245 results about "Weight prediction" patented technology

Embodiments of this disclosure can provide a system and method for training a perception model to perform an autonomous driving task. During operation, the system can obtain labeled training data comprising images captured by multiple cameras mounted at different locations on a vehicle, and the perception model can generate, in parallel, a prediction output associated with the task and a confidence score based on the labeled training data. The confidence score can indicate a level of uncertainty associated with the prediction output. The system can generate an uncertainty-weighted prediction based on ground truth indicated by the labeled training data, the prediction output, and the confidence score; compute a loss function based on the uncertainty-weighted prediction; and update the perception model based on the loss function.

The invention relates to a sea cucumber growth character recognition and measurement method based on machine vision and a measurement system thereof, and belongs to the field of image processing, and the method comprises the following steps: S1, image acquisition and preprocessing; s2, instance segmentation and morphological feature extraction; s3, measuring the length and width of the sea cucumber; s4, pixel-to-actual size conversion is carried out; s5, constructing a body weight prediction model; and S6, outputting a result. The method has the advantages that the sea cucumber image is obtained by using the high-definition image acquisition technology, the contour information of the sea cucumber is accurately extracted through the instance segmentation algorithm, and the problems of changeable and irregular sea cucumber shapes and the like are effectively solved by combining the optimized image processing technology, so that the measurement precision is remarkably improved. Meanwhile, a machine learning regression model is introduced, the weight of the sea cucumber is predicted based on various morphological characteristics, and the dimensionality and the utilization value of measured data are further enriched.

The embodiment of the invention discloses a dynamic computing power scheduling method and device based on deep learning and a medium, belongs to the technical field of cloud computing, and solves the problem that cloud computing resources are seriously wasted due to the fact that a cloud computing scheduling method in the prior art is prone to unbalanced resource allocation. Comprising the following steps: preprocessing acquired cloud computing data to generate a structured feature vector; wherein the cloud computing data at least comprises cloud node state data, computing task attributes and environment data; through a multi-head attention mechanism, task resource space-time correlation features corresponding to the structured feature vectors are extracted; the task resource space-time correlation features are input into a weight dynamic prediction module and a resource matching degree scoring module at the same time so as to carry out parallel processing of weight prediction and matching degree scoring calculation; and solving a multi-objective optimization function corresponding to computing power scheduling based on a parallel processing result, and generating a computing power scheduling instruction based on a solving result.

The invention relates to a resource elastic scaling decision-making method, system and device and a medium. The method comprises the following steps: collecting real-time operation data of a security service node, and performing multi-dimensional security index analysis according to the real-time operation data to obtain a portrait data packet; predicting the security service weight value to obtain a prediction result, performing dynamic error compensation on the prediction result to generate a corrected weight prediction value, and generating a control instruction based on the corrected weight prediction value and the active session state; and when the instruction is a migration instruction, analyzing a session state snapshot of the instruction, calling a preset kernel state locking function to lock a memory session block of a source node, obtaining incremental state change data to generate a migration snapshot packet, and performing block verification injection operation on a target node. According to the method, by integrating multi-dimensional safety index analysis, prediction error compensation and stateful transition verification mechanisms, the accuracy and response efficiency of resource elastic scaling decision making are improved, and the continuity of stateful service transition and the consistency of safety strategies are enhanced.

The invention discloses an ultrahigh-definition panoramic image adaptive HDR fusion method based on data driving, and relates to the technical field of digital image processing, and the method comprises the steps: carrying out the coarse alignment and high dynamic range fusion of a low-resolution line graph sequence, and generating a low-resolution HDR panoramic image; carrying out sharpening processing on the low-resolution HDR panoramic image to obtain a low-resolution sharpened HDR image, and meanwhile, obtaining a multi-scale residual spectrum generated by sharpening; performing up-sampling on the multi-scale residual spectrum, and performing small-sample element learning fine tuning on the weight prediction network in combination with the low-resolution sharpened HDR graph to obtain a fine-tuned weight prediction network; performing weight prediction and layer-by-layer fusion on the registered high-resolution tile sequence on a multi-scale pyramid based on a fine tuning weight prediction network model; according to the method, the sharpening processing is implemented on the low-resolution HDR panorama, and the multi-scale residual spectrum is extracted, so that the detail sensitivity and adaptability in a complex scene are improved.

