53 results about "Identifier scheme" patented technology

The invention relates to the technical field of electroencephalography, in particular to a sleep staging automatic identification method and system based on multi-mode electroencephalography, and the method comprises the following steps: extracting a waveform period to construct dominant distribution, correcting label boundary positioning mutation response, unifying a time axis to form an alignment structure, clustering mutation fragments to establish an alternating relation, and identifying the sleep staging based on multi-mode electroencephalography. And identifying a main channel range to construct an automatic identification scheme. According to the method, a reference channel is positioned through area distribution difference, a boundary change section is judged in combination with a frequency band energy change trend, a channel region with dense and stable mutation points is extracted as an anchor point, time axis unified adjustment is completed according to main response starting and ending time, and an alternating fragment graph is further constructed through a cross-channel time coverage relation of the mutation points. The sequential structure coordination ability and mutation form aggregation expression efficiency among multi-source signals are improved, and accurate recognition of boundary drift and asynchronous response and dynamic extraction of steady-state rhythm in sleep stage division are achieved.

The invention discloses a human and house data comprehensive treatment method, system and equipment based on a multi-level address model, and a medium, belongs to the technical field of data treatment, and aims to solve the technical problem of how to effectively integrate human and house data sources of multiple parties and improve the accuracy, consistency and availability of human and house data. The defects that in the prior art, people and house data are scattered, address description is not standard, the incidence relation is inaccurate, data quality is uneven and a unified identification system is lacked are overcome. According to the technical scheme, the method comprises the steps that a five-level core data table comprising a community table, a building table, a house table, a people and house relation table and a personnel information table is constructed; the method comprises the following steps: accessing original human and house data from a plurality of data sources, performing intelligent analysis and standardization processing on address information, and obtaining data after standardization processing; establishing a house unique identification system and a human-house multi-dimensional relation model based on the standardized data; evaluating and repairing the data quality; and standardized human and house data service is provided for the outside through the data service interface.

The invention discloses an open source environment software hidden vulnerability patch identification method, which is based on a multi-stage architecture and collaborative relationship modeling, and provides a new patch group identification scheme: firstly, obtaining candidate code submission from an open source warehouse, extracting correlation characteristics of vulnerabilities and candidate code submission, including rule-based characteristics and semantic characteristics, calculating a correlation score and screening high-correlation submission; pairwise pairing the high-correlation submissions, and fusing multi-dimensional features to predict a cooperative relationship between the submissions; and finally, constructing an undirected graph based on the correlation score, dividing a maximum connected sub-graph, fusing the features in the group through maximum pooling, calculating the correlation with the vulnerability, and outputting an optimal patch group. Meanwhile, an existing patch identification technology based on sorting learning is combined, an enhancement method based on a submission cooperative relation is provided, ranking logic submitted by candidate codes is updated through internal association between code submission and by means of correlation between group vectors and vulnerabilities, and the identification precision in a multi-patch scene is improved.

The invention discloses a five-axis machine tool geometric error identification method based on single-axis and multi-axis measurement, and belongs to the field of numerical control machine tool precision detection. Error kinematics and multi-axis motion models are established based on a multi-body system theory, and error identification models of a linear axis and a rotating axis are respectively constructed. On the aspect of a linear axis, an identification method based on measurement of a positioning error and a straightness error is provided, two measurement / identification schemes of single-axis motion and two-axis linkage are designed, and 18 linear axis geometric error parameters related to the position and 3 linear axis geometric error parameters irrelevant to the position are obtained through calculation. In the aspect of a rotating shaft, 12 rotating shaft geometric error parameters related to the position and 4 rotating shaft geometric error parameters irrelevant to the position are obtained on the basis of an existing linear shaft result based on a measurement and identification scheme of a double-ball-rod instrument in three directions and six positions. According to the method, positioning and straightness error detection are completed in combination with auxiliary equipment such as a laser interferometer and a double-ball-rod instrument, and the method has good universality.

