406 results about "Discriminant model" patented technology

The invention relates to a risk management and control method and system based on real-time behavior analysis, and the method comprises the steps: carrying out the structural processing of multi-source behavior data through lightweight protocol decoding and behavior label embedding, and constructing an original behavior data set of a user and an entity; extracting multi-dimensional behavior characteristics by using a sliding window analysis and sparse representation mechanism, and constructing a user behavior graph by combining graph embedding learning; constructing a time-sensitive behavior trend model through streaming modeling and an incremental learning strategy, identifying an abnormal evolution trajectory in real time, and introducing a dynamic risk threshold regulation and control mechanism; adopting a high-throughput flow data processing and fast similarity matching algorithm to construct a fusion discrimination model, giving risk levels to abnormal behaviors and classifying the abnormal behaviors; and finally, performing closed-loop optimization in combination with a historical treatment effect. The system has the advantages of high real-time performance, high calculation efficiency, adaptability to complex network environments and the like, and the network security protection capability can be effectively improved.

The invention relates to an underground equipment fault early warning and diagnosis method based on big data analysis. The method is suitable for equipment operation state monitoring and intelligent diagnosis in underground operation scenes such as mines. The method comprises the following steps: collecting multi-source data such as an equipment running state, environment parameters and operation behaviors and preprocessing the multi-source data; multiple signal features are extracted and fused to construct a unified feature vector; performing health modeling by using the residual self-encoder model to generate a health index; an early warning threshold value is dynamically set through clustering analysis and Bayesian reasoning, and anomaly recognition is achieved; after early warning is triggered, fault type identification is carried out by adopting the fusion discrimination model; performing causal reasoning and maintenance suggestion generation based on the equipment fault knowledge graph; and continuously optimizing the model in combination with operation and maintenance feedback information, and constructing a closed-loop diagnosis mechanism. The method has the characteristics of high recognition precision, high response speed, explainable result and sustainable optimization of the model.

The invention relates to the technical field of artificial intelligence and natural language processing, and particularly discloses a text content generation method based on artificial intelligence, and the method comprises the steps: obtaining natural language text input, and extracting a semantic recognition feature vector; acquiring context state data and encoding the context state data into a state recognition feature vector; generating a fusion feature vector containing a semantic and state association relationship through fusion analysis; constructing a causal discrimination model based on the fusion features, and outputting the matching confidence of semantics and states; dynamically adjusting a generation strategy according to the confidence coefficient, if the matching degree is high, generating a standard text, otherwise, triggering an error correction mechanism to output a corrected text; and finally, performing logic consistency verification on the generated text to ensure that physical constraints, technological procedures and safety standards in the industrial field are met. According to the method, by introducing multi-level feature fusion, causal reasoning, intelligent error correction and rule verification mechanisms, context perception and safety controllability in the text generation process are achieved.

The invention relates to the field of tunnel engineering geology and support design, and discloses a weak surrounding rock tunnel deformation risk discrimination method based on shear expansion-shear constitutive, which comprises the following steps: constructing a nonlinear coupling model between a shear expansion angle and shear stress, normal stress and joint parameters, and obtaining the shear expansion angle; establishing a volumetric strain rate discrimination formula; inverting an initial crustal stress tensor field; reconstructing an irregular tunnel boundary; constructing a risk level discrimination model, and outputting a risk level; supporting schemes such as supporting rigidity, anchor rod parameters and spraying layer thickness are matched according to the risk grades; establishing a model to predict a risk trend; a support adjustment suggestion is generated; collecting monitoring data to dynamically correct model parameters; and all the modules are integrated in a deployment system. According to the method, the coupling relation between the shear expansion angle and the shear strength is introduced, the coupling type constitutive discrimination model is established, the risk grading system and the support correction strategy associated with the support response are constructed, and active early warning of the high-risk section and dynamic adjustment of the support rigidity are achieved.

The invention discloses an online public opinion research, judgment, analysis, classification and grading early warning method and system, and the method comprises the steps: constructing a malicious public opinion data detection and judgment model, and carrying out the iterative training of a pre-training neural network through large-scale online public opinion historical data, so as to achieve the fine adjustment and optimization of the model; after malicious public opinion data are detected, fine-grained classification of malicious public opinions is realized through deep semantic comprehension and multi-dimensional text structure analysis by adopting BERT and ERNIE deep learning models and combining semantic role labeling and dependency syntactic analysis technologies; a malicious public opinion data grading and distinguishing model is constructed, deep correlation between public opinion events is captured through multi-layer convolution operation, and the recognition capacity for public opinion sensitivity and urgency degree is improved. According to the invention, malicious public opinion data related to a specific industry field can be efficiently and accurately identified and classified, and the malicious public opinion data is subjected to severity grading, so that the accuracy and the real-time performance of network public opinion management are improved.

