135results about How to "Quick update" patented technology

The embodiment of the invention relates to the technical field of large language modules, and discloses a multi-task processing method of a large language model fused with LoRA, which comprises the following steps: receiving task processing information input by a user, and performing feature extraction on the task processing information to determine each scene task feature; performing data matching with a pre-LoRA module parameter set based on the scene task features to obtain a LoRA module parameter set associated with the scene task parameters, and performing parameter updating operation on the large language model according to the LoRA module parameter set; and inputting the scene task features into the large language model subjected to the parameter updating operation for identification to obtain corresponding identification result information, and outputting the corresponding identification result information. According to the scheme of the invention, by fusing the LoRA module and the routing mechanism, the adaptive selection and optimization capability for specific tasks is realized, and the performance of the model in processing diversified tasks is improved.

The invention relates to multi-mode signal processing, and provides a signal processing method of a computer system. The method comprises the following steps: acquiring different modal signal data by using a plurality of sensors and performing timestamp synchronization; a depth time sequence alignment network and a multi-layer dynamic time warping algorithm are utilized to ensure signal time consistency; multi-modal signal features are extracted and fused through a multi-stream depth feature extraction network and a cross-modal attention mechanism; the system adopts a dynamic convolution kernel generation and meta-learning technology to carry out adaptive signal processing; noise suppression and signal optimization are carried out by using a multi-mode noise reduction auto-encoder and a generative adversarial network, and parameters are optimized and processed through closed-loop feedback control and reinforcement learning strategies. According to the method, synchronization, fusion, adaptive processing, noise suppression and real-time feedback control of multi-mode signals are realized, and the method has the advantages of high precision, high robustness and real-time performance.

The invention relates to the technical field of natural language processing, and discloses a natural language processing system based on Transform, comprising: a neural architecture search assembly for automatically searching and optimizing a Transform model architecture for natural language processing; the multi-task learning component is used for simultaneously training a plurality of natural language processing tasks on the basis of the shared model layer; the meta-learning component is used for optimizing parameter initialization of the model, so that the model can quickly adapt to a new task through a small amount of data and training steps; and the dynamic data enhancement component is used for generating diversified training data and enhancing the generalization ability and robustness of the model. According to the method, knowledge sharing and resource efficiency are enhanced through the multi-task learning and meta-learning component, dependence on a large amount of data is reduced, the generalization ability and rapid adaptability of the model are improved, the adaptability to a new scene is enhanced through the dynamic data enhancement component, and the robustness of the model is improved.

The invention discloses an ODPS-based customer service data aggregation method and server, and relates to the field of electric digital data processing, and the method comprises the steps: carrying out the field classification of collected source data, and obtaining a plurality of source data sets; loading the source data sets to an ODPS platform, and performing first aggregation processing on the source data in the plurality of source data sets by using the ODPS platform to obtain a plurality of field wide tables; pulling the field wide table to a DRDS platform, and performing distributed data processing on the field wide table by using the DRDS platform to obtain a plurality of pieces of distributed data; determining a target data feature corresponding to a target report in a report request instruction according to the report request instruction sent by the user terminal; and extracting multiple pieces of target data corresponding to the target data features in the multiple pieces of distributed data, and performing second aggregation processing on the target data to generate a target report. By implementing the method, the data processing efficiency can be improved when the multi-azimuth data display of the customer service staff is carried out.

The invention discloses a cloud platform rapid deployment system and method. The system comprises a Docker, a container arrangement tool, a database service, a cloud platform application, a monitoring and management tool, a network infrastructure and a security component. The Docker is responsible for managing the complete life cycle of the container and providing a virtualization environment; the container arrangement tool is used for managing and arranging a Docker container; the database service is used for storing information in the cloud platform and carrying out communication through the network service of the Docker; the cloud platform application is a set of service logic and functions deployed in a Docker container; the monitoring and management tool is used for monitoring a running state and managing log information; the network infrastructure is used for ensuring network interconnection between the servers and supporting network communication of the Docker container; and the security component is used for protecting the cloud platform from external attacks and internal threats. According to the method, the deployment process can be simplified, the deployment time is shortened, manual intervention is reduced, the deployment efficiency is improved, resources are optimized, and the operation cost is reduced.