The embodiment of the invention discloses a target recognition method, device and equipment, a medium and a computer program product. The method comprises the steps of obtaining continuous multi-frame point cloud data corresponding to each target object in a target space; dividing the point cloud data corresponding to the target range according to the center of the point cloud data of each target object in each frame; based on the point cloud features corresponding to the divided point cloud data, generating target features corresponding to each target object in each frame; and performing target identification processing on each target object based on the target features to obtain a target identification result. According to the target identification scheme provided by the invention, the point cloud is associated with the target object, so that each target has exclusive point cloud information, identification is carried out according to the point cloud information of each target, the effect of tracking and identifying each target in the space is realized, multiple targets in the space are accurately identified, and the accuracy of target identification is improved.

The application provides an AI-driven software project full life cycle intelligent management method and system, relates to the technical field of software project management, and comprises the following steps: realizing demand structural analysis by fusing NLP and a knowledge graph, automatically supplementing implicit demand and compliance requirements; completing task automatic disassembly, work hour estimation and resource dynamic scheduling based on reinforcement learning; identifying risks in real time by fusing multiple source data and generating a treatment scheme; completing online verification according to an optimal path and realizing data closed loop self-optimization; the system comprises five modules, namely, demand analysis, task management, risk control, online verification and data storage, and breaks data islands by means of a unified data platform, a distributed message queue and a globally unique identification system. The application effectively solves the problems of demand deviation, progress lag, risk lag and data island, significantly improves demand accuracy, management efficiency, risk response speed and delivery quality, and is suitable for intelligent management of software project full life cycle in multiple fields.

The invention provides a dynamic identification generation and updating method for a data cross-border service, a switch and equipment, which realize breakthrough in the aspects of structural design, state management, scheduling strategy, propagation mode and the like of a data identification mechanism. A complete dynamic identification system supporting multi-service isolation, state sensing updating, non-intrusive propagation and boundary analysis auditing is formed, the method is particularly suitable for cross-domain and cross-border data application scenes with high requirements for supervision intensity, data identification precision and real-time performance, and the safety of data propagation is guaranteed. On a critical path of data cross-border transmission, identifiability of service identities, updatability of state information, controllability of identification behaviors and traceability of propagation paths are realized, and capabilities of cross-domain data streams in aspects of compliance supervision, behavior auditing, abnormity traceability and the like are remarkably improved; the method is suitable for data security management requirements in multi-service complex scenes such as government affair cloud, financial science and technology, industrial internet and the like.

The invention discloses a multi-industrial equipment-oriented ELM fault identification method fusing feature attention and manual LU decomposition, and belongs to the technical field of industrial equipment fault diagnosis. The core of the method is that a feature attention mechanism and a manual LU decomposition module are fused: firstly, the feature contribution degree is quantified through a channel attention mechanism, key fault features are automatically focused, weak correlation feature interference is inhibited, and the quality of a hidden layer output matrix is improved; secondly, the manual LU decomposition module realizes autonomous controllability and numerical stability of matrix inverse operation through row principal component selection and regularization optimization, and is separated from the dependence of a third-party math library. According to the method, the accuracy of equipment fault identification is remarkably improved, the omission ratio and the false alarm rate are effectively reduced, memory limitation of edge equipment is overcome, and an efficient and reliable real-time identification scheme is provided for bearing abrasion and gearbox abnormity of a wind power variable pitch system and faults of other industrial equipment such as a machine tool and an industrial motor.

The invention relates to the technical field of high-altitude platform station operation and maintenance management, and provides a multi-load airship platform OMC system equipment identification and management method, which comprises the following steps of: distributing an equipment identification number for each managed equipment on an airship platform, pre-defining a unified management model for each equipment type, and adopting a layered micro-service distributed system architecture, an equipment management function is realized through cooperation among layers; and in the equipment access adaptation layer, a JavaScript-based plug-and-play plug-in framework is adopted, and a JS script corresponding to an equipment type is dynamically loaded and executed, so that protocol adaptation and format conversion of a management command and reported data are realized. The core problem of unified management of heterogeneous equipment in a multi-load airship platform is solved by introducing a hierarchical equipment identification system, a model-driven unified management framework, a micro-service architecture and a JS plug-in adaptation technology.

