of a bridge structure based on bridge health monitoring deformation data. The method comprises the following steps: establishing an initial data set of bridge deformation monitoring data and performing multi-scale decomposition processing to generate deformation component data of different time scales; inputting the deformation component data of different time scales into a pattern recognition engine, and recognizing a characteristic pattern data stream associated with the structural rigidity; constructing a rigidity influence factor sequence based on the characteristic mode data flow, and calculating a statistical characteristic quantity of the rigidity influence factor sequence through a sliding time window; performing multi-dimensional matching analysis on the statistical characteristic quantity and a historical reference database, and outputting a stiffness anomaly probability index; and activating a hierarchical verification mechanism according to the stiffness anomaly probability index, and confirming a stiffness change trend through a cross validation algorithm. Reliable data support is provided for bridge structure health condition evaluation.

The invention discloses a building engineering data information intelligent management method, and relates to the technical field of information management, and the method comprises the steps: collecting and aligning construction site multi-source heterogeneous data, forming an engineering alignment data set, extracting construction state features based on the alignment data set, and calculating a comprehensive deviation index; according to the method, the multi-source heterogeneous data and the dynamic twinborn model are fused, real-time monitoring, prediction and optimization control of the construction site state are achieved, the dynamic twinborn scene is constructed, the health state mapping of the key component is generated, and the construction optimization strategy is generated based on the health state mapping and the deviation index and is controlled. The structural deviation is quantitatively analyzed, the optimal construction strategy is automatically generated, model self-evolution and parameter self-optimization are achieved through feedback learning, the safety and intelligent level of constructional engineering are remarkably improved, the energy consumption and the construction period delay rate are reduced, and conversion from static monitoring to active decision making is achieved.

The invention relates to the technical field of electrical digital data processing, and discloses a source code security vulnerability semantic detection method based on a large language model, which comprises the following steps: analyzing a source code to be detected to extract execution path constraint features, and constructing an orthogonal feature base vector sequence through orthogonalization feature extraction; inputting the source code to be tested into the large language model to obtain an initial semantic tensor; orthogonal projection is carried out on the initial semantic tensor to a code security constraint subspace constructed by an orthogonal feature basis vector sequence, weighted aggregation is carried out in combination with attention weight distribution information entropy, and a refined semantic vector is generated; according to the hidden logic offset vulnerability recognition method, business semantic noise is eliminated by utilizing a logic subspace projection mechanism, the association between a detection conclusion and code execution logic is established, and the precision of recognizing hidden logic offset vulnerabilities is improved.

Some embodiments of a method may include: decoding a motion feature by accessing a motion bitstream; predicting a predicted feature based on the motion feature and one or more reference point cloud frames; decoding a first feature representing an occupancy status of a child level voxel; predicting a second feature based on the first feature and the predicted feature; and decoding a tree voxel occupancy status of the child level voxel via the second feature.

The invention discloses an image enhancement method and system based on semantic constraint degradation modeling. The method comprises the steps that semantic masks and multi-scale degradation features are extracted based on a low-resolution image used for training; performing deep fusion on the extracted semantic masks and the multi-scale degradation features based on a double-flow parallel architecture to generate semantic-structure fusion features; forming a multi-modal guide condition, taking the multi-modal guide condition and the semantic-structure fusion feature as input together, and reconstructing a high-resolution prediction image through a diffusion generation model; constructing a structure consistency optimization total loss based on the high-resolution prediction image and the corresponding target image, and optimizing a diffusion generation model based on the structure consistency optimization total loss; and inputting a low-resolution image to be predicted into the optimized diffusion generation model to obtain a high-resolution image corresponding to the low-resolution image. According to the scheme of the invention, comprehensive and refined understanding of low-resolution images is realized through multi-module cooperation and deep fusion.