The invention provides a position prediction method and device based on an ultra wide band (UWB) digital key, and a position prediction system applied to the UWB digital key, the position prediction system comprises a UWB tag and a plurality of UWB anchor points arranged on a vehicle, and the position prediction method comprises the following steps: determining whether UWB positioning information of the UWB tag is invalid; when the UWB positioning information is invalid, acquiring a velocity vector and a historical track set of the UWB tag and orientation information of the vehicle; and inputting the velocity vector, the historical trajectory set and the orientation information into a position prediction calculation model to calculate a predicted position of the UWB tag. The technical problem that when an existing automobile digital key is blocked by an obstacle and the digital key cannot be connected to an automobile UWB anchor point, the function of the digital key is in a failure or invalid state, and the user experience is affected is solved.

The invention discloses an intelligent recommendation method and system for an industrial process monitoring model, and the method comprises the steps: firstly, solving a problem that industrial time series data lacks visual features; extracting data set features of each case from four aspects of general statistical features, distribution shape features, task related features and exogenous knowledge embedding by adopting a data set feature engineering method; secondly, designing a mapping relation learning strategy based on width learning, mapping the data set feature of each case with the corresponding model score vector, and thus endowing the width learning network with the capability of predicting the model performance; and finally, providing a triple increment strategy for driving multiple increment demands of an industrial scene, and designing an updating strategy of a wide learning network for monitoring case increment, candidate model increment and recommendation node increment respectively, thereby continuously expanding and enhancing the recommendation capability of the method. According to the method, the detection model with the optimal performance can be recommended for the industrial process.

The invention relates to the technical field of power system analysis, in particular to a power system planning method based on multivariable dynamic analysis, and the method comprises the steps: obtaining the related data of a power system, and constructing a DC optimal power flow model with the minimum power generation cost; converting the DC optimal power flow model based on multi-parameter planning; considering the influence of the dynamic change parameters on the power system and constructing a critical region; and solving a planning result according to the critical region and the mapping relation between the dynamic change parameters and the power system planning result. The objective of the invention is to solve the limitation caused by the change of technical and economic parameters during the planning and deductive analysis of a power system, achieve the deductive analysis of continuously changing technical and economic parameters, and improve the reliability of a planning result.

The invention relates to the technical field of intelligent robots, in particular to an intelligent human-shaped accompanying robot and a voice interaction system thereof, and the system comprises a voice input collection and preprocessing module which collects a composite voice signal and a voice triggering instruction in a surrounding environment in real time through a distributed microphone array; the multi-sound-source voice separation module is used for combining and independently processing the potential sound source signals and separating the potential sound source signals into a plurality of independent sound source signals; the target user sound source identification and extraction module is used for predicting the change of a target sound source in continuous time frames by utilizing a dynamic Bayesian network and extracting a target user sound source signal; the semantic analysis and output module is used for generating a natural and smooth voice response through a voice synthesis technology; according to the method, the problem that the target user sound source is difficult to extract in a noisy environment, multi-sound-source overlapping and voice switching scenes is solved, and the robustness and the adaptive capacity of the system in a complex voice environment are greatly improved.

The invention discloses a knowledge graph ontology constrained vegetation remote sensing classification method. The method comprises the steps of data acquisition and data preprocessing, deep learning model selection and fusion of a knowledge graph ontology and a deep learning model. According to the method, the knowledge graph ontology and the FCN8s-ResNet 50 model are fused in the two stages of model training and classification post-processing. The result shows that 1) the classification accuracy of the deep learning model can be remarkably improved by fusing the mountain vegetation knowledge graph ontology in the two stages of deep learning model training and classification post-processing: the accuracy of the deep learning model constrained by the knowledge graph ontology is improved by 10%, and the overall classification accuracy of vegetation types reaches 89.2%; and 2) the deep learning model constrained by the mountain vegetation mapping knowledge domain ontology not only obtains excellent results in fine classification of vegetation types, but also improves the universality and interpretability of the model, so that powerful support is provided for wide-range remote sensing vegetation classification and application of the deep learning model.

