269results about How to "Improve scalability" patented technology

This application discloses a method and apparatus for constructing a multi-scale adaptive gating MambaPlus network, belonging to the field of artificial intelligence and machine learning technology. The method includes: initializing the network configuration and constructing the basic structure; preprocessing the input data to generate a standard dataset; mapping the input data to the hidden space via an input mapping layer, and extracting backbone features from the Mamba backbone; constructing at least two parallel scale branches in the hidden space to obtain a multi-scale feature set; inputting the backbone features and multi-scale features into an adaptive gating module, dynamically allocating weights and adaptively fusing them through a hierarchical gating mechanism to generate fused features; further enhancing the features through cross-scale attention and feedforward enhancement, and then superimposing the residuals to generate the final discriminative features; finally, completing category prediction and model training evaluation. This application, while retaining the advantages of Mamba's long-range dependency modeling, addresses the problems of insufficient utilization of multi-scale information, poor adaptive feature fusion, and low robustness in complex scenarios.

The invention relates to the technical field of information management, in particular to a material digital management system and method based on a cloud platform. The method comprises the following steps: acquiring material data; distributing a material storage space by using the material data; receiving an operation message of the storage space, and pushing the operation message to an inventory database; creating a circulation record table and an inventory state table; using the operation message in the inventory database to update a circulation record table, and recording circulation data; automatically updating an inventory state table according to the circulation data so as to determine the material inventory capacity; when the material inventory capacity is lower than or equal to a preset safety inventory threshold value, management early warning information is generated; and when the material inventory capacity is higher than a preset safety inventory threshold value, updating the management safety field of the inventory state table. The material inventory is monitored in real time based on the information management technology, the replenishment information is intelligently generated, inventory dynamic management is achieved, and the material management efficiency is improved.

This invention discloses a method for fabricating a highly sensitive long-wavelength infrared superconducting single-photon detector. It fully utilizes and amplifies the acoustic mismatch effect between the superconducting thin film and the substrate to achieve phonon localization enhancement on the superconducting nanowires. The effect of this innovative design is to increase the energy utilization rate of long-wavelength infrared photons on the superconducting nanowires, thereby improving the response sensitivity of the superconducting nanowires to long-wavelength infrared photon signals. Compared to existing designs that reduce the cross-sectional size of the superconducting nanowires, create holes beneath the nanowires, or increase the buffer layer, this invention is easier to implement and has higher scalability.

This invention discloses an automated centrifuge and its control system based on material purification. The centrifuge body includes a centrifuge body and an automated control system used in conjunction with it. The centrifuge body is electrically connected to the automated control system, enabling automated sensing, decision-making, and execution throughout the entire material purification process, constructing a closed-loop control system of "collection-analysis-control-feedback." The centrifuge body includes a rotating drum, a drive system, a feeding system, a discharging system, an automatic unloading mechanism, and a safety protective shell. The rotating drum, as the core component for material separation and purification, is made of stainless steel or titanium alloy. This invention relates to the field of material purification equipment technology, specifically enabling automated control of the entire material purification process, improving purification accuracy and production efficiency, enhancing equipment operational safety, realizing process formula storage, data traceability, and production line linkage, and reducing reliance on manual labor.

The application discloses a multi-direction locking type tooling fixture for a numerical control machining center, which comprises a cabinet containing a platform plate, a clamp assembly is arranged on the platform plate, each clamp assembly contains two groups of linear motors, one group of linear guide rod modules, one group of rotary cylinders and one group of clamp carriages, the two groups of linear motors are arranged in parallel, the linear motor contains a first moving carriage, a bearing assembly is arranged on the first moving carriage, a limiting track is arranged on the lower surface of the linear guide rod module, the limiting track is in limiting sliding connection with one of the bearing assemblies, the linear guide rod module contains a second moving carriage, the rotary cylinder is fixedly installed on the second moving carriage, the rotary cylinder is in driving connection with the clamp carriage, and the clamp main body is installed on the clamp carriage. The linear guide rod module is driven by the two groups of linear motors, the linkage control of the rotary cylinder is combined, multi-dimensional accurate positioning of the clamp main body in the horizontal, vertical and rotary directions is realized, and the adaptability to special-shaped workpieces is remarkably improved.

