533results about How to "Lower the threshold" patented technology

The invention discloses a distributed heterogeneous computing power reasoning task dynamic scheduling method and system. The system comprises a computing power architecture adaptation module, a task scheduling module, a model updating module, a capability service module, a dynamic scheduling module, a resource evaluation module and a state sensing module. The computing power architecture adaptation module is used for identifying and matching a local heterogeneous computing power architecture and a remote heterogeneous computing power architecture; the task scheduling module is used for selecting a proper computing power terminal from a distributed heterogeneous computing power environment; the model updating module is used for updating the AI model in real time; the capability service module is used for providing a unified AI model service calling interface; the dynamic scheduling module is used for dynamically allocating computing power resources; the resource evaluation module is used for monitoring and evaluating the state of the computing power resource; the state sensing module is used for monitoring the health state of the computing power node in real time and responding to a fault in time; the scheduling efficiency of computing power resource scheduling is improved, and stable operation of the reasoning task is guaranteed.

The invention relates to the field of financial risk management, in particular to a risk identification method, device and equipment. The risk identification method comprises the steps that user appeal information is acquired, and the user appeal information comprises a business scene and a risk identification demand under the business scene; querying first service data matched with the risk identification demand in the service scene; according to the user appeal information and the first business data, generating a risk identification rule by using a large language model; querying second service data matched with the risk identification rule in the service scene; performing reinforcement learning on the risk identification rule by using the second business data to obtain an optimized risk identification rule; optimizing the large language model according to the business scene and the optimized risk identification rule; and performing risk identification by using the optimized risk identification rule. According to the embodiment of the invention, the limitation of poor adaptability and adjustability of manual design rules is overcome, and the efficiency and quality of risk identification rule generation are higher.

The invention discloses a power load accurate prediction system based on artificial intelligence, and the system comprises a data collection module which is used for collecting multi-source data through a preset path; the data preprocessing module is used for preprocessing the multi-source data according to the data type to obtain processed data; the data fusion module is used for carrying out fusion processing on the processed data by utilizing data fusion to obtain input data in a unified format; the model training module is used for training a constructed deep learning model based on the input data to obtain a prediction model; and the prediction module is used for reasoning current data by adopting the prediction model to obtain a power load prediction result. And the effectiveness and accuracy of power load prediction are improved.

The invention discloses a 2D government affair digital human real-time interaction method and system based on an AI painting technology, and the method comprises the steps: making a digital human image based on a modeling mode combining AI painting and a Live2D technology; acquiring an audio input by a user in real time, and identifying the audio input by the user to obtain an audio identification text and a reply text; according to the reply text, synthesizing and outputting a result audio based on a half-stream speech synthesis method of dependency syntactic analysis; and performing mouth shape synchronous real-time driving and audio broadcasting of the digital human image on the manufactured digital human image based on the result audio. According to the invention, the low-cost construction and deployment of the digital human are realized based on the combination of AI drawing and Live2D technology; when system deployment is carried out on the digital people, a semi-streaming speech synthesis scheme based on dependency syntactic analysis is adopted, and the interaction real-time performance and fluency of government affair digital people belonging to service digital people are improved.

The invention discloses a tensor program generation method and system based on a large language model, and the method comprises the following steps: (1) generating a random ONNX computational graph through an AI compiler test tool nnsmith; (2) importing an ONNX calculation graph generated by nnsmith into a TVM, and converting the ONNX calculation graph into a RelayIR intermediate representation; (3) based on a model intermediate representation RelayIR in the TVM, adjusting and optimizing the model through an automatic adjusting and optimizing device Ansor in the TVM, generating an optimal tensor program, and outputting a JSON file measured during adjusting and optimizing; (4) continuously carrying out model tuning and tensor program measurement, generating a large number of measurement records during tuning, and forming a tensor program data set which comprises features of a tensor program and a corresponding optimization pass list; and (5) training a large language model according to the generated tensor program data set to realize understanding and generation of the tensor program. Through combination of an automatic tool and a large language model, efficient tensor program generation and optimization can be realized.