The invention discloses an AI and algorithm fused intelligent geological layering system and method for geotechnical engineering investigation, and the method comprises the steps: data importing and preprocessing, data verification and gross error elimination, automatic classification of related / non-related variables, cleaning and iterative updating of layer mean values and boundaries, and hierarchical division. According to the method, the data is automatically cleaned, the layer group is verified, the main layer / sub-layer / sub-layer / interbed is divided, the layering accuracy and efficiency are improved, the artificial intelligence technology is applied, and the layering efficiency is improved. Technicians are replaced to complete the heavy work flow of drawing grass sections, manually layering, inputting soil layer information into a computer, manually judging the rationality of in-situ test, geotechnical and rock test data and layering, modifying sketches, secondarily proofreading the rationality of results and completing division, and the exploration efficiency is improved.

The invention provides a multi-modal pig weight measurement method and system based on attention fusion, and the method comprises the steps: S1, data collection: employing a depth camera to synchronously obtain an RGB image and a depth image of a pig, and converting the depth image into point cloud data; s2, multi-modal feature extraction: extracting 2D features from the RGB image, and extracting 3D features from the point cloud data; s3, feature fusion: performing cross-modal feature fusion based on an attention mechanism on the extracted 2D features and 3D features to obtain attention fusion features; and S4, MLP regression prediction: inputting the attention fusion features into an MLP weight regression prediction model, and predicting the weight of the pig through the MLP weight regression prediction model. According to the method, the 2D features and the 3D features are fused, so that the adaptability of the system in a complex scene is enhanced, and the accuracy of weight prediction can be greatly improved.

The invention discloses a logistics return transportation capacity intelligent matching system and method based on journey prediction, and the method comprises the steps: carrying out the transportation capacity departure time window prediction, journey time consumption prediction, space correction prediction, sudden factor correction prediction and bearable weight prediction of the multi-source heterogeneous data of each vehicle in the current collection period; obtaining a transport capacity departure time window, an optimal return path, return time consumption and an estimated loadable weight of the target vehicle; and generating a transport capacity label based on the transport capacity departure time window, the optimal return path, the return time consumption and the estimated bearable weight of the target vehicle, performing multi-dimensional adaptation on the transport capacity label and the order information to obtain a comprehensive score of the transport capacity label, and performing order matching and state updating operation of the target vehicle based on the comprehensive score of the transport capacity label. Accurate pre-judgment of the transport capacity state is achieved, and the invalid matching rate is greatly reduced.

The invention belongs to the field of robot control, and provides a multi-view feature fusion operable part semantic segmentation method and system, and the method comprises the steps: obtaining the point cloud data of an operable part, and generating a corresponding multi-view image; carrying out target detection, extracting features for each view angle, and obtaining a two-dimensional bounding box and a semantic tag; based on the extracted two-dimensional bounding box, processing by using SAM to obtain a foreground mask; on the basis of the extracted two-dimensional bounding box, the relevance of the same semantic target under different visual angles is captured on the global scale by using the constructed global visual angle interaction module, and the feature consistency is enhanced; and processing the fused bounding box features obtained by the global view angle interaction module by using a weight prediction network, predicting the response weight of each bounding box at each super point, combining the obtained foreground mask and the predicted response weight, and back-projecting to a 3D point cloud space through view point information to obtain a final 3D semantic segmentation result. According to the invention, the segmentation accuracy is improved.

The invention discloses a method for realizing foreground multi-modal detection based on adaptive dynamic weight fusion, which comprises the steps of multi-modal data acquisition, multi-modal data preprocessing, multi-modal feature extraction, multi-modal feature fusion, foreground detection and prediction result post-processing. The multi-modal feature fusion step comprises the following steps: a) respectively inputting the extracted RGB features and IR features into a weight prediction sub-network, wherein the sub-network generates a spatial weight map for each modal; each pixel value in the weight map represents the local confidence degree of the mode for foreground discrimination at the corresponding spatial position; b) normalizing the spatial weight maps corresponding to the modalities to enable the sum of the weights of the modalities at the same spatial position to be 1; and c) multiplying the normalized spatial weight map by the corresponding modal features element by element, and adding the modal weighted feature maps element by element to obtain a final comprehensive feature map for foreground detection. According to the method, the foreground detection precision and robustness are remarkably improved.