The invention discloses a small-size high-efficiency large-current lithium battery active equalization circuit and a control method thereof, the small-size high-efficiency large-current lithium battery active equalization circuit comprises active equalization modules which are connected in parallel between two connected battery cells, and each active equalization module is composed of four switching tubes and a group of resonance branch circuit which is formed by connecting an inductor and a capacitor in series; n-1 active equalization modules are needed for a battery pack composed of N battery cells. The LC series branch serves as an energy temporary storage unit to replace a traditional inductance type single inductor or capacitance type multiple capacitors, the number of elements is reduced, meanwhile, low-loss energy transfer is achieved through resonance characteristics, and the equalization scheme has the advantages of being small in size, large in equalization current, small in equalization energy loss between battery cells, high in efficiency and the like.

This invention provides a coherent anti-Stokes Raman microscope system, its driving method, and its microscopic apparatus. The coherent anti-Stokes Raman microscope system includes: a femtosecond laser front-end module, a dual-wavelength module, a pulse broadening and amplification module, a time delay module, a spectral broadening and filtering module, a second harmonic generation module, a dispersion compensation module, and a microscope module. The self-phase modulation spectral selection (SESS) driven coherent anti-Stokes Raman microscope system of this invention uses a self-phase modulation spectral selection (SESS) method to convert a 1550 nm pulse wavelength to 1600 nm. Through a frequency doubling crystal, a second harmonic generation process is achieved, converting the 1600 nm pulse wavelength to 800 nm, thus realizing coherent anti-Stokes Raman microscopy imaging, i.e., selective imaging of various biological tissue components. Compared with other wavelength conversion methods, the device is simple, has high conversion efficiency, is tunable, and has strong scalability.

The application belongs to the technical field of direct current substation, and particularly relates to a battery internal resistance detection device. The application comprises an embedded monitoring host module for receiving and displaying battery internal resistance data, having an alarm and a background monitoring communication interface, the embedded monitoring host module being connected with a communication module, the embedded monitoring host module being connected with an information convergence module through the communication module, the information convergence module being connected with a single-section internal resistance sampling module, the single-section internal resistance sampling module being connected with a battery module; wherein the number of the single-section internal resistance sampling modules is the same as the number of the battery modules, each single-section internal resistance sampling module is directly connected with each corresponding battery module through a short connecting line, and is connected in parallel to the information convergence module through a pressure contact crystal head connecting line. The application has the advantages of efficient expansion, adaptation to different sizes of battery groups, short sampling line distance, simple installation process, high detection precision, low cost and good adaptability of the device.

This invention discloses a remote sensing image description evaluation method based on cycle consistency, comprising: generating fine-grained text descriptions of the original remote sensing image using a remote sensing multimodal large model; inputting the generated long text descriptions into a text-to-image model for image reconstruction, and introducing a weighted basis image strategy to compensate for information loss during the text-to-image conversion process; using a multimodal quality assessment model to compare and score the cycle consistency of the original image and the reconstructed image from dimensions such as information sufficiency, scene consistency, and spatial layout, and performing quantitative evaluation according to a preset six-level alignment standard, followed by normalization processing to obtain the final description quality evaluation result. This method solves the problem that existing remote sensing image description evaluation indicators fail in long text and fine-grained scenarios, achieving objective and reliable automatic evaluation without relying on reference text. It is suitable for applications such as remote sensing multimodal model performance evaluation, large-scale dataset quality screening, and intelligent geographic information perception.