The invention discloses a visual database system supporting real-time collaboration and computer equipment, and the system comprises a space management module which is used for managing a database space and dividing a database into a plurality of spaces, and the use permissions of different users are restrained to be in the corresponding spaces; the user system module is used for managing user attributes and user permissions of the users in each space based on configuration operation of an administrator on the visual page; the database management module is used for performing data storage in each space based on preset database configuration data, and querying the data stored in the corresponding space based on a data query instruction input by a user on the visual page; the operation auditing module is used for querying and auditing the data modification information of all the spaces based on the user permissions of different users and recording the data modification information; and the API management module is used for accessing a RESTful interface to query and modify the data in the database management module.

The invention belongs to the field of software testing, and relates to an ROS 2 software package defect detection method based on a large language model and coverage rate guidance. According to the method, function information is extracted from a to-be-tested program, and a regular expression is used for analyzing function definition, parameter characteristics and other information; the method comprises the following steps: presetting verification standards related to a to-be-tested program, and extracting physical constraints and document constraints from the verification standards; integrating the extracted function information, physical constraint and document constraint into accurate information, filling a cue word template with the accurate information, inputting the cue word template into a large language model, and generating an initial test seed; a mutation strategy is introduced, and diversified mutation operation is performed on the generated seeds, so that test cases with higher diversity are generated; optimizing the test case by adopting a coverage rate guiding mechanism, and guiding to generate a new test case by monitoring the coverage rate in real time; through a defect verification mechanism, dominant defects such as program crash and the like can be detected, and hidden defects such as violation of physical constraints or violation of document constraints and the like can be identified.

The invention belongs to the technical field of protein structure prediction, and relates to a protein structure prediction method, system and device based on a light quantum computer, and the method comprises the steps: 1) obtaining a multi-sequence alignment MSA matrix of a target sequence based on the target sequence alignment of a to-be-predicted protein; 2) encoding the obtained multi-sequence alignment MSA matrix into a {0, 1} matrix; 3) converting the amino acid interaction of the protein into an undirected graph model of {0, 1} state nodes, and training by adopting a Boltzmann machine training mechanism and using a light quantum computer to obtain a weight coefficient of a connecting edge between the nodes in the undirected graph model; and 4) obtaining an interaction coefficient between different amino acids of the protein based on the weight coefficient. According to the method, the calculation efficiency and the solving result of protein structure prediction are improved, and the problems that complex scheduling problems are difficult to calculate and accurate solutions cannot be obtained in the prior art are solved.

The invention relates to the technical field of big data processing and analysis, in particular to a visual SQL generation method and system for FLINK real-time data processing, and the method comprises the following steps: data source management, which is used for managing a data source classification directory; the data type management is used for converting various database management data types; the visual SQL configuration is used for combining and adjusting a data source, a database and a data table in the data source management system and data type information in the data type management system in a visual mode, and finally generating an SQL capable of being executed in an Flink environment; the visual SQL generation method and system for FLINK real-time data processing have the advantages that the technical threshold is lowered, non-technical users can easily define data processing logic, complex Flink codes do not need to be written, and the threshold for using Flink is lowered; the development efficiency is improved, the data processing flow is simplified through a visual operation interface, the development period is shortened, and the development efficiency is improved.

The invention belongs to but is not limited to the technical field of PCBA defect detection, and discloses a PCBA AI defect detection item demand response method and system, which adopts multi-scale long-range fusion, context information, super-resolution and block detection to improve a target detection algorithm, and performs detection and identification of a tiny target. Identifying the fine granularity of the target by adopting a contextual information and depth feature matching method; 3D point cloud and X-ray perspective are adopted, a multi-sensor fusion detection frame is used, and invisible defect detection is carried out. According to the method, AI technologies such as deep learning are utilized, the detection precision of small defects on the PCBA board is effectively improved, and missing detection and false detection are reduced. The AI technology can automatically identify and learn defect features, and efficiency and accuracy are significantly improved. The AI model can optimize the self algorithm through continuous learning, so that the detection system can adapt to different PCBA boards and production environments, and the universality and flexibility of the detection system are enhanced.