The invention provides a wind power prediction method and system, and relates to the field of wind power prediction, and the method comprises the steps: collecting weather forecast data and actually measured wind power data at all historical moments, and constructing a historical data set; using a BP neural network to learn the relationship between each meteorological factor in the historical data set and the wind power to obtain a power prediction model, and performing preliminary prediction on the wind power; based on the historical data set, establishing an error inference model by using a Seq2Seq-AM algorithm, and performing prediction error inference on the preliminary prediction result; and establishing an error correction discrimination model by using an auto-encoder algorithm, and discriminating whether to use the prediction error to correct the preliminary prediction result so as to obtain the final wind power of the distributed wind power plant. Based on the Seq2Seq-AM and the auto-encoder, the prediction error is deduced, and whether the prediction error is used for correction or not is judged, so that accurate prediction of the wind power of the target station is realized.

The invention discloses a tunnel drainage pipeline blockage condition monitoring and processing method and system, and the method comprises the steps: collecting data, such as optical images, three-dimensional point cloud, ultrasonic reflection, gas components, infrared imaging, sound wave images, electromagnetic induction and flow velocity, through a multi-mode sensor array, carrying out the denoising and normalization preprocessing, and extracting multi-scale features; fusing and analyzing the feature data by using a self-attention mechanism and a graph neural network to generate optimized multi-modal comprehensive feature data; based on the data, a full-connection neural network discriminates the type, the number, the size, the blockage degree and the cleaning difficulty of blockages, and dynamic correction is carried out through real-time data feedback and a self-learning mechanism; and finally, cleaning scheme data is generated, and a cleaning task is executed through the mechanical arm. According to the method, the comprehensiveness of data acquisition, the depth of feature extraction, the intelligentization of feature fusion and the accuracy and the real-time performance of a discrimination model are improved, and comprehensive monitoring and efficient processing of the blockage condition of the tunnel drainage pipeline are realized.

The invention discloses a tank equipment sealing micro-leakage intelligent early warning and positioning method, which comprises the following steps of: deploying acoustic sensors, vibration sensors, gas concentration sensors, temperature and humidity sensors and the like aiming at different positions of a tank to realize multi-mode original signal acquisition; various signals are standardized and corrected through denoising, normalization and structural difference mapping, and the influence of the environment and the material structure on the signals is compensated; extracting multi-dimensional feature parameters, and carrying out space-time weighted fusion by adopting an attention mechanism to generate feature vectors marked by confidence; according to the method, a self-learning anomaly recognition algorithm is combined, a judgment model is continuously optimized through transfer learning and online sample accumulation, micro-leakage anomaly judgment is achieved, accurate positioning of a tank leakage source point is executed based on a physical propagation model, and the sensitivity and the self-adaptive capacity of micro-leakage detection are improved; and leakage identification and positioning requirements in a complex environment and a dynamic working condition can be effectively met.

The invention provides an HCG and LH test paper identification method and system based on optical characteristics and a storage medium. The method comprises the following steps: firstly, collecting and generating a reflectivity curve of a test paper detection area and a quality control area; secondly, extracting a reflection characteristic value of a curve in a detection area for test paper validity verification, and entering a core identification stage if the test paper is qualified; preliminary judgment is carried out by calculating a similarity deviation value of a quality control area curve and an HCG / LH standard curve, when the deviation is significant, a preliminary recognition result is generated and enters a cross validation link, and a result is confirmed by comparing a reflection characteristic value with a dynamic threshold range; when the similarity deviation is not obvious or the cross validation is not passed, starting a pre-trained discrimination model to perform final arbitration; by constructing a quadruple judgment mechanism of validity verification, preliminary recognition, cross validation and model arbitration, the operation complexity and misjudgment risk of manual test paper type selection are overcome, and the automation degree of the detection process and the accuracy and reliability of the recognition result are improved.