The invention relates to a computer vision and target detection technology, in particular to an enhanced target detection method and device based on feature fusion and a medium. The method comprises the following steps: acquiring aerial image data; extracting a multi-level initial feature map through an initial feature extraction network; respectively extracting local features and global features of the initial feature map of each level through a local feature extraction network and a global feature extraction network which are deployed in parallel; self-adaptively fusing the local features and the global features through a gating fusion module to generate a fused feature map, and calculating a feature competition map and generating a spatial dimension gating weight map by adopting a local-global feature competition mechanism oriented to an aerial photography scene to realize feature weighted fusion of spatial positions one by one; and performing multi-scale fusion on the multi-level fusion feature map, and finally outputting a target detection result. According to the invention, the detection precision and robustness of the multi-scale target in the aerial image are effectively improved.

The invention relates to the cross technical field of signal processing and artificial intelligence, in particular to a multi-source noise removal method and system based on DAE, and the method comprises the steps: 1, carrying out the data collection and feature extraction of multi-source noise and pure signals; step 2, constructing a de-noising recognition knowledge base based on feature analysis; step 3, constructing a deep denoising auto-encoder model based on a knowledge base; step 4, hierarchical training and optimization guided by a mixed loss function of the deep denoising model; step 5, denoising processing of a target signal and output evaluation based on a discrimination model; according to the invention, noise data in multiple fields such as electromagnetism, remote sensing and biological signals are integrated, a dynamic mixing strategy and a data enhancement technology are adopted, a training set which highly simulates a real environment is constructed, and a unique cross-scene adaptation module can perform adaptive adjustment according to signal characteristics of different application scenes; the problem that a traditional method is poor in scene adaptability is solved.

The invention discloses a wide-frequency-domain weak signal data acquisition method, system, equipment and medium, and relates to the technical field of power system monitoring and fault diagnosis, and the method comprises the steps: collecting an input signal, analyzing the signal characteristics in real time, obtaining the frequency composition and amplitude change information of the signal, and dynamically adjusting the sampling rate according to the signal characteristics. Performing noise reduction processing on the collected signals, eliminating noise interference and retaining effective signal components, performing time alignment processing on the data to form a data set with a unified time reference, extracting multi-category features based on the data set, performing fusion judgment, identifying whether the system is in a fault state, and when it is judged that a fault occurs, judging whether the system is in a fault state or not; if yes, fault analysis and positioning are executed, and an analysis result is generated and output. According to the method, fast Fourier transform analysis is carried out on the signals, the dominant frequency and harmonic components can be accurately recognized, the energy ratio can be calculated, and a reliable data basis is provided for signal feature extraction and fault diagnosis.

The invention discloses a mixed Mamb-Attention air quality prediction model based on multi-scale decomposition and a construction method of the mixed Mamb-Attention air quality prediction model. The method comprises the following steps: firstly, constructing a multi-scale decomposition module (MSD), decomposing an input time sequence into a trend term, a season term and a residual term through a parallel sliding window group, and realizing cross-scale feature fusion by utilizing group normalization and convolution; then designing a periodic pyramid module, extracting multi-level periodic features based on fast Fourier transform (FFT), and enhancing the perception ability of the model to different time scale periodic laws; the Mama branch is used for capturing long-range dependence, the self-attention branch is used for extracting a local dynamic mode, and the output of the Mama branch and the output of the self-attention branch are fused through residual connection and layer normalization; and finally, the prediction head module completes feature aggregation and result output. The model gives consideration to long sequence modeling capability and calculation efficiency, can accurately capture multi-scale dynamic change and non-stationary features in air quality data, improves the precision and stability of air quality prediction, and has good practical value and popularization prospect.