The invention discloses an AI calculation training system which comprises an AI training platform, and the AI training platform comprises a data set management module which is used for creating a data set, labeling picture data or video data and screening data in the data set; the model training module is used for a user to select a model application type, a training data set, a training label and an adjusting and optimizing mode in a model training stage, and the system realizes model automatic training through an automatic deep learning technology; the model management module is used for managing various models and screening, deleting or checking data models; and the model verification module is used for helping a user to simulate an actual production environment through a model verification function and further evaluating a model effect. According to the method, a full-process graphical operation interface from data management to model release is provided, a user can complete model development work without any coding capability, the AI application development period is remarkably shortened, and the requirement for quick online application is met.

The embodiment of the invention provides a parameter configuration method and computing equipment. The parameter configuration method is used for performing parameter configuration on a to-be-configured hard disk in a server. The method comprises the following steps: receiving a parameter viewing request for a to-be-configured hard disk from terminal equipment, wherein the parameter viewing request is used for requesting to view parameters of a Mode Page of the to-be-configured hard disk; in response to the parameter viewing request, parameters of the Mode Page of the hard disk to be configured are sent to terminal equipment, and the terminal equipment is used for displaying the parameters of the Mode Page; under the condition that a user modifies the parameters of the Mode Page displayed by the terminal equipment, parameter modification information sent by the terminal equipment is received, and the parameter modification information comprises configuration parameter values obtained by modifying the parameters of the Mode Page; and indicating the to-be-configured hard disk to modify parameters of the Mode Page based on the configuration parameter value. According to the method, the parameter setting step can be simplified, the operation difficulty is reduced, and the working efficiency of parameter configuration is improved.

The invention relates to the technical field of battery management, in particular to a lithium battery remaining life prediction method and device, terminal equipment and a storage medium, and the method comprises the steps: determining the working parameters and environmental parameters of a target battery and the cell type of the target battery; determining a corresponding pre-constructed life prediction model according to the type of the battery cell; the life prediction model is constructed according to the working parameters and the environmental parameters; and determining the current cycle charging and discharging times of the target battery, and obtaining the residual life of the target battery according to the life prediction model, the cycle charging and discharging times, the environmental parameters and the working parameters. Therefore, more relevant factors are considered in the life prediction of the lithium battery, the prediction result is more accurate, and the prediction process is more convenient.

The invention discloses a generative model reasoning deployment method and device based on AI questions and answers, and the method comprises the steps: constructing the basic configuration of an AI questions and answers system, and carrying out the model deployment; designing a database and storing model metadata, a model example code, a GPU (Graphics Processing Unit) configuration guide and a Dockerfile script template; receiving a user reasoning service request, generating a deployment task ID and recording task information; performing model existence check, and if the model is missing, automatically downloading model metadata according to a stored Git warehouse address; extracting corresponding configuration data and example code details from the database, and dynamically filling a script template with the configuration data and example code details to generate a Dockerfile script; executing a mirror image starting script constructed in the container, running a mirror image and returning a service address; according to the invention, the efficiency and convenience of model service deployment are improved, the threshold of AI technology application is reduced, the resource configuration is optimized, and the user experience is enhanced.

The invention discloses a digital role audio and video generation method and a digital role live broadcast interaction method, and the generation method comprises the steps: generating explanation data of a to-be-explained object according to a digital role material generation task of a requester; performing data conversion processing on the explanation data to obtain audio data corresponding to the explanation data; performing role image digital processing on the obtained role image reference data to obtain a digital role image video; carrying out fusion processing on the audio data and the digital role image video to obtain a digital role driving video for explaining the object to be explained by using a digital role; returning the digital role driving video to the requester, so that the requester processes the digital role driving video before distribution to generate a digital role audio / video; therefore, computing resources are saved, the cost is reduced, and the coverage rate of digital role audio and video playing is improved.