The invention relates to the field of fruit quality nondestructive testing, in particular to a winter peach cold storage shelf life classification and quality nondestructive testing method based on spectrum-image feature fusion and stacked model fusion, and the method is used for accurately classifying the winter peach cold storage shelf life and predicting the key quality of the winter peach cold storage shelf life based on hyperspectral imaging and deep learning technologies. The method comprises the following steps: acquiring a hyperspectral image in a wave band range of 400-1000nm, and extracting a region of interest through mask processing; phenotype data of the samples are measured and classified, and abnormal samples are removed; through spectrum preprocessing and characteristic wavelength screening, effective pixel size and area parameters are extracted, and a multi-task integration model is constructed. The model not only realizes 100% accurate classification of the shelf life of the winter peaches, but also predicts the sugar degree (R2 = 0.8451) and the hardness (R2 = 0.8798), and supports a pseudo-color visual distribution diagram and fruit diameter and weight prediction. The method is high in precision, good in stability and high in automation degree, and effective technical support is provided for quality management and market circulation after agricultural products are harvested.

The invention relates to the technical field of deep learning, and provides a mammary gland blood oxygen dynamic imaging method and system based on multispectral fusion, and the method comprises the steps: irradiating mammary gland tissues, and collecting mammary gland tissue reflected light image sequences under different wavelengths to obtain multispectral time sequence data; constructing an adaptive weight prediction network model, and dynamically predicting the fusion weight of each spectral wavelength at different moments in the future according to the blood oxygen change intensity and historical time sequence information at the current moment; performing multi-resolution time sequence modeling on the multi-spectral time sequence data, respectively extracting short-term, medium-term and long-term time sequence features, and performing weighted fusion on the features of different spectral wavelengths through a fusion weight to obtain fusion feature data; and performing inversion processing on the fused feature data, and calculating oxyhemoglobin saturation distribution of a continuous time sequence in real time to obtain a dynamic blood oxygen image sequence. According to the invention, the blood oxygen detection precision and the depth resolution are improved, and high-precision dynamic imaging of the blood oxygen saturation of the breast tissue is realized.

The invention discloses a chicken body size and weight automatic estimation method based on multi-view image fusion. The method comprises the steps of multi-view image acquisition and preprocessing, and body size estimation key point detection. Body weight estimation; visual angle fusion and parameter calculation; calculating a pixel-real distance conversion factor F through the solid circular array calibration plate, and converting a pixel distance into a real entity ruler parameter; and establishing a body size-body weight incidence matrix, inputting the body size parameter and the body weight prediction value into a full connection layer, and outputting the optimized body weight prediction value. Compared with traditional manual measurement, the method has the advantages that the measurement time consumption of a single chicken is greatly shortened, and rapid and automatic detection is realized; equipment such as a high-cost depth camera does not need to be used, hardware investment is remarkably reduced, meanwhile, labor dependence is reduced, labor cost is effectively controlled, and the batch monitoring requirement of a large-scale farm is highly met.

The invention provides an AI-based multi-user interactive language learning system and method, and relates to the technical field of interactive language learning. Real-time structured analysis of a multi-user mixed voice stream is realized through a voiceprint separation technology, a grammar error density spectrum is constructed to quantify user grammar error distribution characteristics, and a culture conflict intensity field is established; cross-cultural dialogue conflict intensity is dynamically calibrated, a semantic transition trajectory is generated, a logic fault evolution law is accurately captured, and a traditional single-dimensional static learning mode is broken through; a language adaptation index is dynamically generated on the basis of a weight prediction model, collaborative modulation of a grammar reconstruction task package, a culture tuning script and a semantic bridging task is achieved, culture conflict mediation deeply conforms to real-time conflict intensity, and dialogue fault features are accurately matched through semantic logic repair; through a three-channel task distribution mechanism of directional insertion, multicast broadcast and random allocation, efficient collaboration of personalized grammar training, immersive culture drill and distributed logic repair is ensured.

The invention relates to a unit excitation system fan energy-saving control method based on multi-parameter dynamic adjustment. The method comprises the following steps: acquiring joint regulation and control data of a target fan; inputting the combined regulation and control data into a trained dynamic weight prediction model, wherein the dynamic weight prediction model outputs a dynamic weight corresponding to each type of data based on the relevance between each type of data in the combined regulation and control data and the fan energy-saving effect of the excitation system of the target fan; based on the dynamic weight corresponding to each type of data in the joint regulation and control data, determining a comprehensive regulation coefficient; based on a preset first mapping relation table, determining a target fan rotating speed matched with the comprehensive adjustment coefficient; and controlling the target fan to operate according to the target fan rotating speed. According to the scheme, the fan energy consumption is effectively reduced.