The invention relates to the technical field of biological feature recognition, and particularly provides an anti-migration PPG recognition method based on rate perception and a state space model. The method comprises the following steps: performing physiological feature front-end extraction on an original single-channel PPG signal to obtain a high-dimensional shallow feature sequence; performing double-flow cooperative processing on the high-dimensional shallow-layer feature sequence, and distributing the high-dimensional shallow-layer feature sequence to two parallel branches, namely a control flow branch and a data flow branch; in the control flow branch, an amplitude spectrum and an instantaneous physiological rate curve are obtained; in the data stream branch, acquiring a deep global feature sequence with rate invariance; obtaining multi-scale refinement features based on the high-dimensional shallow feature sequence and the deep global feature sequence; according to the multi-scale refinement features, a final biological feature recognition result is obtained, the method can actively sense the physiological rate change, and efficient nonlinear modeling can be achieved with the extremely low parameter quantity.

The invention relates to a circuit board on-line defect detection method and system, and the method comprises the steps: carrying out the synchronous collection and structural integration of multi-source technological parameters such as production line environment temperature and humidity, equipment operation states, material batches and the like, and defect detection images, and achieving the construction of large-sample original data in a production process; through standardization and de-noising preprocessing, multi-modal features are fused, and a distribution mapping model of process and defect features is established by using algorithms such as mutual information analysis and principal component analysis. Based on a feature distribution model and real-time data, a dynamic anomaly detection threshold is adaptively generated, Bayesian inference and evidence reasoning are combined, multi-level confidence levels and risk response suggestions are output, and self-learning evolution of the model and the threshold is realized through a closed-loop feedback mechanism. According to the scheme, the accuracy of anomaly detection, the response timeliness and the risk disposal intelligent level are improved, the method adapts to complex working conditions and batch changes, and the closed-loop optimization and safety control capability of the production process is remarkably enhanced.

The invention relates to the technical field of voice processing, and provides a rhythm migration method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a decoupled rhythm feature based on a source rhythm voice, and a decoupled tone feature based on the voice of a target speaker, the decoupled rhythm feature represents the rhythm of the source rhythm voice, and the decoupled tone feature represents the tone of the target speaker; the decoupled timbre features represent the timbre of the voice of the target speaker; generating a target voice vector sequence based on the text features of the target text, the voice features of the voice of the target speaker, the decoupled rhythm features and the decoupled timbre features; and synthesizing a target audio based on the target voice vector sequence. According to the method and the device, the decoupled rhythm features and the decoupled timbre features are acquired, and the target voice is generated based on the features, so that the problem of feature mixing is effectively relieved, the timbre purity of the target speaker in cross-person rhythm migration is ensured, the expressive force of rhythm migration is improved, and the synthesized audio is more natural and vivid.

The invention provides an AI-driven equipment health state assessment method and system, and relates to the technical field of intelligent operation and maintenance. The method comprises the following steps: acquiring equipment operation data, and performing time and dimension unification and quality control to form a multi-source operation data set and an environment context; generating an initial state feature based on the mechanism feature library, and obtaining a general representation through self-supervised pre-training; executing calibration learning by using a preset health label, and establishing a fusion evaluation model containing time sequence consistency and physical boundary constraint; carrying out distribution alignment and uncertainty estimation on the basis of scene differences to obtain alignment characterization and credibility scores so as to optimize a model gating strategy; performing joint mapping on the new data, outputting health index, fault probability and residual life estimation, and generating a root cause clue; lightweight online updating is executed under drifting detection, health indexes and root cause clues are written back to a mechanism feature library, early warning levels and maintenance suggestions are generated, and therefore complete-cycle intelligent sensing and self-adaptive optimization of the equipment state are achieved.

The invention relates to the technical field of multi-peak spectrum demodulation, and particularly provides a fiber bragg grating multi-peak spectrum demodulation method and system. The method comprises the following steps: extracting local spectral features based on an experimental reference spectrum to form initial atoms, and performing translation offset and normalization processing to obtain an over-complete spectral atom dictionary; performing global offset preliminary estimation based on the dictionary, and obtaining preliminary estimation values of peak sites of the measurement spectrum and the reference spectrum through cross-correlation calculation; executing constraint orthogonal matching pursuit sparse recovery based on the estimated value, and recovering atomic displacement from the measurement spectrum by using block sparsity, translation consistency and non-negative constraint; and performing wind speed inversion and calibration based on the atomic displacement, and converting the wind speed into a wind speed estimated value through a nonlinear calibration model to obtain a final result. According to the method, a dictionary based on experimental data is constructed, dependence on large-scale labeled data is reduced, a physical mechanism and sparsity prior are fused, and the problems that a traditional method is insufficient in precision and weak in generalization ability in a complex environment are solved.