The invention discloses a point cloud imaging and positioning method based on a chip-level millimeter wave radar, and the method comprises the steps: receiving echo information, carrying out the windowing of each frame of echo signal, and carrying out the fast time dimension and slow time dimension FFT, and obtaining a plurality of frames of RDM; combining multiple frames of RDM, and setting a coarse-grained label by using a low threshold value; respectively extracting an azimuth angle and a pitch angle from each frame of RDM, and converting a target point cloud set into a Cartesian coordinate system for clustering and denoising; performing fine-grained marking on the RDM according to the de-noised point cloud; taking the RDM of each frame and the RDM with the fine-grained label as input and output respectively, and training a deep learning network to obtain an RDM target detection network; the multi-frame RDM of a to-be-detected target is obtained in the moving process of the mobile radar, the multi-frame RDM is sent to the RDM target detection network to be processed, point cloud data are extracted, and registration is carried out to obtain the pose of each frame of the radar and the final point cloud image of the to-be-detected target. According to the invention, high-quality dense point cloud images can be efficiently generated.

The invention discloses a three-dimensional model generation method based on a text cue word, and the method comprises the steps: predicting a distance sign function SDF through a constructed neural network according to a dynamically optimized random sampling camera posture and an illumination vector, and carrying out the volume rendering, and obtaining an initial rendered image; through a pre-trained diffusion model, the rendering image quality and the signed distance function prediction accuracy are improved according to the encoded text cue word information; and through the trained neural network, predicting rendering colors and a signed distance function according to known observation point coordinates, obtaining three-dimensional model surface point cloud data in an interpolation mode, and finally generating a three-dimensional model format file. According to the method, the problems that a professional needs to apply professional software to construct the three-dimensional model, time is long and the like are effectively solved, the three-dimensional model constructed through the method can be directly printed through three-dimensional printing equipment, and finally a three-dimensional object entity associated with the text cue word can be obtained.

The invention relates to the technical field of lubricating oil analysis, in particular to a method and system for quantitative analysis of a lubricating oil additive. The method comprises the following steps: acquiring base oil data and lubricating oil additive data; performing mixed oil treatment based on the base oil data and the lubricating oil additive data to generate mixed lubricating oil additive data; performing mixed oil concentration detection on the mixed lubricating oil additive data to generate mixed oil concentration detection data; performing chromatographic column selection on the mixed lubricating oil additive data based on the mixed oil concentration detection data to generate chromatographic column selection data; and performing boiling point heating on the chromatographic column selection data to obtain mixed oil boiling point heating data. By optimizing data acquisition, chromatographic separation, component identification and quantitative analysis as well as result analysis and visualization, the accuracy of quantitative analysis of the lubricating oil additive is improved.

The invention discloses a method and a system for automatically extracting a complex road network topological relation under a multi-breakpoint condition, relates to the technical field of road network data processing, and solves the technical problem that a road network topological structure is inconvenient to extract due to the fact that a road section space connection relation cannot be effectively utilized under the multi-breakpoint condition in the prior art. The method comprises the steps that road sections of road network data are split, the split road sections are traversed, breakpoints are determined, target adjustment straight line numbers corresponding to the breakpoints are recorded, then the road network data are converted into an undirected graph according to the breakpoints and the corresponding target adjustment straight lines, the undirected graph is divided into connected sub-graphs, and the connected sub-graphs are connected with the road network data. A base straight line is selected from the connected subgraph, nodes of other straight lines are projected to the base straight line, coordinates of the nodes are adjusted according to projection positions of all projection nodes on the base straight line to obtain adjusted road network data, and a topological relation is regenerated according to the adjusted road network data; according to the method, the threshold of road network topological relation extraction is reduced, and the labor cost is greatly saved.

The embodiment of the invention provides a simulation scene generation method, device and equipment, and the method comprises the steps: obtaining scene description information of a target simulation scene, and the scene description information is used for describing the movement information and road condition information of traffic participants; obtaining a global map of the target area, and determining a target local map corresponding to the target simulation scene in the target area based on the scene description information and the global map; and generating a target simulation scene based on the target local map and the scene description information. According to the simulation scene generation method provided by the invention, the construction efficiency and controllability of the simulation scene can be improved.

