Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

22results about How to "Lower integration costs" patented technology

Game anti-addiction face age assessment method and device

The invention provides a game anti-addiction face age evaluation method and device, and relates to the technical field of artificial intelligence, and the method comprises the steps: in response to a verification request of a game application, collecting a face image of a user through an image collection device of terminal equipment; inputting the face image into a lightweight age evaluation model deployed in the terminal device for reasoning, executing living body detection and age evaluation, and determining a living body detection result and age evaluation data; wherein the lightweight age evaluation model is constructed based on a lightweight convolutional neural network, and an age sensitive feature attention module is configured in the model and is used for dynamically improving an endowed weight for an age-related feature region in the face image; the living body detection process and the age evaluation process are carried out cooperatively; and generating a control instruction for the game application according to the living body detection result and the age evaluation data in combination with a preset local anti-addiction rule base.
Owner:CHINA ACADEMY OF INFORMATION & COMM

A method for online search of particle spelling variants and properties available to an agent

This invention discloses an online search method for particle spelling variants and properties usable by intelligent agents. The steps are as follows: inputting the baseline spelling of a particle and the constructed particle-specific prompt words into a large language model to generate a core variant corresponding to the particle; integrating the core variant of each particle with the authoritative benchmark information corresponding to the particle to form a full variant set for each particle; each record in the full variant set contains a unique identifier for a particle, a multi-scene benchmark name, a core variant, and extended attributes; dynamically mapping a remotely callable interface set on the server to a local proxy object on the client; after receiving a query request, the local proxy object parses the particle spelling parameters in the query request and performs multi-dimensional spelling matching in the full variant set to realize online search for particle spelling variants and properties. This invention effectively improves the ability of AI intelligent agents to handle domain-specific tasks.
Owner:INST OF HIGH ENERGY PHYSICS CHINESE ACAD OF SCI

A neural network heterogeneous partition rewriting method and system based on graph intermediate representation

PendingCN122132174AReduce end-to-end inference latencyControl transmission overheadProgram initiation/switchingResource allocationCode generationAlgorithm
This invention relates to the field of artificial intelligence compiler technology, and discloses a method and system for rewriting heterogeneous partitioning of neural networks based on graph intermediate representation. The method includes: obtaining the intermediate representation of the high-level computational graph of the neural network, the hardware configuration of the heterogeneous hardware platform, and the description of the heterogeneous backend devices; extracting subgraph nodes and tensor edges of core operators from the intermediate representation of the high-level computational graph; constructing a cost model based on the hardware configuration; solving the heterogeneous backend device allocation result for each subgraph node; rewriting the allocation result back to the intermediate representation of the high-level computational graph in the form of attributes; generating a wrapper function marked with heterogeneous backend devices; and rewriting the computational graph structure inside the target forward function. This invention realizes hardware-aware graph-level automatic partitioning, tightly coupling the partitioning result with the high-level IR, and connecting to backend code generation through a unified wrapper function interface, solving the problems of existing technologies that rely on static rules, lack quantitative cost models, and have loose coupling between partitioning results and backend calls.
Owner:UNIV OF ELECTRONICS SCI & TECH OF CHINA

Miniature pesticide spraying device for pest control

The utility model discloses a miniature pesticide spraying device for pest control. The miniature pesticide spraying device comprises a pesticide box, a mounting plate and a pesticide spraying assembly, the pesticide box is mounted on the mounting plate, a pesticide outlet is formed in the bottom, one or more pesticide inlets are formed in the top, and a pesticide inlet cover covers the pesticide box and is used for storing pesticide liquid; the mounting plate is of a rigid structure and is used for mounting the medicine box to an automatic machine; the pesticide spraying assembly is detachably connected with a pesticide outlet in the bottom of the pesticide box and used for spraying pesticide liquid. Moreover, the pesticide spraying assembly comprises an angle adjusting mechanism which is used for adjusting the horizontal direction of pesticide liquid sprayed by the pesticide spraying assembly. The device is simple in structure, capable of being compatible with various automatic machines, high in pest control efficiency and capable of working continuously.
Owner:NANJING FORESTRY UNIV

Intelligent vehicle ramp confluence decision-making method considering advanced confluence of rear vehicles into main line and vehicles behind main line