The embodiment of the invention discloses a logistics package weight information abnormity identification method and electronic equipment. The method comprises the steps of establishing a logistics package weight database of a commodity minimum stock unit SKU dimension according to multiple pieces of forward logistics data; using the data in the database to train a commodity weight prediction model; receiving weight information of a returned goods package uploaded by a logistics service party in the process of providing a returned goods collection service, and determining associated returned goods order information and returned goods SKU information associated with a returned goods order; predicting logistics package reasonable weight interval information of the returned commodity SKU through the commodity weight prediction model; and judging whether the weight information of the returned goods package returned by the logistics service party is within the reasonable weight interval range or not, and performing abnormity judgment. According to the embodiment of the invention, the credibility of the returned goods weight data is improved, and the platform risk is reduced.

The invention relates to the technical field of semiconductor manufacturing, in particular to a semiconductor machine control method and device, a medium and a product, and the method comprises the steps: training an initial weight prediction model according to training time sequence data until a weight prediction error is not greater than a preset target value, and obtaining a target weight prediction model; inputting the test time sequence data into the target weight prediction model, and predicting a target weight; the test time sequence data comprises process parameter time sequence data, machine state parameter time sequence data and process measurement result time sequence data; according to the predicted target weight and the actual feedback data and historical feedback data of the current batch of wafers, obtaining process parameter prediction data of the current batch of wafers; and controlling the semiconductor machine to execute the target technological process on the current batch of wafers according to the technological parameter prediction data. And at least according to the dynamic change of the actual working condition of the machine, the process parameters of the machine can be dynamically tracked and predicted in real time.

The invention provides a heat pump ring fin difference control method and system based on dew-point temperature prediction, and the method comprises the steps: calculating the current dew-point temperature based on the air temperature and the relative humidity, and obtaining a future dew-point temperature sequence through linear weighting prediction based on historical dew-point data; generating a candidate control action set based on the current dew point temperature, the future dew point temperature sequence, the current fin temperature, the current opening degree of the electronic expansion valve, the rotating speed of the fan and the upper limit of the adjustment stride; screening candidate action subsets through an evaluation function; for each candidate action, using a semi-analytical system response simulator to predict a future fin temperature sequence, and calculating a difference sequence between the fin temperature and the dew point temperature based on the future fin temperature sequence; carrying out weighted calculation on a risk score for the difference value sequence of each action, and screening an action set based on a risk score threshold value; and an objective function is constructed, the action with the minimum objective function value is selected as the optimal control action, and the optimal control action is issued to the electronic expansion valve and the fan driver through PWM signals.

The invention discloses a rapid communication method based on wireless channel knowledge map construction, and relates to the technical field of sixth generation communication, the method comprises the following steps: collecting a channel parameter matrix of a user in communication, processing to generate a standard channel parameter matrix of each user, and constructing an adjacent matrix based on the spatial distance of the user; carrying out feature aggregation and relevance capture on the standard channel parameter matrix by adopting a graph auto-encoder, generating a coding feature matrix, and mapping the coding feature matrix into a reconstructed adjacent matrix to verify whether the coding feature matrix is output or not; a partition clustering algorithm based on community discovery is adopted, the coding feature vectors of all the users are iteratively merged, a terminal community set is generated, and a centroid feature matrix is calculated; a weight prediction network is adopted to carry out dimension expansion mapping on a target position, the target position is spliced with the centroid feature matrix, a weight coefficient vector is obtained through a multi-head attention mechanism and dimension expansion mapping, a channel parameter matrix of each user is weighted and summed, and a channel parameter matrix of the target position is obtained; and accurately constructing a channel knowledge map and establishing a communication process between the base station and the user.

The invention relates to a passenger flow prediction method, device and equipment based on Internet map data and a medium, and relates to the technical field of intelligent traffic data processing, the method comprises the following steps: obtaining dynamic monitoring data and historical reference data of a target area based on an Internet map, the dynamic monitoring data comprising a user track position and a real-time moving direction, and the historical reference data comprising the user track position and the real-time moving direction; the historical reference data comprises same-period passenger flow distribution and environmental influence parameters; based on the dynamic monitoring data and historical reference data, extracting time periodicity features, space transfer features and environment association features, and fusing to form a spatial-temporal feature matrix; and inputting the spatial-temporal characteristic matrix into a dynamic weight prediction model, and outputting a passenger flow prediction result including a specific region and a specific time period. By integrating dynamic monitoring data and historical reference data related to an internet map, extracting and fusing time, space and environment features and inputting the extracted and fused time, space and environment features into a dynamic weight prediction model, high-precision passenger flow prediction is realized, and a prediction result can directly support the landing of a subsequent regulation and control strategy.