The invention relates to the technical field of low-altitude equipment, in particular to a low-altitude resource intelligent scheduling method and system based on deep learning, and the method comprises the steps: collecting the real-time state and network load data of a low-altitude flight equipment group, and constructing a dynamic operation data set; generating an operation mode feature set through multi-dimensional airspace situation awareness and analysis, and performing sparse clustering division based on the feature set to form a network resource demand priority mapping table; traversing the mapping table to dynamically calculate the resource demand, determining a multi-dimensional weight coefficient, and performing high-dimensional feature dimension reduction and optimization through a mixed integer nonlinear programming solver to obtain a resource demand feature vector; constructing a resource scheduling strategy optimization model by adopting a deep reinforcement learning algorithm based on the vector; inputting real-time data into the model to execute a resource scheduling decision, and outputting a dynamic allocation strategy; simulation deduction and compliance verification are carried out on the strategy in the digital twin simulation platform, and cooperative intelligent scheduling of communication, calculation and spectrum resources is achieved.

The invention discloses a multi-modal large model hidden danger identification method and system based on feature retrieval enhancement, and relates to the technical field of computer vision. The method comprises the following steps: quickly scanning an input image through a lightweight target detection model, and positioning a hidden danger candidate area; performing feature retrieval in a pre-constructed standard hidden danger knowledge base based on the candidate region to obtain related standard hidden danger reference information; and the candidate region and the reference information are coded and fused and then input into a multi-modal large model, and a fine-grained recognition result containing hidden danger categories, position coordinates, confidence coefficients and professional text description is output. According to the method, a professional knowledge base retrieval mechanism and a multi-modal feature fusion technology are introduced, so that the problems of lack of professional knowledge and insufficient fine-grained discrimination ability in an existing hidden danger recognition method are effectively solved, and the recognition accuracy and the result specialty are improved while the detection efficiency is ensured.

The invention provides a multi-target tracking method and device based on a large model. The method provided by the invention comprises the following steps: acquiring a current frame image of a video, detecting a target in the current frame image through a detector, and generating a target detection frame and a corresponding confidence score; based on a dynamic equation of a Kalman filter, predicting the position of a track fragment in the current frame of image according to the track fragment of the previous frame of image; when the target detection frame is an effective detection frame, associating the target detection frame with the predicted trajectory fragment based on a feature coordinate matching method, and updating the position of the trajectory fragment through an observation equation after successful association; when the target detection frame is not the effective detection frame, generating a mask fragment of the target based on a mask fragment updating method, associating the mask fragment with the predicted trajectory, and updating the position of the trajectory fragment; and outputting the tracking result of the current frame, repeatedly executing the step of outputting the tracking result of each frame, integrating the tracking results of all single frames, and generating a complete tracking trajectory of all targets in the video.

The invention relates to the technical field of intelligent teaching, and discloses a teacher ability assessment and occupational development digital optimization method based on artificial intelligence, which comprises the following steps: S1, multi-modal data preprocessing: collecting and preprocessing multi-modal data related to teacher teaching activities, the multi-modal data comprises classroom video data, classroom audio data, teaching text data, teaching interaction behavior logs and student feedback data; s2, teaching behavior feature extraction: extracting teacher teaching behavior features based on the multi-modal data; s3, constructing and mapping a knowledge graph; s4, performing dynamic evaluation; s5, generating an occupational development path; and S6, updating the model. By introducing a graph neural network and a time sequence prediction model, the relationship between modeling capability nodes and time change characteristics are synthesized, dynamic quantification and interpretability evaluation of teacher vocational capability are realized, the teacher capability change trend is reflected, and a continuous and datamation evaluation result is provided.