The invention discloses a power transformer parallel operation multi-modal data optimization storage method, which relates to the technical field of power data storage, and comprises the following steps: constructing an association identification system based on power transformer parallel operation characteristics; establishing an event-triggered time sequence document collaborative storage model which comprises three layers of data models; scene event driven synchronization is realized based on power grid event triggering, and a parallel operation characteristic parameter out-of-limit event is monitored as a synchronous triggering condition; when an event is triggered, scene identification change information in the event is analyzed through an Flink engine, the time sequence layer and the document layer are linked to execute dynamic storage strategy recombination, and cross-library associated corresponding data is injected. According to the invention, cross-device data association can be realized, traceable linkage of high-frequency measurement and scene parameters of all transformers in a parallel operation group is realized, and meanwhile, the data storage efficiency is improved through an event-driven synchronization mechanism.

The invention discloses an aquatic organism resource identification and traceability method and system based on image identification and deep learning, and is applied to the technical field of aquatic organism resource traceability. The method comprises the following steps: obtaining an image of an aquatic organism resource, sequentially carrying out ROI region extraction, attitude standardization and key point coordinate extraction and identification, establishing a data file of the whole life cycle of the aquatic organism resource, and storing traceability information to a cloud platform; and assigning a unique identifier to each batch of aquatic organism resources and associating the unique identifier with the traceability information, reading all the traceability information of the corresponding batch of aquatic organism resources through the identification identifier, performing identification through the shot image, and determining whether the identification information is consistent with the traceability information recorded in the traceability information. According to the method, the overall calculation amount is greatly reduced through a staged identification scheme, so that real-time processing in the aquatic organism resource transaction process is easier to realize, and the traceability information is ensured to be real and reliable in combination with a data verification technology.

The invention discloses a large language model intellectual property protection method and device based on domain-specific prejudice feature fingerprints, and belongs to the cross technical field of artificial intelligence safety and intellectual property protection. The method specifically comprises the steps that a three-dimensional orthogonal prejudice data set is constructed, and the data set depicts model prejudice features from the three dimensions of the field, the prejudice type and the response mode; screening the three-dimensional orthogonal prejudice data set based on a two-stage screening mechanism so as to extract core sample data with intra-family high stability and inter-family high discrimination; performing domain-specific prejudice feature fingerprint extraction based on the core sample data to obtain a standardized reference fingerprint; and performing ownership judgment on the to-be-tested model by adopting a confrontation verification mechanism based on the standardized reference fingerprint. According to the method, a quantifiable and high-robustness ownership identification scheme can be provided for the closed-source large language model, and the method has important application value for solving the safety problems of infringement and embezzlement of intellectual property of the large language model and the like.

The application belongs to the technical field of image processing, and particularly relates to a crack image segmentation method and system. The robustness of a model to noise is improved by introducing Gaussian noise into image data for enhancement processing. An encoder composed of three layers of progressive stacked residual units is used to extract crack features, and a DropBlock regularization module layer is used to reduce feature redundancy and enhance feature diversity. The image space resolution of the crack features is restored by a decoder symmetrical to the encoder, and the fine structure of the crack boundary is restored. Finally, a dense conditional random field is used to optimize the crack probability distribution map, and the final crack segmentation map is output. The application effectively alleviates the foreground-background class imbalance problem, reduces false positives and false negatives, improves the segmentation performance of the model, and is particularly suitable for deployment on edge devices such as unmanned aerial vehicles and mobile phones, providing a high-precision, low-complexity and strong-robustness crack identification scheme for infrastructure health monitoring.

The invention belongs to the technical field of information, and particularly relates to a method and a device for analyzing data resources based on IPv6 (Internet Protocol Version 6) address identification. The method comprises the following steps: identifying a data resource by using an IPv6 address to form a data resource identifier based on the IPv6 address; and analyzing the data resource identifier based on the IPv6 address by using a DNS (Domain Name Server) system, and querying a target data resource according to an analysis result. According to the method, the massive advantage of the IPv6 address is fully exerted, data resource identification is realized based on the IPv6 address, and the reliability, the safety and the stability are higher than those of a newly established identification system; addressing analysis is achieved based on an existing network architecture and a DNS system, the existing network architecture does not need to be transformed on a large scale, the method has the advantage of being capable of being popularized and deployed on a large scale, and the problem that data resources cannot be accessed due to the fact that IP addresses are not changed in time due to the fact that geographic positions of data are frequently replaced can be solved.