The invention belongs to the technical field of intelligent vehicles, and particularly relates to an intelligent vehicle ramp confluence decision-making method considering advanced confluence of a rear vehicle into a main line and a vehicle behind the main line, which comprises the following steps: acquiring a road level position, a vehicle speed and a longitudinal acceleration of a vehicle, and if the three meet set conditions, judging that the vehicle enters an active confluence decision-making stage; meanwhile, the vehicle carries out panoramic scanning on an acceleration lane behind the vehicle, and potential cooperative vehicles are screened out according to a set mode; behavior characteristics of the potential cooperative vehicles are further monitored, and when set conditions are met, it is judged that the potential cooperative vehicles have an advanced afflux behavior, and safety evaluation is started immediately; and if the potential cooperative vehicle is judged to be safely converged in advance and the collision time of the own vehicle and the potential cooperative vehicle is greater than or equal to 0.8 s, the own vehicle completes convergence with an acceleration curve greater than that of the potential cooperative vehicle, and otherwise, a conservative convergence strategy is executed. The method provided by the invention can improve the success rate, safety and efficiency of intelligent vehicle afflux.
Owner:CHANGAN UNIV

Heat supply operation visualization system based on MCP protocol and large model

The invention discloses a heat supply operation visualization system based on an MCP protocol and a large model. A data resource layer comprises different heat supply business data visualization rules, a scene adaptation rule, a heat supply operation business function module interface building tool, a data visualization tool and a core heat supply business function module interface; the intelligent processing layer identifies a visual display intention through a visual display intention analysis large model, obtains a plurality of visual display sub-tasks, and then determines called corresponding data interfaces and capability units according to the sub-tasks through a visual display business resource matching large model; related information is obtained through interaction of the MCP host, the client and the server, and visual display of each heat supply service function is completed; the user interaction layer is used for obtaining a user visual display intention through a multi-round dialogue mechanism and calling an intention analysis large model; and the server is also used for displaying the visualization effect of each heat supply business function to the user and carrying out iterative optimization according to feedback information.
Owner:HANGZHOU YINGJI POWER TECH CO LTD

Plastic package silicon carbide power semiconductor module

The invention discloses a plastic package silicon carbide power semiconductor module, which relates to the technical field of electronic components and comprises a substrate, a silicon carbide wafer, a plastic package body, power pins, signal pins, a fault detection module, a fault communication module and a linkage protection module. The silicon carbide wafer is arranged on the substrate, and the pins are oppositely arranged on different sides; the plastic package body packages internal devices, and all functional modules are integrated on the substrate. The fault detection module captures various fault signals such as over-current and over-voltage signals in a full-dimension mode and converts the signals into electric signals. The fault communication module establishes two-way communication with an external master controller through a special pin, stably transmits fault data and receives a control instruction; and the linkage protection module constructs closed-loop control, synchronously triggers hardware protection and fault feedback, and executes hierarchical processing. Signal interference is avoided through structure optimization, communication reliability and module adaptability are improved, safe and stable operation of the power conversion system is guaranteed, and the power conversion system is suitable for high-frequency and high-voltage scenes such as electric automobiles and photovoltaic power generation.
Owner:爱微(江苏)电力电子有限公司

Invoice data processing method and related product

PendingCN121860792AReal-time duplication checkingFlexible dockingDatabase updatingFinanceVisual presentationInvoice
The invention discloses an invoice data processing method and a related product. The method comprises the following steps: performing interface configuration on a target system; calling an invoice pool by using the configured interface, and performing multi-dimensional verification on invoice data in the invoice pool to obtain a verified invoice pool; the multi-dimensional verification at least comprises invoice number verification, key field verification and issuing time verification; and if the invoice pool is newly added with the invoice data, updating an input data value based on the newly added invoice data, and visually displaying the updated input data value. According to the invoice management and tax monitoring system and the invoice management and tax monitoring method, the core pain points of high adaptation cost, weak ability of checking true and false invoices, duplication checking lagging, statistical artificialization and the like in a traditional mode are solved, an integrated invoice management and tax monitoring system with automatic acquisition, intelligent checking, real-time updating and visual presentation is further constructed, and the intelligence, compliance and response agility of invoice data processing are remarkably improved.
Owner:太保科技有限公司

Methods and related products for implementing AI Agents to call RPA services based on AI and LLM.