The invention provides a near-infrared light direct curing additive manufacturing slurry and a preparation method thereof. The slurry comprises the following components in parts by weight: 2-15 parts of a first auxiliary agent, 10-85 parts of inorganic material powder, 0.01-5 parts of a dispersing agent and 0.01-5 parts of a photoinitiator. Wherein the first auxiliary agent serves as a curing monomer and a cross-linking agent in the slurry, the refractive index nD of the inorganic material powder is 1.5-2.8, and the slurry does not contain an up-conversion luminescent material. The slurry can be directly cured under the action of infrared light with the wavelength range of 850-1100 nm and the light source energy density of 3 W / cm < 2 >-15 W / cm < 2 >. The near-infrared light direct curing additive manufacturing technology has the advantages of being high in forming speed, high in single-layer thickness, high in precision, low in energy consumption, high in efficiency and the like, and a new solution is provided for large-scale industrial application of additive manufacturing.

The invention provides a universal method for realizing multi-condition and search based on LDAP, and belongs to the technical field of data retrieval of directory service and identity management. A user defines query conditions including attributes, values and logical relationships through a configuration file, a visual interface or an interface; the system parses the query condition model into a query condition model supporting nested combination of AND, OR and NOT by using a parser, and performs legality check and error prompt; and then a generator dynamically generates a filtering statement conforming to the LDAP grammar specification according to the model, and query is executed through an LDAP client. In order to improve the performance, common condition caching and high-frequency statement pre-compiling optimization are introduced, and an asynchronous query mode is supported. And the query result is returned in the form of structured data, and is matched with a display interface to support screening, sorting and exporting, so that the flexibility and maintainability of multi-condition query are improved.

The invention discloses an unmanned aircraft and pilot management and control method and management and control system, and the method specifically comprises the steps: collecting the product identification information of an unmanned aircraft, judging that the product identification information is illegal if the product identification information cannot be obtained, and further judging whether the real-name registration is needed or not if the product identification information can be obtained; if the real-name registration information is needed, acquiring real-name registration information, if the real-name registration information cannot be acquired, judging whether the real-name registration information is illegal, if the real-name registration information can be acquired or does not need to be real-name registration, acquiring flight dynamic data to judge whether the flight dynamic data is in a suitable flight airspace range, and if the flight dynamic data is legal; otherwise, further collecting flight authorization information, and if no authorization exists, determining that the flight is illegal; and if authorization is obtained, judging whether the reported flight plan data is met, if not, judging that the flight plan data is illegal, and if yes, judging that the flight plan data is legal. The management and control system comprises a memory, a processor and a computer program capable of running on the processor and executing the management and control method. The method is wide in applicability and low in application threshold, the accuracy of legal / illegal identification can be greatly improved, and the management and control safety is improved.

The embodiment of the invention provides a data processing method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining transaction process data corresponding to a target transaction, and obtaining a target index and index configuration data of the target index based on the transaction process data, and obtaining a target data source corresponding to the target transaction based on preset data source configuration data, obtaining an index calculation result of the target index from the target data source based on the index configuration data, and analyzing the index calculation result according to the index configuration data to obtain an index analysis result of the target transaction.

The invention discloses a quality infrastructure service efficiency evaluation method based on a large model, which is based on full-chain elements such as construction, management, service and propaganda and promotion of a quality infrastructure one-stop service platform, integrates application of a key technology of the one-stop service digital platform, deeply explores influence factors of online and offline service efficiency of the platform, and improves the quality infrastructure service efficiency evaluation efficiency. And a gradient grading platform evaluation system model is constructed. The objective of the invention is to provide a powerful tool for platform self-service efficiency evaluation and provincial platform gradient grading.