The application discloses a kind of holographic fire management systems and methods, belong to computer and wisdom fire-fighting technical field.The method includes: obtaining fire-fighting facilities three-dimensional point cloud data and establishing equipment unique identification system;Collect equipment operating parameters;Real-time data correlation mapping is constructed, and sensor data is associated with three-dimensional digital model by message queue middleware and edge computing unit;Three-dimensional model and operating parameter are superimposed and displayed in actual equipment position by holographic projection or augmented reality module;Response user's gesture recognition and voice instruction bimodal interaction;Based on equipment logical correlation rule, the linkage relationship between water system, electrical system and control system is exhibited in different visual coding in three-dimensional holographic model.The application also discloses a kind of holographic fire management systems.The application solves the problem that space position and state data are separated in traditional fire management, and linkage logic is difficult to intuitively present, improves the visual management and operation and maintenance efficiency of fire-fighting system.

This invention discloses a method and system for automated generation and optimization of database table structures. The method specifically includes: receiving the user's backend project source code; extracting a semantic information network graph related to data persistence through static code scanning; inputting the semantic information network graph into an intelligent parsing engine to analyze code logic relationships and obtain index recommendation strategies and potential table structure design flaws; based on the index recommendation strategies and potential table structure design flaws, identifying sensitive fields and recommending encryption or de-identification schemes to obtain a first table structure; performing automated stress testing on the first table structure and adjusting the index strategy and partitioning scheme based on the test results; and optimizing the first table structure according to the adjusted index strategy and partitioning scheme to obtain a second table structure. This invention achieves intelligent generation and optimization of database table structures, improving development efficiency, enhancing maintainability, improving user experience, and ensuring data security.

The application discloses a power distribution network cable joint fault monitoring method and system based on voiceprint analysis, and relates to the technical field of power distribution networks.The method comprises the following steps: performing monitoring urgency evaluation of partial discharge fault monitoring based on predicted material state information and predicted scene state information of a cable joint area in a preset time zone, and setting an adaptive identification time length; taking maximum fault identification accuracy and approximation of the adaptive identification time length as scheme optimization targets, and outputting an optimal fault identification scheme; periodically acquiring original acoustic signals of the cable joint area, and performing signal processing and fault identification on the original acoustic signals according to the optimal fault identification scheme, and outputting a fault monitoring result.The technical problem that the existing cable joint fault identification scheme lacks a dynamic adjustment mechanism, cannot be dynamically adjusted according to real-time equipment states and operating environments, results in low detection efficiency, high missed detection rate and inability to realize real-time monitoring is solved.

The application discloses an OIC product full-life-cycle type history traceability data management system, relates to the technical field of data traceability, and comprises a type main data control module, a multi-source data integrated collection module, an assembly pedigree construction module, a history correlation management module, an intelligent traceability engine module, a data security and archiving module and a comprehensive application service module. The application realizes intelligent collection, automatic fragmentation and invalid data elimination of multi-source data by constructing a unified type identification system and a full-cycle digital main line. A hierarchical assembly pedigree is established based on a product unique identification, a state characteristic matrix is formed by fusing multi-dimensional indexes, and risk marking is completed. History data is cleaned and correlated to construct a complete traceability chain, the pedigree structure is dynamically corrected in combination with recall risk and material risk, efficient forward and reverse intelligent tracing is realized, data security, permission control and long-term archiving are simultaneously strengthened, and comprehensive application service is supported.

The invention discloses a virtual machine template library construction method and system for active medical instrument software detection, and relates to the technical field of active medical instrument software detection, a targeted classification system and a four-layer basic template frame are constructed, exclusive dependence, detection tools and compliance configuration are built in, the standard requirements are highly met, and the detection efficiency is improved. The adaptation precision of the template and medical instrument software of different types and versions is greatly improved, the consistency of the detection environment is guaranteed, and the reliability of the detection result is enhanced. Automatic template generation, a multi-dimensional retrieval interface and a full-life-cycle dynamic management mechanism replace tedious manual configuration and screening, the labor cost of environment deployment and template maintenance is remarkably reduced, and the retrieval and management efficiency is improved. The new, iteration and archiving process of the unique identification system and specification avoids the disorder and failure of the template, realizes the standardized management and control of the template library, meets the requirements of multi-scene rapid detection, and provides efficient, compliant and stable environment support for the software detection of the active medical instrument.