The invention discloses an enterprise stock equity atlas construction method based on multi-dimensional data verification and a storage medium, and relates to the technical field of knowledge atlas.The enterprise stock equity atlas construction method comprises the steps that a text is extracted from obtained document data, layout analysis and table structure reconstruction are conducted on the extracted text, and the text obtained after format conversion is input into a large language model; extracting the enterprise entity information and stock equity relationship in the text; constructing a preliminary equity atlas; detecting a loop in the preliminary equity atlas, performing benefit community identification on the equity atlas after contradiction resolution, and outputting a healthy equity atlas; enterprise entity information is extracted in real time, credibility verification is carried out, and the information is stored in a database. A set of complete automation technology chain is formed from data acquisition and OCR table reconstruction to LLM extraction, map construction, contradiction resolution, benefit community recognition and health map comparison, and the manual operation cost is greatly reduced.

The present application discloses a single-frequency GNSS unmanned aerial vehicle positioning method, device, medium and product, which relates to the field of unmanned aerial vehicle positioning. At any current moment: the distance and ionospheric delay of each satellite are determined based on the position coordinates of each satellite and the position coordinates of the unmanned aerial vehicle at the previous moment; the tropospheric delay at the current moment is determined based on environmental parameters; three satellites are determined as first target satellites, and two satellites are determined as second target satellites; pseudo-range observation values ​​are calculated based on the distance, corrected clock error data, ionospheric delay and tropospheric delay of each first target satellite; phase observation values ​​are calculated based on the distance, corrected clock error data, ionospheric delay, carrier phase integer ambiguity and tropospheric delay of each second target satellite; the position coordinates of the unmanned aerial vehicle are determined based on the pseudo-range observation values ​​and position coordinates of each first target satellite, and the phase observation values ​​and position coordinates of each second target satellite. The present application improves the accuracy of unmanned aerial vehicle positioning and reduces costs.

The invention discloses a method for providing container service, a container system and a related device, and belongs to the technical field of containers. The method comprises the steps that a container system provides a basic mirror image and a public mirror image for a user, the basic mirror image is used for creating a container, the public mirror image is mounted, the public mirror image does not need to be downloaded, and occupied storage space of a node where the container is located is reduced. And moreover, the public mirror image independently exists, so that the public mirror image is convenient to manage and quickly update and upgrade. The public mirror image is mounted, so that the container created for the user can be quickly updated by quickly updating the public mirror image. And a public mirror image does not need to be downloaded, and container creation is more efficient through a mounting mode.

The invention relates to a distributed real-time library implementation method of a power monitoring system and electronic equipment, the real-time library is based on a multi-master multi-slave architecture and comprises a plurality of master nodes and a plurality of slave nodes, and the master nodes and the slave nodes belong to the real-time library and map real power equipment; according to the hierarchical data structure defined by the real-time library, equipment in the power monitoring system is added into real-time library model information, the real-time library model information is updated, initialization of the real-time library is achieved, and measured equipment real-time data is synchronized to the real-time library. Compared with the prior art, the method has the advantages of realizing rapid data synchronization of the real-time library under the condition of high concurrent access, reducing data synchronization delay when the main node fails, accelerating data updating efficiency and the like.

The invention is suitable for the technical field of data query, and provides a power marketing data query method and system based on a JointBERT model of a knowledge graph, and the method comprises the steps: obtaining field background knowledge related to the power industry, and constructing the knowledge graph; a JointBERT model is constructed and trained, and the model comprises an input layer used for processing input text data; the coding layer is used for carrying out fusion coding on rich general semantic knowledge in a corpus and domain knowledge in the knowledge graph; the decoding layer is used for generating field and intention corresponding to the input text and slot labels corresponding to words; compressing and optimizing the model by adopting an early leaving mechanism and a knowledge distillation method to obtain a target model; and the power demand problem of the user, which is received in real time, is used as the input parameter and is input into the target model for data query and analysis, so that the enterprise information query efficiency of related workers is improved.