The invention discloses an energy storage system residual capacity dynamic prediction method based on intelligent electric meter load data, and belongs to the technical field of energy storage management, and the method comprises the steps: obtaining the time sequence load, voltage data and external environment data of an intelligent electric meter, and constructing a probability state space of a plurality of potential energy consumption tracks; decomposing a probability state space through causal inference, quantifying a dynamic causal map of time sequence characteristics and energy consumption changes, and deducing preliminary residual capacity prediction including a confidence interval; projecting real-time intelligent electric meter data to a feature space cognitive manifold of the dynamic causal atlas, calculating geometric deviation vectors of amplitude and direction, and quantifying the geometric deviation vectors into confidence-weighted prediction deviation; and when the prediction deviation triggers the learning condition, correcting the preliminary residual capacity prediction, and recursively adjusting the dependency weight in the dynamic cause and effect atlas. According to the method, the technical scheme of integrating probability modeling, causal inference and closed-loop adaptive adjustment is adopted, and the prediction accuracy and robustness can be improved.

The invention belongs to the field of animal husbandry management, and particularly relates to a livestock online weighing error analysis method based on state classification and time-frequency characteristics. According to the scheme, firstly, an original weighing signal is subjected to modal decomposition to obtain a weight prediction value, and a reference error is calculated in combination with a static weighing value; secondly, extracting time-frequency domain characteristic parameters after self-adaptive windowing processing is carried out on the signals, and exploring the relation between the time-frequency domain characteristic parameters and reference errors in different active states; and finally, establishing a state classification model and two types of error prediction models, and performing hyper-parameter optimization on the error prediction models by using a myxobacteria optimization algorithm. In practical application, the state classification model is utilized to perform state classification on the weighing object, and the specified error prediction model is called in combination with the classification result to realize weighing error prediction. According to the invention, the problems of insufficient precision and poor generalization of the existing scheme are overcome; the method can be used for compensating the error of the online weighing platform so as to promote the development of fine breeding.

The invention discloses a service scheduling management method and system based on urban operation, and relates to the technical field of scheduling management, and the method comprises the steps: collecting multi-modal traffic data according to an urban traffic operation carrier, determining graph structure nodes and roadsides according to traffic resources in a city to form a graph structure, forming an adjacent matrix according to a connection relation, and carrying out the operation of the adjacent matrix; and constructing a node feature matrix combination to update an adjacent matrix, constructing a graph convolutional network to update graph node features, re-estimating weights for output nodes, dynamically calculating the change trend of each edge, and predicting passenger flow and route load. According to the method, multi-modal traffic data are integrated into a graph structure, the complex relation between different traffic resources in a city is captured, the load possibility of each edge in the future is directly output through edge weight prediction, a quantitative basis is provided for line resource allocation, node weight prediction reflects the passenger flow dynamic state of each station in the future, and the load probability of each edge in the future is directly output through edge weight prediction. And the generation of a finer-grained scheduling strategy is supported.

The invention provides a signal transmission method, a video encoder and a computer readable storage medium. A computer-implemented signal transmission method performed by an encoder comprises the steps of: transmitting, by a processor, a bitstream comprising weight information for predicting a coding unit (CU) to a video decoder, the weight information indicating that if weighted prediction is enabled for a bidirectional prediction mode of the CU, weighted averaging for the bidirectional prediction mode is disabled, and if weighted prediction is enabled for the bidirectional prediction mode of the CU, weighted averaging for the bidirectional prediction mode of the CU is disabled; the weight information indicates that if weighted prediction is enabled for at least one of luma and chroma components of a reference picture of the CU, weighted averaging for the bidirectional prediction mode is disabled, in which the bitstream includes a flag indicating whether weighted prediction is enabled for at least one of luma and chroma components of the reference picture, and the flag includes a flag indicating whether weighted prediction is enabled for at least one of luma and chroma components of the reference picture. The mark comprises a mark lumaweighted lxflag [i] transmitted for the ith reference picture in the reference picture list Lx, and x is 0 or 1.