The invention provides a partial discharge signal automatic positioning method and system and a storage medium, and belongs to the technical field of signal processing, and the method specifically comprises the steps: determining the partial discharge abnormal data of different switch cabinets according to the detection data of the historical partial discharge data of the switch cabinets, determining an identification scheme of the setting target of the partial discharge online monitoring device based on the partial discharge abnormal data and the potential setting target, and according to the change condition of the setting target and the identification results of the setting target under different partial discharge abnormal types, determining that the reliability degree of monitoring analysis processing does not meet the requirement. According to the method, the monitoring result of the partial discharge exception type in the set target is determined, and the automatic positioning processing method of the exception type is determined in combination with the positioning deviation conditions under different partial discharge exception types, so that the accuracy of positioning processing of the partial discharge source is improved.

This application provides a method, apparatus, device, and medium for identifying and processing recyclable materials. In this method, image data and weight data of the item to be recycled are acquired; the image data are input into a first recognition model and a second recognition model respectively, and inference is performed in parallel to output corresponding material identification results and product identification results; based on the material identification results, product identification results, and weight data, it is determined whether the recyclable item is a qualified recyclable; if it is determined to be a qualified recyclable, a first control command is output, or if it is determined to be an unqualified recyclable, a second control command is output. This solution addresses the problem in related technologies where recyclable material identification schemes cannot meet the requirement of acquiring directly recyclable aluminum cans and determining their product attributes, thus affecting the recycling effect.

The application provides a long-tail distribution object identification scheme fusing an unsupervised domain adaptation technology, comprising an object identification main system, wherein the object identification main system comprises a domain adaptation system, an object library and a model center; the domain adaptation system comprises a prototype level contrast learning and a consistency loss based on a smoothing label; the object library comprises an object image acquisition device and a database; and the model center comprises an object pre-training model and a long-tail identification model. The long-tail distribution object identification scheme can make the photographing object identification module in a shopping / science popularization application program more accurately identify objects in a natural life scene.

The present disclosure provides a method and a system of automatically identifying goods storage and retrieval. A method of dividing key monitoring areas corresponding to entrances and exits and intelligent shelves, activating user positioning and automatically weighing goods can realize intelligent and simple goods storage and retrieval for users without binding Radio Frequency Identification (RFID) tags for each piece of goods, thus greatly reducing consumables and labor cost. At the same time, the method can also avoid the problems of high computing cost and easy misidentification of a machine vision identification scheme. In addition, the storage and retrieval prediction algorithm can quickly determine and broadcast information of the retrieved / stored goods, thus reducing the waiting time of users and improving the user experience.

The application discloses a flow scheduling method and device and a medium, and relates to the technical field of communication. The method applied to a user-side device comprises the following steps: in the case that a first data packet of a data flow is listened to, source identification information of the first data packet is identified, the source identification information comprises user identification information and application identification information; flow identification information is generated, the flow identification information comprises the source identification information; the first data packet and the flow identification information are sent to a network-side device, so as to instruct the network-side device to, in the case that a second data packet of the data flow is received and five-tuple information of the second data packet is the same as five-tuple information of the first data packet, perform flow scheduling on the second data packet according to the source identification information. The application can improve the protocol applicability and compatibility of the flow identification scheme.

The invention relates to blind identification of Turbo codes in a non-cooperative communication environment, and provides a random interleaving mode identification method based on interleaving Turbo codes. The method comprises the following steps: firstly, realizing interleaving mode identification of an outer interleaver and an inner interleaver on the basis of a Turbo code coding system added with the outer interleaver; secondly, an LD-based identification scheme is proposed for the first time in the scene, and a KL divergence-based scheme is further proposed to identify the interleaver in the system, so that compared with the LD-based scheme, the influence of variance is considered, and the performance is further improved; moreover, aiming at the condition that only one RSC code is successfully recovered, a scheme of secondary identification after forward updating is adopted, so that the identification accuracy is further improved. The random interleaving mode identification of the interleaved Turbo code under the low signal-to-noise ratio is realized.