The application provides an efficient interactive elevator control management method and system based on elevator code scanning reservation. Through code scanning reservation, cloud intelligent scheduling, voice interaction and other technical functions, the defects of the background technology are effectively solved, and efficient elevator control and high-quality user experience are realized. The system takes "reservation allocation and execution" as the main line, supports internal users (enterprise employees) to scan and reserve elevators, allocates elevators through intelligent algorithms, and realizes efficient elevator control. The system has good scalability (supports adding new elevators and users), security (identity verification and data encryption), user experience (real-time status display and voice interaction), and can be widely applied to office buildings, shopping malls and other scenarios.

The present application relates to a kind of borehole type static sounding device, it includes hydraulic power device, photoelectric separation cabin, suspension device, valve cabin, collection battery cabin, penetration propulsion device and probe in turn from top to bottom connection;Wherein, hydraulic power device is connected with external electricity by bearing cable interface, photoelectric separation cabin is connected with suspension device by axial positioning elastic card mechanism, suspension device is set in the middle position of borehole type static sounding device, hydraulic power device and collection battery cabin, penetration propulsion device are set to be located in the two sides of suspension device.The borehole type static sounding device of the present application can improve its stability when measuring, improve measurement accuracy.

This invention discloses a parallel optimization system for massive video processing. The system adopts a four-layer integrated parallel and collaborative processing architecture, comprising, from top to bottom: a parallel task layer for video stream access, grouping, splitting, encapsulation, and queue management; a resource abstraction layer for unified hardware modeling, status acquisition, topology construction, and capability assessment; a parallel scheduling layer for task-hardware matching, load balancing, priority scheduling, and dynamic adjustment; and an execution optimization layer for data stream localization, parallel read / write, cache optimization, and zero-copy processing. The parallel task layer, resource abstraction layer, parallel scheduling layer, and execution optimization layer form a complete parallel processing link: task input, resource awareness, accurate scheduling, and optimized execution. This invention implements concurrent rate limiting and smooth access for the input video stream to prevent traffic surges; and sets synchronization points and timing control for parallel tasks to ensure orderly output.

This application provides a portable oil and gas well fluid level monitoring system and method, relating to the field of underground wellbore fluid level monitoring technology. The system includes: a virtual instrument platform configured with a LabVIEW-based program for casing pressure signal acquisition and control, a sound-generating device control program, and a sound wave signal acquisition program; a portable acquisition instrument including: a USB data acquisition card, a pressure transmitter, a sound-generating module, and a microphone; the pressure transmitter acquires pressure information at the inlet of a normally closed solenoid valve within the sound-generating module; when the pressure information reaches a preset pressure, a pressure ready signal is generated; when the pressure ready signal is valid, the sound-generating module is controlled to generate infrasound; and the infrasound reflection signal received by the microphone is acquired, processed by a MATLAB-based signal processing program, and displayed in the user interface of the sound wave signal acquisition program. The implementation method provided in this application is beneficial for improving the efficiency and safety of oil well production.

A kind of distributed video splicing device, it is related to video splicing technical field, including with several local input modules of external image device connection, with the management module of several local input modules connection, with the output module of management module connection, with the network streaming media input module of management module connection, output module is connected with external display device, local input module includes with the input interface of external image device connection, with the input conversion chip of input interface electrical connection, with the input FPGA of input conversion chip electrical connection, input FPGA is electrically connected with management module, input conversion chip is used to output RGB signal to input FPGA.The technical scheme provided by the utility model aims to realize integrated design, and can solve the technical problem of network streaming media data processing.

This invention provides an integrated method for pruning and controlling large-scale models of unmanned aerial vehicles (UAVs), belonging to the field of model lightweighting technology. The proposed method significantly reduces model size and computational load, greatly improves inference speed and control response performance, and enables the deployment of complex neural network controllers on resource-constrained airborne platforms. Simultaneously, the lightweight model maintains or approaches the control accuracy and stability of the original uncompressed model. This method has wide applicability and strong scalability, covering various scenarios from basic flight control to complex mission control, providing a feasible, efficient, and cost-effective solution for applying large models to airborne controllers of UAVs, and possesses high engineering practical value.