The invention discloses an integrated energy system planning and scheduling method and device considering demand side response. The method comprises the following steps: establishing load models of price type demand response, excitation type demand response and alternative type demand response; establishing an initial planning scheduling model by taking the lowest annual operation cost and the lowest demand side response cost of the full-life-cycle integrated energy system as a first optimization target; solving the initial planning scheduling model in combination with operation constraint conditions and user satisfaction constraints to obtain load transfer quantity and reduction quantity after price-type demand response and excitation-type demand response, and a real-time pricing mechanism; carrying out alternative demand response based on a real-time pricing mechanism to obtain an optimized price curve and load substitution quantities of various energy sources; and solving the planning scheduling optimization model taking the lowest annual operation cost of the full-life-cycle integrated energy system as a second optimization target to obtain an optimal capacity configuration result and an operation scheduling plan of each energy device in the integrated energy system.

The embodiment of the invention provides a satellite ephemeris updating method and device and spaceborne equipment, and relates to the technical field of aerospace, and the method comprises the steps: obtaining the current position and current space environment parameters of a satellite; searching a grid adjacent to the current position from the plurality of grids, wherein the grid is associated with the position, the space environment parameter interval and the atmospheric density; predicting the atmospheric density of the orbit where the satellite is located by using the position, the space environment parameter interval and the atmospheric density associated with the adjacent grids, the current position and the current space environment parameters; and updating the ephemeris of the satellite by using the predicted atmospheric density. By applying the technical scheme provided by the embodiment of the invention, the prediction precision of the atmospheric density can be improved and the satellite ephemeris prediction precision can be improved under the condition that the computing power and storage of satellite-borne equipment are limited.

The invention provides a production management system based on a low-code platform, and relates to the technical field of data processing, and the system comprises a data collection module which is used for collecting production resource state data, order task data and process parameter data; the task decomposition module is used for decomposing the production task to generate task unit data; the configuration generation module is used for generating initial task configuration data according to the task unit data; the task scheduling module is used for receiving the initial task configuration data and generating a task scheduling scheme; the execution control module is used for distributing a production instruction to production equipment according to the task scheduling scheme and controlling the execution process of the production task; the dynamic updating module is used for dynamically adjusting task configuration and feeding back to the task scheduling module when a production task adjusting request is received in the production task execution process; the resource allocation scheme of the production management system can be automatically updated based on the data feedback of the production resources.

The invention provides a router fault automatic repair method, system and device and a medium, and the scheme comprises the steps: responding to the occurrence of a fault, and obtaining log data corresponding to the current fault in real time; obtaining a problem classification result and a problem mark of the current fault by using a pre-trained fault classification model based on deep learning based on the obtained log data; based on the problem classification result and the problem mark, accessing a pre-constructed problem solving large model according to a preset public network service path to obtain a solution of the current fault; and executing the currently obtained solution to realize automatic repair of the router fault. According to the scheme, by designing the problem solving large model, the router can automatically repair most of self faults, and by periodically updating the problem solving large model, the robustness of the problem solving large model can be effectively improved.

The embodiment of the invention discloses a graphics processor, an operation method of the graphics processor and electronic equipment, and relates to the field of computers. According to the method, the DMA module is used for directly transmitting data between the SRAM and the configuration register / memory of the ISP, frequent intervention of a CPU is not needed, the configuration process of the ISP is simplified, and the configuration efficiency is improved. The SRAM maps a section of area with the same size as the ISP configuration register / memory, and quick updating and reading of ISP configuration and information are realized through direct reading and writing of the DMA. Compared with a mode of directly using a plurality of registers and memories, the mapping mode has the advantages that the occupation of hardware resources is reduced, so that the area of the whole SRAM is smaller than the total area of a mode of independently using the registers and the memories.