The invention relates to the technical field of surface defect detection, and discloses a gearbox gear defect detection method. The method comprises the steps of collecting a full-surface image of a gear, performing multi-layer feature extraction through image preprocessing in combination with a convolutional neural network, performing gridding segmentation and defect probability calculation through a Bayesian discrimination model, performing defect detection and classification on a defect-prone area and a defect-non-prone area by adopting a single-category model and a multi-category model respectively, and obtaining a defect-prone area and a defect-non-prone area. Generating a defect report; compared with the problems that in the prior art, manual detection is relied on, the speed is low, precision is insufficient, and particularly under the condition of a high-speed production line or uneven illumination, a traditional detection technology is difficult to effectively recognize small defects, real-time and high-precision gear surface defect detection is achieved by combining automatic detection based on deep learning, and the detection efficiency is improved. Therefore, the problems of missing detection and false detection in manual detection are avoided, the detection efficiency and accuracy are remarkably improved, and the method is suitable for a modern automatic production line.

The invention provides a production line product quality tendency defect determination method, medium and system, and belongs to the technical field of production digital data processing.The production line product quality tendency defect determination method comprises the steps that firstly, quality detection data is collected to construct a feature matrix, and a time sequence feature vector is obtained through singular value decomposition; and calculating a defect tendency index, and establishing association mapping in combination with a defect category vector. Then calculating the parameter utility by adopting a multi-level analysis method, and weighting the defect tendency index; a bidirectional long-short-term memory neural network is constructed as a discrimination model, a special defect screening layer is included, and dynamic identification of defect types is realized. According to the method, a defect feature database is established to store historical data, real-time monitoring and early warning of production line products are realized through model training and association rule mining, and finally early warning information is output, so that the technical problem that the quality defect tendency of the production line products cannot be accurately identified and predicted in the prior art is solved.

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 provides a cutting tool wear state identification and residual life prediction method based on a digital twinning technology, and relates to the technical field of intelligent tool health. The method comprises the following steps: constructing a cutting process physical model consistent with a machine tool-tool-workpiece system and process parameters, generating a virtual monitoring signal and a virtual degradation track, and comparing the virtual monitoring signal and the virtual degradation track with actually measured data of a multi-source sensor to form virtual-real synchronous monitoring residual features; constructing a likelihood function based on the residual error, and jointly calibrating cutter individual and material parameters through Bayesian updating to obtain an updated digital twinborn model accurately matched with the current working condition; on this basis, a corrected state discrimination model is constructed, stable recognition of the tool wear state is achieved, a random degradation process model is adopted for quantitative prediction of wear evolution and residual life, and the accuracy and robustness of tool health monitoring and life prediction are effectively improved.

The invention discloses a cheating detection method and system applied to online interview, relates to the technical field of computer vision, and solves the problems that in the prior art, multiple behavior characteristics in interview personnel are often adopted for monitoring in the whole process, multi-data analysis in the presence of non-cheating behaviors causes waste of computing resources, and the monitoring efficiency is high. And the cheating detection method is low in efficiency. A gaze anomaly coefficient is generated according to gaze data; when the fixation abnormal coefficient is greater than the abnormal threshold value, generating a multi-modal feature according to the monitoring video; generating a multi-modal decision weight according to the interview stage label; fusing the multi-modal decision-making weight and the multi-modal features to obtain multi-modal decision-making data; and inputting the multi-modal decision data into the cheating judgment model to obtain a monitoring result, carrying out tendency evaluation on cheating monitoring, and carrying out multi-modal dynamic fusion when the cheating tendency exists so as to accurately judge the cheating condition, so that the cheating monitoring efficiency and the cheating monitoring accuracy are improved.

The invention discloses an integrated energy system rapid energy flow analysis method based on physical mechanism guidance and a multi-head graph attention mechanism, and the method comprises an integrated energy system node and topological feature extraction method based on a graph neural network, which is used for constructing multi-energy flow node information and graph representation of an integrated energy system; constructing an energy flow convergence discrimination model based on a long short-term memory neural network algorithm, wherein the energy flow convergence discrimination model is used for discriminating whether the real-time energy flow to be predicted converges; an energy flow prediction method based on multi-head graph attention and physical mechanism guidance solves the problems that a traditional energy flow analysis method is low in calculation efficiency and a conventional machine learning method does not consider actual physical constraints, meanwhile, a multi-head graph attention mechanism is adopted to expand the visual field of attention, the learning ability of the relation between different nodes of a model is enhanced, and the learning efficiency is improved. Therefore, the operation efficiency and the result accuracy of subsequent energy flow prediction are ensured. According to the method, on the basis of historical operation data of a comprehensive energy system, a graph neural network is used for constructing and representing the relation between nodes and branches, a judgment model is constructed through a long-short-term memory neural network, and whether an energy flow prediction model carries out energy flow prediction or not is guided; and a rapid and accurate energy flow analysis result can be provided for operation scheduling of the integrated energy system.