This disclosure provides a method and related products for implementing AI Agent invoking RPA services based on AI and LLM. The method includes: receiving a semantic instruction; determining at least one automation bearer unit corresponding to the semantic instruction, each automation bearer unit carrying at least one automation service; obtaining a list of automation tools corresponding to each automation bearer unit, each automation tool list including descriptive information of each automation service carried by the automation bearer unit; determining at least one target automation service from each automation tool list whose descriptive information matches the semantic instruction; and invoking the corresponding target automation service from each automation bearer unit. This implementation allows users to directly invoke the corresponding target automation service through semantic instructions.
Owner:BEIJING BENYING NETWORK TECH CO LTD +1

Calling method and device of standard component resources in industrial design software, computer equipment and medium

PendingCN121858115AImprove calling efficiencyreduce callsGeometric CADSpecial data processing applicationsUser inputDesign software
The embodiment of the invention provides a calling method and device for standard part resources in industrial design software, computer equipment and a medium, and the method comprises the steps: receiving a matching condition input by a user on a design software platform, screening out a standard part three-dimensional model meeting the matching condition according to the matching condition, and storing the standard part three-dimensional model in the design software platform; taking the model as a candidate standard component three-dimensional model; selecting an adaptive standard component three-dimensional model from the candidate standard component three-dimensional models according to actual installation requirements as a target standard component three-dimensional model; recommending a matched standard component three-dimensional model according to the target standard component three-dimensional model; and calling the target standard component three-dimensional model and the matched standard component three-dimensional model for installation, and displaying attribute information on the target standard component three-dimensional model and the matched standard component three-dimensional model. According to the scheme, the design process is optimized through standard part full-life-cycle tracing and global consistency management, and the standard part calling and assembling time is shortened.
Owner:SHAANXI AIRCRAFT CORPORATION

Task planning method and device, electronic equipment, medium and product

ActiveCN122047972BImprove conversion abilityImprove generalization abilityUser inputIntent recognition
The application discloses a task planning method and device, electronic equipment, medium and product. The method comprises the following steps: obtaining an interactive sentence input by a target user; performing intent recognition on the interactive sentence to obtain an intent candidate list; generating a structured intent object according to the intent candidate list and multi-round dialogue data of the target user; loading a corresponding business domain task graph according to the structured intent object; and performing task planning according to the business domain task graph to obtain a task execution sequence. The method can combine real-time interactive intent of a user with multi-round dialogue history to form a structured intent expression, and complete adaptive task planning based on a matched business domain task graph, thereby automatically generating a task execution sequence that meets current requirements without relying on a static preset process, and effectively improving dynamic adaptation capability and execution rationality of task planning.
Owner:CHENGDU MINGTU TECH CO LTD

An artificial intelligence silent evolution control system based on a fusible, contractible, auditable and interpretable edge protocol

PendingCN122509359AImproved risk interception efficiencysave computing powerControl systemControl engineering
The application discloses an artificial intelligence silent evolution control system based on a fusible, contractible, auditable and interpretable edge protocol and belongs to the technical field of artificial intelligence system safety and reliability management. The system comprises an edge protocol bus, an interpretable module, a fusible module, a contractible module, an auditable module and a silent evolution management layer. The interpretable module calculates the confidence and the feature attribution path of the model output. The fusible module triggers the fuse according to the preset confidence condition and publishes an event. The contractible module executes dynamic resource contraction in response to the fuse event. The auditable module generates a three-tuple association log of the confidence, the fuse action and the contraction action. The edge protocol bus provides standardized event communication, module registration and version negotiation. The application solves the technical problems that the existing AI management module is isolated and cannot actively defend and heat evolve based on the output quality, and is suitable for high-risk AI application scenarios such as finance, medical treatment and automatic driving.
Owner:徐义才

Local area network image transmission method based on two-dimensional code session guidance