The utility model provides a kind of system for simulating SPI interface to realize the display of embedded device graphic interface, comprising: main processor, graphic display device, main processor includes embedded main control chip, graphic display device includes LCD display module, embedded main control chip sets up GPIO interface, and GPIO interface connects LCD display module.The utility model can be in the case where hardware SPI interface resource is insufficient or occupied, by using GPIO interface to simulate SPI interface, make full use of the existing GPIO resource of equipment, without additional increase the chip with hardware SPI interface, reduce hardware cost and design complexity, improve the utilization of hardware resources;Through the accurate control of GPIO pin, ensure the stability of simulated SPI interface in data transmission process, effectively reduce the phenomenon such as lag, flicker when graphic display, improve the quality and stability of graphic display, and flexibility and scalability are good.

This invention discloses an efficient QR matrix decomposition method based on FPGA for implementing systolic matrices, belonging to the field of software-defined radio signal transmission technology for airborne networks. It is based on the design of a multi-channel digital front-end implemented on an FPGA chip. The method includes: designing a general handshake coordination mechanism, designing a timing flow for systolic matrix multiplication, performing fast QR matrix decomposition single-loop operations by calculating the Householder mirror transformation matrix, designing a top-level loop pipeline for QR matrix decomposition, and finally verifying the hardware scheme of the fast QR decomposition algorithm. This invention converts resource consumption into timing consumption by designing the overall pipeline logic and loop operations, enabling fast processing and accurate decomposition of digital information in a multi-channel digital front-end on an FPGA, effectively saving on-chip resources. In addition, this invention has excellent scalability, supporting 64 to 128-bit high-precision fixed-point number formats, and the data matrix order can reach up to 128.

This utility model relates to the field of electronic device accessories technology, specifically to a tablet keyboard protective cover and a tablet computer assembly. The tablet keyboard protective cover includes: a first sleeve and a second sleeve rotatably connected. The keyboard is connected to the inner side of the first sleeve. The second sleeve includes a first support plate and a second support plate rotatably connected. The first support plate is rotatably connected to the first sleeve via a first flexible connecting part, and a protective shell is connected to the second support plate. A support member is rotatably connected to the side of the keyboard near the first flexible connecting part. The support member is connected to the first support plate via a second flexible connecting part, so that when the tablet keyboard protective cover is closed, the width direction of the support member rotates to be perpendicular to the plane of the keyboard; when the tablet keyboard protective cover is open to support the tablet computer, the width direction of the support member rotates to be parallel to the plane of the keyboard. This method extends the effective support surface during use, which helps to improve the stable support for the tablet computer.

The utility model discloses a mobile hand material lifting platform with adjustable structure, which is characterized by comprising a lifting mast group, a hand crank mechanism for driving the lifting mast group to lift and adjust, a rope connecting the hand crank mechanism and the lifting mast group, and a moving base installed at the bottom of the lifting mast group and capable of driving the lifting mast group to move vertically or horizontally. The moving base comprises a moving base, a support frame connecting the moving base and the lifting mast group, a plurality of universal wheels installed at the bottom of the moving base and capable of driving the lifting mast group to move vertically, and a large pulley and an angle-adjustable horizontal wheel installed at both ends of the support frame and capable of enabling the lifting mast group to move horizontally. The utility model drives the lifting mast group to move vertically through the universal wheels on the moving base, turns over the moving base and the lifting mast group to enable the large pulley and the horizontal wheel on the support frame to touch the ground, drives the lifting mast group to move horizontally, thereby expanding the range of use and improving the convenience of moving the lifting mast group.