The present invention belongs to the technical field of aviation component mold production, and in particular, relates to a PIID coating detection method for aviation component molds, and the steps are as follows: full coverage scanning and detection of the surface of the aviation component mold; comprehensive detection of various defects according to different requirements of detection sensitivity and resolution for different coating defect types; training a large number of qualified and unqualified coating sample data to obtain a discriminant model for coating performance, inputting the coating data obtained by the detection into the discriminant model, and quickly and accurately identifying unqualified coatings. The present invention uses machine vision to obtain three-dimensional point cloud data and reconstruct the model, determine the optimal detection path, and achieve full coverage scanning; use machine learning to train the discriminant model to achieve rapid and accurate identification of coating quality, effectively solving key problems such as blind spot elimination, comprehensive defect detection and performance evaluation in mold coating detection, and significantly improving detection efficiency and accuracy.

The invention discloses a single-channel speech enhancement method and device based on a mean-reversal Schrodinger bridge, and the method comprises the steps: forming a sample pair set through a pure speech sample and a noisy speech sample, carrying out the preprocessing and Fourier transformation of the speech samples, and obtaining a frequency spectrum complex matrix pair; constructing discriminant models, estimating a frequency spectrum complex matrix of the pure voice, adjusting model parameters according to the difference between the frequency spectrum complex matrix of the pure voice and the frequency spectrum complex matrix of the real pure voice, and taking the discriminant model corresponding to the optimal parameter as a target discriminant model; constructing a score model, estimating a reverse optimal migration score in a reverse generation process, adjusting model parameters according to the difference between the reverse optimal migration score and a real reverse migration score, and taking the score model corresponding to the optimal parameter as a target score model; and giving noisy voice, inverting the reverse optimal offset score of the Schrodinger bridge by using the trained target score model parameterization mean value, and carrying out a reverse generation process to generate pure voice. The method is lower in calculation cost and better in speech enhancement effect.

The invention relates to a target object detection method and device, computer equipment and a readable storage medium, and the method comprises the steps: inputting a to-be-detected image into a target detection model for preliminary detection, and obtaining a preliminary detection result of a target object; inputting the preliminary detection result into a category discrimination model for prediction to obtain a probability score of a non-target object; comparing the preliminary detection result with a pre-stored feature queue to obtain a similarity score; the feature queue is used for storing non-target features in the historical detection result; correcting the probability score based on the similarity score to obtain a target score; determining a false detection result from the preliminary detection result based on the target score; and the false detection result is eliminated from the preliminary detection result to obtain a final detection result, so that the problem of poor target object detection effect is solved, the false detection recognition accuracy is improved, and the quality of the target detection result is improved.

The invention discloses a disturbance of consciousness electroencephalogram discrimination method and system adaptive to channel deletion. According to the method, on the basis of an electroencephalogram classification model of a deep network, automatic discrimination of disturbance of consciousness (MCS and UWS) is carried out with high accuracy according to resting-state electroencephalogram energy of a patient. The model utilizes a time domain double-branch CNN module and a space-time Transform module to learn features in electroencephalogram data hierarchically; and a model architecture with variable channel dimensions is introduced, an inter-channel association learning method fusing electrode position information and a training strategy of channel random discarding are introduced, so that the model can adapt to a scene in which part of channels are missing, and the clinical applicability of the method is improved.

The invention provides an ewe oestrus detection method and device based on multi-modal feature fusion, and relates to the technical field of livestock behavior monitoring, the method comprises the following steps: acquiring behavior data and audio data of an ewe to obtain a first behavior data set and a first audio data set; inputting the first behavior data set into an ewe behavior recognition model to obtain a second behavior data set, and inputting the first audio data set into an ewe sound detection model to obtain a second audio data set; performing feature extraction on the second behavior data set and the second audio data set to obtain a multi-modal feature fusion data set; and inputting the multi-modal feature fusion data set into an ewe oestrus discrimination model to obtain an ewe oestrus detection result. According to the ewe oestrus detection method based on multi-modal feature fusion, on the basis of fully learning the multi-modal features of the ewe, whether the ewe is oestrus or not is intelligently judged through the ewe oestrus judgment model, and the accuracy and efficiency of ewe oestrus detection are improved.