The invention relates to the technical field of devices used for replacing metal wires on the surface of any transparent hard substrate to transmit power for electrical equipment, in particular to a transparent power supply device with a pre-etching circuit and a preparation method of the transparent power supply device. The transparent power supply device comprises a low-sheet-resistance conductive film coiled material and an electrode connecting kit matched with the low-sheet-resistance conductive film coiled material. The low-square-resistance conductive film coiled material is connected with the electrode connecting kit, and the electrode connecting kit is connected with a power supply for taking electricity; the low-square-resistance conducting film coiled material comprises a conducting layer, a prefabricated circuit is arranged on the conducting layer and comprises a plurality of parallel conducting areas, an etching circuit is arranged between every two adjacent conducting areas, and the etching circuits are used for insulating every two adjacent conducting areas; comprising the following steps: preparing a prefabricated circuit of a low-sheet-resistance conductive film coiled material, namely processing and forming the prefabricated circuit on the low-sheet-resistance conductive film coiled material by adopting a roll-to-roll continuous etching process; the roll-to-roll continuous etching process comprises the following steps: carrying out circuit pre-preparation on a low-square-resistance conductive film coiled material by adopting roll-to-roll continuous etching equipment; the roll-to-roll continuous etching process comprises the following steps: roll-to-roll continuous etching is carried out along the material pulling direction of equipment; the roll-to-roll continuous etching process comprises the step of roll-to-roll continuous etching in the direction perpendicular to the material pulling direction of the equipment. Compared with the prior art, the transparent power supply device with the pre-etching circuit and the preparation method of the transparent power supply device can effectively overcome the defects of the existing technical scheme, improve the product yield and reduce the production difficulty and threshold, so that the technology can be widely accepted and applied by the market.

The embodiment of the invention provides a self-planning method for a mathematical geometry problem of an intelligent agent, which can be applied to the technical field of artificial intelligence. The self-planning method for the mathematical geometry problem of the intelligent agent comprises the following steps: generating a seed image according to scene demand information of the mathematical geometry problem; performing complex feature planning on the seed image based on a preset guide search rule until a target mathematical problem set is output; and performing problem solving path matching on at least one target mathematical problem in the target mathematical problem set through a preset language model to generate a problem solving planning path set. The embodiment of the invention further provides a self-planning method and device for the mathematical geometry problem of the intelligent agent, equipment, a storage medium and a program product.

The invention discloses a search intention recognition method based on an NL2SQL model, relates to the technical field of data retrieval, and solves the problems that in the prior art, SQL query writing in the data analysis process needs a large amount of manual participation, time is consumed, and the requirements for the technical level and experience of operators are high. According to the method, the original data is converted into the data items by constructing the business knowledge base and the business dictionary base, and the SQL statements are constructed based on the data items. The method has the advantages that natural language query can be understood and converted, so that non-technical users can easily perform database query. Natural language query is converted into SQL statements by training a pre-training language model, so that the threshold of using a database by a user is greatly reduced. According to the method, the retrieval accuracy is improved, the user experience is enhanced, and database query becomes more visual and convenient. The method brings innovation and progress to the field of information retrieval.

The invention provides a progressive cloud loading method and system based on strength analysis and intelligent process design software, and relates to the field of industrial software cloud loading, and the method comprises the steps: S1, carrying out the micro-service transformation of key modules and services in CAE and CAPP systems; s2, packaging the micro service to form a standardized container mirror image; s3, constructing an adaptation layer to ensure data security interaction between the on-cloud service and a legacy local system, and maintaining data consistency and continuity of an on-cloud and under-cloud system by using a data synchronization service; s4, implementing a progressive migration strategy, gradually moving the micro-service to a private cloud environment, and maintaining integrated collaboration with a non-migrated part by using an adaptation layer; s5, integrating a CI / CD process, and completing software testing, construction and deployment in an automatic mode; and S6, setting safety compliance measures. By means of the method, an enterprise can gradually achieve intellectualization and cloud collaboration of the design and manufacturing process with the lowest cost and risk.

The invention relates to the technical field of monitoring management systems, in particular to a power network communication equipment monitoring and early warning system, which comprises a plurality of monitoring equipment, and further comprises a region construction unit for marking power network nodes as power grid nodes and dividing the power grid nodes into a plurality of independent rectangles; a plurality of maintenance points are preset, and the affiliation of each independent rectangle is determined; the icing early warning unit is used for acquiring environment data and transmission conductor data in an independent rectangular range, analyzing and calculating a line influence value, a wind speed influence value, a humidity influence value and a temperature influence value, and calculating an icing early warning value for judging the future icing condition of the transmission conductor; and the icing monitoring unit obtains the current wind speed, the current wind direction and the tension at the two ends of the power transmission conductor, constructs a wind vector by combining the wind speed and the wind direction, performs vector decomposition to obtain a second vector on the cross section of the power transmission conductor, and analyzes and calculates attachment parameters for judging the current icing condition of the power transmission conductor based on the second vector and the conductor gravity vector.