This invention discloses a method and system for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle. It relates to the field of carbon accounting and includes: classifying solid waste according to its compositional characteristics and establishing an identification system; collecting dynamic parameters at each stage through monitoring equipment to form a basic database; constructing a phased quantitative model based on the data; correcting the accounting results by combining a scenario-carbon emission correction matrix; using an LSTM neural network to predict carbon emission curves for the 5-10 year operation period; establishing a carbon balance equation for recycled aggregates in the waste stage; and finally analyzing parameter uncertainties through Monte Carlo simulation, dynamically updating the database, and outputting the carbon emission placement confidence interval. The advantages of this invention are: by refining the entire life cycle stages and combining a phased quantitative model, scenario correction, LSTM prediction, recycled carbon balance calculation, and Monte Carlo uncertainty analysis, it accurately, dynamically, and comprehensively completes the carbon accounting of coal-based solid waste road base courses, providing a basis for long-term planning.

The invention provides a premixed concrete quality detection method and device, a medium and a product, and relates to the technical field of image processing, and the method enables a target image detection model to capture and enhance time sequence change information related to concrete internal flow from an input image sequence through a space-time dynamic sensing module and a space-time attention module. Characteristics containing dynamic information are purified on different scales through a channel-space double attention module, and key apparent characteristics under each scale are focused. And finally, the adaptive feature pyramid network dynamically adjusts the fusion strategy according to the image content, and performs intelligent weighted fusion on the multi-scale dynamic features. The mode of combining time sequence dynamic analysis and self-adaptive multi-scale space analysis enables the target image detection model to form more comprehensive identification on the state of the concrete, and the technical problem that a concrete state identification scheme is poor in identification effect under complex and dynamic actual working conditions in a transportation tank car is solved.

The embodiment of the specification discloses a service recommendation method and device, a storage medium and an electronic equipment, wherein the method comprises the following steps: converting a user original behavior log into a semantic structure unit containing service item attributes and action features by using a preset interaction data collection burying point identification system; constructing an interleaved behavior sequence representing a user intention evolution track by alternately coding the extracted attributes and action features according to time; injecting candidate service item features into the sequence to form a target perception transduction sequence; reasoning a target perception intention feature; and finally, integrating the feature as an enhanced input into a preset recommendation model to output a prediction score. The embodiment of the specification effectively improves the intention capturing ability and prediction accuracy of the recommendation system by interleaved sequence modeling and target perception reasoning.

The embodiment of the invention discloses a service recommendation method and device, a storage medium and electronic equipment, and the method comprises the steps: employing a preset interaction data collection burying point identification system, converting an original behavior log of a user into a semantic structure unit containing service item attributes and action features, and storing the semantic structure unit in a database; the extracted attribute and action features are coded alternately according to time to construct an interlaced behavior sequence representing a user intention evolution trajectory, candidate service item features are injected into the sequence to form a target perception transduction sequence, target perception intention features are obtained through reasoning, and finally the features serve as enhanced input and are fused into a preset recommendation model, so that the target perception intention evolution trajectory is obtained. And outputting a prediction score. By adopting the implementation of the invention, through staggered sequence modeling and target perception reasoning, the intention capture capability and prediction accuracy of the recommendation system are effectively improved.

The invention provides a test board card management method, device, system and equipment, and the method comprises the steps: determining a corresponding relation between a server and a test board card through a system bus scanning operation, and carrying out the scanning operation based on the corresponding relation between a slot position and a test channel; the method comprises the following steps: controlling a test channel corresponding to each slot to perform level switching and detecting an output level change condition of each test board card to determine a corresponding relationship between the slot and the test board card, thereby generating a target corresponding relationship table capable of representing the corresponding relationship among a server, the slot and the test board card; compared with a mode of manually checking and recording the mapping relation in the prior art, an automatic identification scheme for the mapping relation in combination with software operation and hardware electric signals is provided, automation of mapping relation confirmation can be achieved, the labor cost is reduced, the accuracy of configuration records is improved, an unsupervised batch configuration mode can be supported, and the user experience is improved. The method is suitable for a large-scale production environment.