The invention relates to the technical field of wireless network communication, real-time audio and video transmission and rapid equipment pairing, in particular to a local area network image transmission method based on two-dimensional code session guidance, which comprises the following steps: after an image transmission device is connected with routing WiFi or a client hotspot, starting an HTTP control service and an audio and video stream pushing service; the plug flow service comprises plug flow based on a UDP (User Datagram Protocol) and an optional RTSP (Real Time Streaming Protocol) session service; and the device obtains a local area network IP and a port to construct a control URI and a media URI, and packages control or media access information, a one-time session token, a validity period, a device public key commitment value and a transmission strategy parameter into JSON session guide information to generate a two-dimensional code for display. The user side scans the two-dimensional code and then verifies the commitment value and the session token, a secure session bound with the equipment is established, and a control channel is upgraded to a WebSocket two-way channel after HTTP handshake; a media channel adopts UDP transmission and feeds back time delay change and packet loss statistics through an RTCP, a sending end performs congestion control based on time delay gradient estimation and an adaptive threshold value, and adaptively selects mixed NACK or FEC packet loss recovery and jitter buffer target delay according to round-trip time delay and a packet loss rate. Therefore, quick connection, impersonation resistance and low-delay stable preview can be realized under the weak network condition.
Owner:SHENZHEN ZKSWE TECH CO LTD

A method for accurately measuring the opening and closing time of a disconnecting switch

This invention relates to the field of online detection and condition monitoring technology for power system switchgear, and discloses a method for accurately measuring the opening and closing time of disconnecting switches. By synchronously collecting voltage data on both sides of the disconnecting switch to construct a voltage difference sequence, the action interval is extracted through differential and sliding window energy analysis. Within this interval, differential features are used to determine the action index and convert it into the action time. The average voltage difference before and after the action time is calculated to determine the action type. The stability of the action time is evaluated by combining a reference action time and statistical analysis of multiple measurement deviations. Simultaneously, key data from the entire process are compiled into an identifiable dataset. Through these steps, the measurement of the opening and closing behavior and action time of disconnecting switches is achieved, overcoming the problems of asynchronous single-sided sampling or mechanical triggering, poor reliability of threshold and zero-crossing detection, and difficulty in quantifying time stability, thus providing an improved means for disconnecting switch condition assessment without adding complex hardware.
Owner:SHANGHAI ELECTRIC PORCELAIN WORKS CO LTD

Flexible tactile sensor and pressure tactile measurement method thereof

PendingCN121994386ASolve the problem of signal-to-noise ratio degradationstable reflection echoForce measurementMeasurement of force applied to control membersTactile sensationPMUT
The invention provides a flexible tactile sensor and a pressure tactile measurement method thereof, and the flexible tactile sensor comprises a piezoelectric micromachined ultrasonic transducer (PMUT) array layer which is used for transmitting an ultrasonic signal and receiving a reflected ultrasonic signal; the ultrasonic reflecting layer is arranged on a sound wave transmitting path of the PMUT array layer and is used for reflecting the ultrasonic signal; and the flexible ultrasonic coupling layer is arranged between the PMUT array layer and the ultrasonic reflecting layer and is used for conducting ultrasonic signals and generating deformation in the thickness direction in response to external pressure so as to change the propagation path length of the ultrasonic signals in the flexible ultrasonic coupling layer. According to the invention, an ultrasonic reflection mechanism irrelevant to target acoustic characteristics is realized in the PMUT flexible touch sensor for the first time, the problem of degradation of the signal-to-noise ratio in a low-impedance / sound absorption target scene is fundamentally solved, and the universality and measurement consistency of the sensor are remarkably improved.
Owner:TIANJIN UNIV

Method for filling the gap of a clamping jaw for adapting to different sizes of medicine bottles

PendingCN122290124Aaccurate identificationPrevent incorrect completionMedicineImage segmentation
This invention relates to the field of image data processing, and in particular to a method for filling gaps caused by grippers on medicine bottles of all sizes. The method includes acquiring an original image containing the medicine bottle and grippers; extracting and preprocessing an initial segmentation mask of the bottle body using an image segmentation model; extracting the edge contour of the preprocessed bottle body mask; determining and locating the gap region and extracting gap parameters; planning a filling path based on the gap parameters to fill the gap region pixel by pixel; optimizing the contour smoothing of the filled complete bottle body mask; verifying the contour integrity and connection accuracy of the filled mask; and outputting the complete bottle body segmentation mask to a size measurement module after passing the verification. This application achieves the filling of gaps caused by grippers on medicine bottles of all sizes, with a single-frame mask filling time ≤10ms. The filled bottle body mask is used for subsequent pixel-level size measurement, with a size measurement deviation ≤0.1mm.
Owner:美蓝(杭州)医药科技有限公司