The invention relates to the technical field of multi-modal behavior anomaly detection scheme design, in particular to a multi-modal behavior anomaly detection method and system under the condition that encrypted traffic is not decrypted. The method comprises the following steps: capturing a network encrypted traffic data packet in real time; on the premise that decryption is not carried out, multi-mode non-decryption features such as flow layer statistics, time sequence interaction, encryption handshake and context association are extracted in parallel; establishing a dynamic normal behavior baseline model of each feature based on historical data; the real-time features are compared with the baseline model, comprehensive judgment is carried out through a multi-mode correlation analysis algorithm, and an anomaly detection result is output; and when the abnormal condition is judged, the network control equipment is automatically linked for blocking. The method thoroughly gets rid of dependence on traffic decryption, realizes high-precision and self-adaptive detection and rapid automatic response to abnormal behaviors in encrypted traffic through multi-dimensional feature fusion and dynamic baseline technologies, and effectively solves the problem of failure of a traditional detection technology in an encrypted environment.

The invention relates to the technical field of power load prediction, in particular to a long-term power system load prediction method and system based on multi-scale decomposition fusion. The method comprises the steps of performing data preprocessing based on time sequence data; performing multi-scale decomposition and feature embedding on the preprocessed data to obtain a multi-scale load feature vector set; performing gating adaptive filtering and attention double-path fusion under the multi-scale load characteristics based on the multi-scale load characteristic vector set; performing independent prediction and prediction fusion on a fusion result based on a space-time attention gating mechanism; and evaluating a result after prediction fusion. According to the multi-scale prediction result space-time attention fusion mechanism provided by the invention, prediction information on different scales can be adaptively integrated, deviation caused by a single scale is avoided, the comprehensive performance of long-term prediction is further improved, and the method is suitable for various power system planning and operation scenes.

The invention relates to the technical field of power equipment operation and maintenance and state monitoring, in particular to an optical fiber transformer health state assessment method and system based on feature fusion. According to the invention, sampling signals of a plurality of optical fiber current transformers are collected and preprocessed; constructing a historical data matrix for principal component analysis, establishing a principal component space and calculating a square prediction error control limit value; constructing a simulated fault data set and projecting the simulated fault data set to a principal component space to calculate a square prediction error time sequence; performing wavelet packet decomposition on the square prediction error time sequence to extract a normalized energy feature vector, and training a one-dimensional convolutional neural network model; projecting the real-time sampling signal to a principal component space to judge whether the real-time sampling signal exceeds the limit or not; if yes, the health state level is output through the trained model. According to the invention, accurate grading and real-time online evaluation of the health state of the optical fiber transformer are realized.

The invention relates to the technical field of machine vision and intelligent quality inspection, in particular to a bolt surface tiny defect feature extraction method based on an image enhancement technology, which comprises the following steps: acquiring an original two-dimensional image sequence of a bolt surface acquired at the same imaging view angle and a preset multi-light-source angle, and calculating a surface normal vector field and a surface albedo; obtaining basic decoupling feature data including two-dimensional coordinates, a gray value, a surface albedo and a normal vector; reconstructing three-dimensional point cloud data of the bolt surface by using the surface normal vector field and a preset integral boundary condition, and determining a target topological region set and a corresponding spatial enhancement weight; enhancing the basic decoupling feature data based on a spatial enhancement weight to generate a defect risk feature map; calculating a defect risk value based on the feature intensity value, the spatial distribution, the connected region area of the high-risk feature and the spatial aggregation degree, and respectively generating an alarm instruction, a recheck instruction or a qualification instruction; according to the method, the sensitivity to hidden cracks in a weak area is greatly improved.