The application provides a regional controller and a vehicle, the regional controller comprising a basic system module and a plurality of independent function modules, wherein the independent function modules are integrated with hardware resources, and the independent function modules are connected to the corresponding first system connectors arranged by the basic system module through the second system connectors arranged thereby, so that the independent function modules are connected to the basic system module. In this way, the independent function modules for realizing functions in the application adopt a modular structure instead of being integrated in the regional controller, and when the configuration requirement of the regional controller changes, the independent function modules can be added or the hardware resources in the independent function modules can be changed in the regional controller, so that the overall hardware development is not required, thereby shortening the development cycle of the regional controller, reducing the cost, and improving the expandability of the regional controller.

The application provides a deep learning model security evaluation system and method, comprising the following steps: S1, collecting information of a to-be-tested model and performing pretreatment; S2, testing the to-be-tested model to obtain a test result; the test comprises adversarial attack test, data poisoning robustness test and backdoor trigger reverse test; S3, summarizing all test results and performing analysis to output a final evaluation result. The application provides security analysis for a trainer of a deep learning model, realizes integration of various attack algorithms, including adversarial sample attack, backdoor attack and data poisoning attack, etc., analyzes performances before and after model attack and defense, enables a user to clearly understand performance changes of the user's model under various attack types, and enables the user to comprehensively evaluate the user's model on a unified platform and better grasp security of the model.

This invention discloses a dual-spectrum hierarchical architecture and CT detector based on X-rays, relating to the field of medical CT equipment technology. It includes: a scintillator layer capable of receiving high-energy X-rays and low-energy X-rays respectively; and several sets of metal sheets positioned on one end face of the scintillator layer, wherein the metal sheets can block high-energy or low-energy X-rays irradiating a portion of the scintillator layer. This invention solves the technical problems of complex structure, low yield, and high cost of existing dual-spectrum CT detectors.

This disclosure provides a motion planning method, apparatus, computer device, and storage medium for robots, applied to each robot in a robot swarm; each robot maintains its own local factor graph; the local factor graphs maintained by the multiple robots constitute a global factor graph; the method includes: in response to a state update event corresponding to the current robot being triggered, determining a trigger domain, and determining a replanning subnet from the current global factor graph based on the trigger domain; performing message passing processing based on Gaussian belief propagation in the replanning subnet with the goal of reducing the energy change of the local factor graph corresponding to the current robot, to obtain an updated local factor graph corresponding to the current robot; determining the confidence distribution information of the variables corresponding to multiple variable nodes in the local factor graph of the current robot based on the updated local factor graph of the current robot; and determining the motion replanning result of the current robot based on the confidence distribution information.

The application discloses a kind of microcavity optical frequency comb ultra-high repetition frequency precision measurement system and method, solve the existing microcavity optical frequency comb repetition frequency measurement system because of facing electronic device frequency bandwidth bottleneck lead to the problem that measurable range is seriously limited, including optical rough measurement unit, comb tooth processing unit, modulation frequency shift unit, reference equalization unit, beam combination beat frequency unit and microwave precision measurement unit;Optical rough measurement unit is used to measure the repetition frequency of the light beam to be measured, and is divided into measurement beam and reference beam;Comb tooth processing unit is used to filter out target comb from measurement beam;Modulation frequency shift unit is used for electro-optic modulation to target comb;Reference equalization unit is used to control the reference comb power of reference beam to obtain better beat frequency detection effect;Beam combination beat frequency unit is used for optical beam combination beat frequency between target comb and reference comb and is divided into two beams;Microwave precision measurement unit is used to detect the beat frequency signal between target comb and reference comb, to realize the precision measurement of repetition frequency.

The application provides a W-band active phased array antenna, comprising: an antenna array surface, a TR component, a power division module, a wave control module, an extensible structure support and a heat dissipation device; the antenna array surface adopts a waveguide slot array line array antenna form, a feed line is integrally designed with the antenna array surface, loss is low, and efficiency is high; radio frequency connection of the antenna array surface, the TR component and the power division module adopts a waveguide direct connection mode, which can significantly reduce radio frequency connection loss; the TR component and the wave control module are interconnected through a low-frequency connector wire throwing, which is flexible and convenient; the application has simple structure, is convenient to expand into a large-scale array, has high structural reliability, is suitable for a star-borne environment, and is particularly suitable for the fields of star-borne communication or radar technology.