Method, device and storage medium for improving repeatability of spectral prediction results

The application relates to a method and device for improving the repeatability of spectral prediction results and a storage medium, and belongs to the technical field of spectral measurement. The method comprises the following steps: obtaining sampling data of a to-be-measured sample, each sampling data comprising N sub-sample data; for each sampling data, performing data rejection on the N sub-sample data based on the median of the N sub-sample data in the sampling data and a pre-set range threshold, to obtain processed sub-sample data; constructing M normal distributions based on the processed sub-sample data based on the median; determining a sampling result of the sampling data based on the average value of the median of each normal distribution, and the sampling result is used for spectral prediction. Since the sub-sample data in any distribution form is converted into a connection composed of M normal distributions, the repeatability of the spectral prediction result can be improved, meanwhile, a spectral acquisition device with higher repeatability accuracy is not required, and the research and development cost can be reduced.
Owner:INTELLIGENT ANALYSIS SERVICE CO LTD

Assembly type network informatization integration platform

The invention discloses an assembly type network informatization integration platform which comprises a platform service bus module, a module interface specification module, a module registration and version management module, an assembly arrangement module and at least one of an integration adaptation module, a configuration template and dependency verification module, a deployment delivery module and an operation monitoring and auditing module. The platform carries out registration and version management on the function module through a standard interface and a contract, an assembly list is generated through assembly arrangement, protocol conversion and data mapping are achieved through an adapter unit, a verification module completes dependency and configuration consistency verification, a delivery module supports release and rollback, and a monitoring and auditing module achieves operation monitoring and trace auditing. And the integration efficiency and the traceability are improved.
Owner:WUXI HUISHAN DISTRICT PEOPLES HOSPITAL

A radar frequency band fusion extension method based on MP and root-MUSIC algorithm

The application belongs to the technical field of radar target recognition, and particularly relates to a radar frequency band fusion expansion method based on MP and Root-MUSIC algorithm. The method comprises the following steps: step 1: setting two adjacent radars, the observation angles of which are the same, and the working frequencies of which are different, so that a full-polarization GTD model of low-frequency and high-frequency electromagnetic echo is established for a stationary target composed of scattering centers; step 2: using echo data and conjugate data, incoherent parameters, i.e. linear phase and fixed phase, are estimated; step 3: using an improved algorithm to process coherent data, poles and amplitudes of a wideband full-polarization model are estimated, and then original data of low-frequency subband and high-frequency subband are used to calculate a fusion signal of a full frequency band. The multi-frequency band fusion method provided by the application uses original data and a conjugate form to construct a new covariance matrix, so that the modified algorithm has better parameter estimation performance under the same signal-to-noise ratio.
Owner:AIR FORCE UNIV PLA

Body intelligent robot skill runtime data analysis and scheduling method

PendingCN122584290AMeet the needs of ultra-high precision scenariosImprove resolution accuracy
The present application relates to embodied intelligent robot technical field, specifically to embodied intelligent robot skill runtime data analysis and scheduling method. The method comprises: a skill data receiving module receives skill data packets of external data sources and completes format verification and integrity check; a multi-level semantic analysis engine carries out three-level progressive analysis of syntax layer, semantic layer and intention layer on the skill data packets, and generates executable action sequences; a real-time scheduling controller combines the current state of the robot and environmental information to prioritize and time slice allocate the action sequences, and generates a scheduling scheme; an adaptive optimization module continuously iterates scheduling strategy parameters based on historical execution data and real-time feedback. Through the technical framework of multi-level semantic analysis and dynamic scheduling cooperation, the present application solves the problems of information loss, response delay and insufficient adaptability in skill transfer process, and significantly improves the precision and task completion efficiency of robot skill execution in complex dynamic environment.
Owner:深圳复现范式科技有限公司