The invention belongs to the technical field of software engineering, and particularly relates to a unit test method for generating a discrimination model based on retrieval. Searching a to-be-tested function, namely a test case pair, and establishing a code library; constructing an efficient code retriever based on code embedding and mixed feature extraction; constructing a multi-input matching discriminator network, preparing an own data set, and training the network; integrating the code library, the retriever and the discriminator into a local knowledge base, and performing fine adjustment on the large model based on the local knowledge base to obtain a retrieval generation discrimination model; and generating a test case based on the retrieval generation discrimination model, and testing. According to the method, the discrimination model is generated by using retrieval, corresponding change is automatically carried out according to the previous test case, and the test with higher coverage rate and quicker speed can be realized.

The invention discloses a method and device for constructing a discrimination model based on solid waste fingerprint features, and belongs to the field of solid waste discrimination. The method comprises the following steps: acquiring a plurality of groups of solid waste samples of a predetermined category in a predetermined area, and selecting a K alpha spectral line to analyze the oxide content of the sample as various fingerprint factors to obtain a plurality of groups of preliminarily screened fingerprint factors; according to the statistical magnitude of the multiple groups of preliminary screening fingerprint factors, removing the fingerprint factors which do not meet the screening conditions to obtain secondary screening fingerprint factors; on the basis of the fingerprint factors screened for the second time, new fingerprint factors are introduced one by one according to a preset principle, and an optimal fingerprint factor combination is screened out; and constructing a discrimination model based on the optimal fingerprint factor combination, inputting the solid waste sample of the to-be-discriminated category into the discrimination model to obtain a discrimination score, and obtaining the category of the solid waste sample of the to-be-discriminated category according to a discrimination rule of a predetermined category and the discrimination score. According to the invention, unknown solid wastes can be accurately and rapidly identified.

The invention relates to the technical field of fire safety monitoring, and discloses a linear fire detection method and system based on infrared spectrum analysis and an infrared spectrum detector, and the method comprises the steps: arranging a plurality of infrared spectrum detector nodes; the collected infrared spectrum signals are preprocessed; mapping the preprocessed spectral signal to a pre-constructed low-dimensional feature space; calculating the conditional probability density of the current infrared spectrum signal in various states; judging whether the posterior probability exceeds a set fire judgment threshold value or not; the system comprises an infrared spectrum detection module, a preprocessing module, an embedded mapping module, a discrimination module and a linkage control module. The infrared spectrum detector comprises a spectrum sensing unit, a signal processing unit, a data communication unit and a data uploading interface. According to the method, the technical effects of keeping high sensitivity and low false alarm rate under complex and changeable working conditions are achieved through a combined discrimination model of embedding mapping, kernel density estimation and Bayesian posteriori.

The invention discloses a liquid qualification detection system and method, and belongs to the technical field of intelligent detection and data analysis, multi-dimensional response signals of to-be-detected liquid under excitation of different frequencies are obtained, and a multi-dimensional response data set containing optical, acoustic, electrical and thermal physical channels is constructed; extracting a liquid component behavior characteristic spectrogram through a multi-modal characteristic fusion module; a liquid qualification judgment model is constructed based on a clustering enhanced convolutional neural network and a multi-scale attention mechanism, and classification prediction of the liquid qualification state is realized; further combining confidence coefficient calculation with historical sample database comparison, performing adaptive calibration on a prediction result, and outputting a detection result and a risk level; if the detection result is unqualified or the risk level exceeds the threshold value, an alarm mechanism is automatically triggered, and abnormal feature parameters are marked; the method has the advantages of high detection precision, high response speed, high anomaly recognition capability, excellent traceability and the like, and is suitable for scenes of liquid quality control, environment detection, biological sample analysis and the like.

The invention provides a cross-omics sparse feature selection system and method based on hierarchical causal modeling, and the system comprises a data input and preprocessing module which is used for receiving multi-omics original data of a multivariate sample; the hierarchical causal structure learning module is connected with the data input and adaptive preprocessing module and is used for constructing a cross-omics hierarchical causal topology; the causal-oriented sparse feature selection module is connected with the hierarchical causal structure learning module; and the model retraining and integration module is used for constructing a three-layer weighted integration discrimination model based on the screened markers, optimizing the fusion weight of each layer through a gradient descent algorithm, and outputting a final prediction result. According to the method, the protein-metabolism biological hierarchy relationship and serum-urine complementary information are fully utilized, and the method has good generalization ability and can be widely applied to marker mining and prediction modeling of cancers, metabolic diseases and the like, so that the accuracy and reliability of precise medical treatment are improved.