This invention discloses a cross-shaped microstrip filter, belonging to the field of microwave RF passive device technology. The filter includes a dielectric substrate, an input microstrip line, an output microstrip line, two sets of series-arm resonant units, and two sets of parallel-arm resonant units. The series-arm resonant units are composed of microstrip inductors and interdigital capacitors connected in series, with the resonant frequency corresponding to the filter's center frequency. The parallel-arm resonant units are composed of microstrip inductors and surface-mount capacitors connected in series, generating low-frequency and high-frequency transmission zeros respectively. The four sets of resonant units converge at a common node to form a cross-shaped topology. This invention, through a modular cross-shaped architecture, allows direct mapping between the resonant units and the frequency response, simplifying debugging and increasing design efficiency. Modular cascading can achieve high-order filtering, significantly improving band-edge steepness and stopband suppression, making it suitable for microwave RF systems such as communication, radar, and satellite navigation.

The application discloses a multi-unmanned aerial vehicle target search method based on self-attention and reinforcement learning, which converts the target search task of the unmanned aerial vehicle into a multi-agent cooperation problem, takes each unmanned aerial vehicle as an independent agent, adopts a strategy network SA-MADDPG to search for a target, simultaneously adopts a Voronoi diagram to reasonably divide a search area, and uses a target probability map TPM to help evaluate the possible distribution of the target, so that the unmanned aerial vehicle preferentially searches a high target probability area, and finally adjusts a search strategy according to real-time feedback during the execution of the search task, so as to adapt to the dynamic change of the target.

The utility model discloses a modularization pet cabinet, including cabinet, connecting cabinet and buckle connecting piece, and the buckle connecting piece includes first buckle and second buckle, first buckle includes gourd shape card slot, and the first card body that is corresponded with the small groove size of gourd shape card slot is provided on second buckle, and the cabinet is fixed with connecting cabinet through the buckle connecting piece, wherein, first buckle is fixed on the cabinet, and second buckle is fixed on the connecting cabinet. The utility model discloses through the design of buckle connecting piece, can carry out modularization assembly to pet cabinet, make it can according to pet quantity, species and the change of habit and carry out flexible combination and adjustment, satisfy the diversified demand of many pet family, and convenient handling and adapt to different home environment simultaneously.

The application provides a laser-assisted biological 3D printing gun, and belongs to the technical field of biological 3D printing guns. The application integrates substrate printing and biological printing in one printing gun through a control unit to control all printing processes, so that the printing gun is automated. External inflation devices, mechanical arm components and other equipment are used to reduce the weight of the printing gun, so that the overall structure is compact, the volume is small, the expansibility is high, the production cost and the labor cost are saved, and the printing mode assisted by the laser can be used to achieve fine control and output control of biological materials in micro-nano precision, so that the precision of 3D printing is improved.

The embodiment of the application provides an automatic protocol conversion system and method. The system comprises a core board and at least one conversion board; the conversion board is used for acquiring first attribute information of a collection device and first data collected by the collection device; the core board is used for acquiring the first attribute information in the conversion board, determining a first protocol corresponding to the first data according to the first attribute information; the core board is also used for loading a protocol library corresponding to the first protocol in the hard disk to the conversion board in the case that the protocol library corresponding to the first protocol is detected in the hard disk; and the conversion board is also used for converting the first data into second data according to the protocol library corresponding to the first protocol, and sending the second data to the core board, so that the automatic protocol conversion system can realize the communication connection between the newly-added collection device and the core board by means of the conversion board when the hardware interface or the protocol of the newly-added collection device is incompatible with the core board, thereby enhancing the scalability of the automatic protocol conversion system.