The invention discloses a target identification method and device in an infrared image, equipment and a storage medium. The method for identifying the target in the infrared image comprises the steps of training a deep learning identification model fusing vision and track features based on a marked target data set; time sequence visual features of the target are extracted from the continuous multiple frames of infrared images; based on the position change of the multi-frame detection frame, reversely deducing the motion trail of the target in the physical space, and extracting the time sequence change characteristics of the trail; and fusing the time sequence visual features of the target and the track time sequence change features to form a joint feature vector, and identifying and classifying the joint features of the target through a deep learning identification model. According to the method, a target trajectory feature modeling mechanism is introduced, continuous multi-frame infrared image feature extraction is combined, and visual appearance features and physical space motion features are fused, so that the distinguishing capability of targets with similar appearances such as an unmanned aerial vehicle and a flying bird in a motion behavior dimension is effectively enhanced, and the accuracy of recognizing a complex target in an infrared image is improved.

The invention belongs to the cross technical field of artificial intelligence and image recognition, and discloses a cursive script image sequence recognition method for modeling by fusing a stroke structure, and the method comprises the steps: preprocessing and normalizing a cursive script image, eliminating interference information, normalizing the form of the cursive script image, and guaranteeing the stability and effectiveness of a subsequent processing flow; based on stroke decomposition and feature coding of structural analysis, stroke-level structural features are extracted through character skeleton modeling, and cursive writing characters are converted into time series data; modeling a cursive script recognition neural network fused with structural semantics, and recognizing a stroke time sequence by adopting a BiLSTM neural network; an identification output and result optimization module is designed to ensure the final identification quality; through application of scene integration and platform deployment interface design, practical deployment of the model and butt joint of a front-end platform are realized, and a complete cursive script recognition ecology is constructed. According to the method, not only is the structured representation capability of the cursive script image improved, but also the recognition precision and generalization performance of the model on complex handwriting are remarkably enhanced.

The invention relates to the technical field of energy prediction, in particular to an energy short-term load prediction method and system based on SE-Block improved Transform. The method comprises the steps of performing reversible normalization preprocessing based on acquired multi-element load sequence data; carrying out feature extraction and fusion on the preprocessed data by utilizing improved cross-scale interaction Patching, wherein the feature extraction and fusion comprise multi-scale feature extraction, cross-scale interaction alignment, residual error correction and dynamic fusion; and performing feature screening on the fused features based on a channel attention mechanism, wherein the feature screening comprises feature response based on improved SE-Block and non-linear interaction of context vectors. Aiming at the non-stationarity of the actual load caused by the influence of meteorological conditions and user behaviors, the model accurately depicts the fluctuation details of the load curve by automatically eliminating the noise interference among multiple variables, and the robustness of the model in the multi-element load prediction of the integrated energy system is reflected.

The invention discloses a remote sensing image semantic change detection method based on semantic enhancement Transform and three-dimensional convolution, and aims to solve the problems of insufficient semantic and change information fusion and difficult narrow and long ground feature feature extraction in the existing method. According to the method, an SET-3DC network is constructed, 'feature extraction-feature interaction-semantic enhancement-change information fusion 'is taken as a core link, multi-scale dual-time-phase features are extracted through a dual-branch feature extractor (DFE), cross-branch feature interaction is realized through a channel feature exchange module (CFE), and multi-scale dual-time-phase features are extracted through a multi-scale dual-time-phase feature extraction module. A semantic branch decoder (SEAA-SD) based on semantic enhancement and axial attention strengthens local details, global semantics and narrow and long ground feature features, a change information extraction module (CIE) based on three-dimensional convolution fuses multi-source change information and aligns time-space correlation, and a network is optimized in combination with a joint loss function. Through multi-module cooperation and innovative structural design, the precision and internal consistency of semantic segmentation and change detection are improved, and the method is suitable for a high-precision remote sensing image semantic change detection scene.