The invention provides an intelligent agent-based perioperative health education method for an elderly fracture patient, and the method comprises the steps: inputting multi-dimensional target dialogue content in a target interface of an intelligent agent in response to the elderly fracture patient based on a preset input form, analyzing the target dialogue content based on a preset natural language processing model to obtain a corresponding analysis result, generating personalized health education content corresponding to the target dialogue content according to the analysis result in combination with a preset target health education knowledge base, and determining an interaction mode of the elderly fracture patient for the identity category of the elderly fracture patient. And feeding back personalized health education content to the elderly fracture patient in real time based on an interaction mode, acquiring physiological data and feedback data of the elderly fracture patient, adjusting the personalized health education content based on the physiological data and the feedback data, and feeding back the adjusted personalized health education content to the elderly fracture patient in real time.

The invention discloses a cloud map updating method and device for a strip mine area in the technical field of point cloud maps. The method comprises the steps that the strip mine area is divided into a terrain change frequent area and a non-map change frequent area; carrying out map data acquisition on the frequent terrain change region, generating instant update acquisition data, and uploading the data to a cloud in real time; the cloud map is updated at the cloud according to the real-time updating collection data; carrying out map data collection on the non-topographic-change frequent region, generating delayed update collection data, and uploading the delayed update collection data to a cloud end in a delayed manner; the cloud map is updated at the cloud according to the delayed updating collection data; different acquisition modes correspond to different map updating modes, so that the map updating frequency of an infrequently changed area is reduced, the accuracy and the real-time performance of the map are ensured, the server pressure is reduced, and the user experience is improved.

The utility model discloses a lighter similar to a diesel engine, which comprises a protective shell, decorating parts are arranged on the front surface, the rear surface and the right surface of the protective shell, a bearing is fixedly connected to the surface of the decorating part on the rear surface of the protective shell, a gyro wheel is rotatably connected to the surface of the bearing, and a fire outlet pipe is fixedly connected to the left surface of the protective shell. A fire behavior limiting cover and a shell sealing cover are arranged at the upper end of the fire outlet pipe and the upper end of the protective shell correspondingly, a button is arranged on the surface of the shell sealing cover, a fuel storage box and an electric igniter are arranged in the protective shell, a valve is arranged on the right surface of the fuel storage box, and a right-angle guide pipe is arranged in the valve. The output end of the electric igniter and the surface of the right-angle guide pipe are provided with a power line and a rubber sleeve correspondingly, and the upper surface of the valve is rotationally connected with a winding rod. By means of the structure, the compactness of internal components of the lighter is improved, and a user can conveniently replace a power line of a fire outlet area.

The application discloses a self-luminous replaceable plate type photocatalytic experimental device and method with sound-light coupling reinforcement, relates to the technical field of photocatalysis, and comprises a self-luminous catalytic substrate, a reaction tank integrated with an ultrasonic transducer, an online monitoring unit and an external pumping circulation system; the self-luminous catalytic substrate is made of a semi-transparent optical fiber material and has an optical fiber guiding structure inside, can concentrate and guide external light to the inside of the substrate and uniformly transmit the light out of the surface, and serves as not only a support carrier of a catalyst but also an efficient in-situ surface light source; the ultrasonic cavitation reinforces the liquid-solid interface mass transfer and peels off a toxic layer in real time; the online monitoring unit acquires reaction efficiency data in real time and accurately guides physical replacement of the substrate according to a kinetic decay curve; the application realizes full-process zero-loss transmission of light energy and orders-of-magnitude improvement of mass transfer efficiency, solves the problem of low photocatalytic efficiency caused by limited light transmission and mass transfer bottleneck, and is suitable for evaluation of high-performance catalytic materials and research on deep treatment of high-viscosity / high-turbidity systems.

The invention relates to the technical field of vehicle control, and discloses a base station selection method and device, equipment and a storage medium, and the method comprises the steps: carrying out the speed measurement of a network base station connected with a vehicle based on a speed measurement data packet issued by a management platform, and generating a base station speed measurement data packet; uploading the base station speed measurement data packet to a management platform, so that the management platform updates the network state map according to the base station speed measurement data packet; receiving an updated network state map fed back by the management platform, and selecting a target base station according to the updated network state map; and controlling the vehicle to switch and connect to the target base station. The vehicle can be used for replacing the speed measurement equipment to perform network speed measurement on the network base station, so that the network state map can be quickly updated, the timeliness of the network state map is improved, the base station selection of the vehicle based on the updated network state map is more in line with the actual scene, and the good network state of the vehicle is ensured.