The invention discloses a weighing method and device of a washing machine and the washing machine. The method comprises the steps of collecting a current value of a motor of the washing machine in the process that the washing machine runs according to a first speed; multiple target current values are extracted from the data set corresponding to the current values, and the multiple target current values are current values which completely describe at least one wave crest in the current values; acquiring a curve characteristic value of a current curve corresponding to the target current value; a weighing value corresponding to the curve characteristic value is determined through a weight prediction model, the weight prediction model is a model obtained by using multiple groups of training data through machine learning training, and each group in the multiple groups of training data comprises a sample curve characteristic value and a sample clothes weight corresponding to the sample curve characteristic value; and obtaining the weight of the clothes in the washing machine according to the weighing value. According to the invention, the technical problem of low weighing accuracy of the washing machine in the prior art is solved.

This application provides an image processing method performed by a computer device. The method includes: obtaining vertex features of vertexes in a mesh model of an object and topology information of the mesh model; obtaining vertex relationship indication information indicating vertexes having a constraint relationship between two different vertexes in the mesh model; and predicting a skin weight of the mesh model by applying the vertex relationship indication information, the vertex features, and the topology information of the mesh model to a skin weight prediction model, the skin weight of the mesh model indicating a deformation influence degree of the joint point of the object on each vertex in the mesh model in a deformation process of the mesh model. By predicting the skin weight of the mesh model, the accuracy of predicting a skinning matrix of the mesh model is improved.

The present application belongs to the field of weld weight prediction, and particularly relates to a bus steel structure weld weight prediction method and system, which comprises the following steps: reading a three-dimensional model and generating an initial weld path set; filtering, continuity repairing and logical segmentation based on process avoidance features are performed on the path to obtain an effective weld information set; the information set is input into a process decision rule library, and process specification parameters are decided according to structural features, wherein at least the large-section profile specific joint is decided as intermittent welding and the conversion rate is determined; finally, the weld weight is calculated and summarized according to the process parameters and geometric features. The method realizes the full-process automation and intelligentization from the model to the weight prediction, and significantly improves the prediction accuracy and efficiency.

The invention discloses a lightweight load prediction method and system for a distribution transformer, relates to the technical field of power systems, and solves the problem that efficient and accurate lightweight prediction cannot be carried out on the load of the distribution transformer, and the method comprises the steps: analyzing the historical load of the distribution transformer in different dates in the past; binding the historical average temperature of the power utilization area in different dates in the past with the corresponding historical load; analyzing the load condition of the distribution transformer according to the date type of the power utilization area in the current date and the real-time average temperature, and analyzing to obtain a reference load interval or a reference load of the distribution transformer in the current date; predicting the load of the distribution transformer at the next time node according to the reference load interval to obtain a predicted load interval of the distribution transformer at the next time node; and analyzing the load condition of the distribution transformer at the next time node according to the predicted load interval. According to the invention, efficient lightweight prediction of the load of the distribution transformer is realized.

The embodiment of the invention provides an electric power industrial control system security detection method and device, and the method comprises the steps: obtaining network traffic and log data, extracting traffic features and log features, predicting the weight of each expert model through employing an expert weight prediction model according to the extracted features, and predicting the detection task probability through employing a task prediction model, performing enhancement processing on the weight of each expert model by using a preset attention weight to obtain an enhanced weight of each expert model, calculating an expert matching degree according to the detection task probability and a preset expert ability matrix, calculating a correction weight of each expert model according to the expert matching degree and the enhanced weight of each expert model, and obtaining a correction result of each expert model; and based on the correction weight of each expert model, selecting a predetermined number of expert models, and performing detection by using the selected expert models to obtain a detection result. According to the method, the accuracy and the dynamic adaptability of routing decision and the accuracy and the integrity of a detection result of a multi-expert system in a complex electric power industrial control scene can be improved.

The invention discloses a dynamic updating method and device for biological characteristics, equipment and a storage medium. Comprising the steps of collecting real-time authentication features of a user, and extracting historical fusion features of the user; calculating a feature quality factor, a similarity factor, a time interval weight and a historical cumulative weight based on the real-time authentication feature and the historical fusion feature; training a fusion weight prediction model based on the multi-layer perceptron neural network model; inputting the feature quality factor, the similarity factor, the time interval weight and the historical cumulative weight into a fusion weight prediction model, and outputting a dynamic fusion weight; and based on the dynamic fusion weight, fusing the real-time authentication feature and the historical fusion feature to obtain an updated fusion feature. According to the method, the newly collected features and the historical fusion features are combined to update the feature library, the feature fusion weight is dynamically adjusted through the multi-dimensional feature factors and the nonlinear model, and the complex relationship among the factors is captured through the model, so that the weight adjustment is more flexible and accurate.