The invention provides a low-voltage flexible DC system fault detection method based on complex domain analysis, and the method comprises the steps: obtaining and sampling a transient current signal in a low-voltage flexible DC system, fitting the transient current signal into a linear combination of a group of exponential functions, and obtaining a fitting exponential function; solving the fitting exponential function in a Z domain based on a Pade approximation method to obtain a complex index of the fitting exponential function; determining a fault threshold frequency according to the operation parameters of the low-voltage flexible DC system; the complex index is used as an analysis object, and the natural oscillation frequency of the low-voltage flexible direct-current system is obtained through formula calculation; and taking the fault threshold frequency as a state circle radius, constructing a fault discrimination model based on a complex plane, comparing the natural oscillation frequency with the fault threshold frequency, and outputting a fault detection result of the low-voltage flexible DC system in combination with a real part of a complex index. According to the method, the sensitivity to noise can be reduced, and meanwhile, the requirement of fault detection for speed is met.

The invention discloses a smart city public facility supervision method based on the Internet of Things, and the method comprises the steps: carrying out the multispectral collection and space marking of a surface block of a to-be-detected circuit board, and achieving the standardized preprocessing and physical feature extraction of multi-label high-dimensional pixel data; according to the method, material spectral features and surface texture data are combined, a clustering analysis and supervision discrimination model is used for accurately discriminating the type of a block material, material attributes and multi-spectral features are fused, and pixel-level thickness prediction and dynamic adaptive correction are realized through a machine learning model. The system can automatically monitor drifting of the material and the mapping relation, and performs real-time optimization on the mapping model based on periodic calibration data.

The invention discloses a complex nonlinear MIMU (Micro Inertial Measurement Unit) signal noise reduction method based on improved modal decomposition. Dynamic adaptive optimization of noise parameters is realized by introducing Hilbert spectrum analysis into a CEEMDAN (Complex Empirical Empirical Mode Decomposition Number) algorithm; an IMF component discrimination model is constructed based on multi-dimensional feature fusion, and accurate classification of IMF components is realized; and aiming at a discrimination result, adopting a hierarchical processing strategy of combining variational mode decomposition and empirical wavelet transform for different types of mode components to realize high-quality signal reconstruction. Compared with the prior art, the method has the advantages that the signal decomposition quality is remarkably improved, the multi-feature fusion discrimination model is excellent in performance when the boundary fuzzy region is processed, the problem of discontinuity of a traditional hard threshold method at the feature boundary is solved, the signal-to-noise ratio is remarkably improved, the root-mean-square error is greatly reduced, and the noise reduction effect is remarkably improved.

The invention discloses a fault positioning optimization method and system for relay protection of a new energy power grid, and belongs to the technical field of fault positioning optimization. The system comprises a data acquisition module, an initial model generation module, a feature set optimization module, a hyper-parameter optimization module, a model optimization module and a protection control module. The particle swarm optimization algorithm is utilized to perform hyper-parameter global search and optimize the penalty coefficient and the kernel parameter, then local optimization is performed through the genetic algorithm, the penalty coefficient and the kernel parameter of the SVM fault judgment model are cooperatively optimized, the particle swarm global search and the local optimization capability of the genetic algorithm are fused, so that the hyper-parameter of the support vector machine is efficiently adapted, and the fault judgment accuracy is improved. The nonlinear fault classification robustness is remarkably improved, and the problems that in the prior art, the model parameter optimization efficiency is low, and local optimum is prone to occurring are solved.

The invention provides a generation and discrimination collaborative optimization industrial visual detection system and method, and relates to the field of computer vision and artificial intelligence. The decoder comprises a unified encoder, a reconstruction decoder, a condition generation decoder and a discrimination head; the unified encoder is used for extracting general visual features based on the to-be-detected image; the reconstruction decoder is used for obtaining a reconstruction image without defects; the condition generation decoder is used for generating a defect image containing a specified defect; the discrimination head is used for identifying the reconstructed image and outputting confidence based on an identification result; training is carried out based on a defect image generated by a condition generation decoder, and a defect category and a category probability are generated; and the collaborative optimization module is used for constructing a collaborative loss function, a task loss function and a total loss function, reversely updating a discrimination head parameter based on a gradient generated by the total loss function, and generating a decoder parameter and a unified encoder parameter. The joint evolution of the generative model and the discrimination model is realized, and the performance of industrial visual inspection can be effectively improved.