The present application relates to the technical field of hydraulic engineering modeling, and discloses a method, device, equipment and storage medium for generating a hyperbolic arch dam BIM model. The method comprises: based on the elevation parameter information and parameter equation of the target arch dam model, creating the basic shape of the target arch dam model through a Dynamo programming plug-in; determining the target components required for the target arch dam model according to the simulation analysis tasks to be performed as needed; loading the basic shape and the target components in the Dynamo programming plug-in; when the type of the target component is one or more, running a fusion code, combining the basic shape and the target component, and generating the target arch dam model. The present method can quickly generate high-precision arch dam models of different structures, improve the modeling efficiency and modeling accuracy of the arch dam model, and save time and labor costs.

The embodiment of the invention provides a table data processing method and device, equipment, a medium and a product. The method comprises the steps of displaying at least part of records of a target table in a target interface of a client in response to a data access request which is input by the client and aims at the target table; the record refers to a row of data in the target table; receiving a sub-record editing operation for the target table through the target interface, determining hierarchical description data corresponding to the edited sub-record according to the sub-record editing operation, and updating display data of the target table in the target interface according to the hierarchical description data; the hierarchical description data is used for describing child records in the target table and father-child relation data corresponding to father records containing the child records. According to the technical scheme provided by the embodiment of the invention, the records of the table can be loaded as required, the changed records can be quickly and effectively updated in a targeted manner, and the processing efficiency and the use experience of the table are improved.

The application discloses a routing association relationship updating method, a routing communication method and related devices. The method determines the shortest path of each node in the autonomous network when updating the path, and determines the problem node according to the comparison result between the shortest path of each node and the historical path. Finally, according to the routing prefix in the problem node association prefix set, the target next hop block associated with the routing prefix is selected from the pre-stored next hop block. Then, the next hop block associated with the first prefix is updated according to the target next hop block and the shortest path of the problem node. The next hop block of the application is an index representing the forwarding address of the message. The update of the next hop block associated with the routing prefix is the update of the forwarding address of the message. The flow does not need to re-decide the next hop address of each routing prefix based on the shortest path of all nodes when updating the path, which saves the computing power and improves the network performance.

The present invention discloses a remote grinding database management system with a user-base-process-knowledge progressive structure and an efficient and low-consumption intelligent grinding method, which belongs to the field of intelligent manufacturing. Aiming at the problem of automatic search for flexible optimal processing solutions that urgently needs to be solved in data-driven intelligent grinding technology, the present invention provides a grinding process knowledge acquisition and intelligent control processing method based on the power signal of the grinding process monitoring. At the same time, the present invention integrates the efficient management of users, bases, processes and knowledge data of the entire grinding industry chain, and can realize remote control of the automatic production monitoring system to the workshop and 1 to N network resource sharing, enhance the intelligent and flexible production capacity of grinding processing, and improve the core competitiveness of grinding processing products.

The invention relates to the technical field of campus information acquisition, in particular to a digital information acquisition device based on campus clothes images, which comprises a school uniform body, a clamping groove, an information display screen, a detection door and a positioning device, and is characterized in that the clamping groove is formed in the school uniform body, and the information display screen is clamped in the clamping groove; the detection door is installed at a school gate security check position and is used for detecting whether information between a school entering person and the information display screen is correct or not so as to prevent a non-school person from entering a school; the plurality of positioning devices are installed in the campus, and the positioning devices are used for determining the position of each student according to the information of the information display screen; by arranging the information display screen, the information of students is displayed on the information display screen in real time, so that teachers can quickly obtain various information of the students in a school, and the information display screen is charged through the solar panel and the conducting strip, so that the students can work for a long time.