39 results about "Embedded processing" patented technology

This invention discloses an intelligent management method for digital creative marketing scenarios based on multimodal feature fusion, relating to the field of multimodal feature fusion technology. The method includes the following steps: processing raw data to obtain short-term and long-term interaction sequences; performing fixed-length processing on the short-term interaction sequences to obtain fixed-length short-term interaction sequences; performing fusion processing, first embedding processing, and second embedding processing on the multimodal feature set of creative content to obtain creative content embedding matrices, location embedding matrices, and time embedding matrices; processing the fixed-length short-term and long-term interaction sequences to obtain user short-term and user long-term interest vectors; adaptively fusing the user short-term and user long-term interest vectors to obtain a user dynamic interest representation; calculating a balanced interest score matrix based on the user dynamic interest representation; and performing propagation processing through an improved graph neural network model to generate a personalized creative content recommendation list.

This invention discloses a multi-module adaptive wall-climbing cleaning system and its control method. The system includes: a sensing module integrated on a main frame, which acquires wall environment data in real time and inputs it into a control module; a control module, including an embedded processor platform and multiple communication interfaces, which identifies contaminated areas by processing the received wall environment data, generates path instructions using a gray wolf optimization algorithm, and generates cleaning and adsorption instructions based on the contaminated area identification results; a motion module, which moves and adjusts the system's posture based on the path instructions; a cleaning module, including a replaceable cleaning tool unit, which matches the corresponding cleaning tool to the cleaning tool unit based on the cleaning instructions and performs cleaning; and an adsorption module, which dynamically adjusts the adsorption force based on adsorption instructions and real-time acquired adsorption state data to attach the system to the wall surface to be cleaned. This invention enables efficient, safe, and precise cleaning of complex walls, improving the level of automation.

Embodiments of an implantable system and apparatus for providing electrical and optical deep brain stimulation are disclosed. Through the biological introduction of light-responsive opsins to neurons, the system is operable to transform neurons into optogenetically programmable state and history-dependent logical transformers. Embedded processors of the system are operable to stimulate individual neurons using phase-shifted electrical and / or optical signals through a neural interface comprising a plurality of carbon nanotubes, then electrically and / or optically measure the neurons. Due to the parallelism and chemical superposition of the human brain, the implantable system is operable to establish a quantum-like processing ability using its neural interface.

A large model-based medical information recommendation method, a device, and a computer-readable storage medium. The method includes the following steps of: S11, receiving basic medical inquiry information; S13, performing sentence embedding processing on the basic medical inquiry information and converting it into a first digital vector; S15, performing matching processing in a large model based on the first digital vector to obtain first recommendation information, wherein the first recommendation information includes several first information items and a similarity degree of each first information item, and the several first information items are arranged in order of the similarity degree; S17, returning the first recommendation information; S21, receiving a selection of one or more first information items in the first recommendation information from a user; S23, optimizing sentence embedding processing of the basic medical inquiry information and training the large model based on selected first information items.

The application relates to a multi-stage Boot starting method of an embedded processor single starting mirror image and an embedded system; the method comprises the following steps: configuring a BIF file of the starting mirror image, the BIF file encapsulates and packs program segments in an ELF; adding a configuration segment type in the type of the program segments of the BIF file, including segment contents executed by the configuration segment type in the program segments of the configuration segment type, and setting a segment moving destination address as an entry address of the segment contents executed by the segment; a Bootloader program runs processor execution single-stage loading, and loads program segments in the BIF file according to the BIF file; judging that the currently executed program segment is a configuration segment, moving the segment contents of the configuration segment to an execution address according to the segment moving destination address, and handing over processor running permission to the configuration segment to execute the segment contents; after the execution is completed, handing over the processor running permission back to the Bootloader program, and continuing to load subsequent program segments. The application improves the flexibility and expandability of the single-stage loading Bootloader program under the condition of saving ROM storage space.

The application discloses a gait real-time analysis method and system based on visual skeleton feature extraction, which comprises the following steps: collecting a user walking video through a single camera; extracting human key skeleton points in real time by using a lightweight AI model; calculating anatomical coronal plane, sagittal plane parameters and gait space-time parameters based on the skeleton point coordinates; superimposing the skeleton and parameters on the video in real time and providing voice feedback; starting and stopping through a gesture recognition non-contact control system; automatically storing data and generating a rehabilitation report; the system comprises a USB camera, an embedded processing end and a feedback end with display and audio; the application replaces expensive motion capture equipment with a low-cost single camera, realizes objective quantitative evaluation of gait training, real-time multi-modal feedback and automatic management, significantly reduces medical costs, improves the efficiency and effect of rehabilitation training, and is especially suitable for primary medical institutions.

A semantically aligned low-altitude remote sensing image super-resolution method and system are disclosed. The method includes fusing high-precision 6-axis IMU data and visual sensor information in real time to accurately calculate the relative displacement of four neighboring frames, obtaining a low-resolution image sequence with known, accurate, and stable pixel displacement relationships; and constructing an end-to-end image super-resolution reconstruction network based on offset pixel semantic alignment. The system includes a global shutter camera, a 6-axis IMU, an embedded processor, a synchronization module, and an output module. This invention combines precise hardware control with powerful deep learning reconstruction capabilities, achieving reliable and high-quality super-resolution reconstruction of low-altitude remote sensing images.

This application provides a multimodal safety monitoring system and method for engineering vehicles, belonging to the field of engineering machinery safety monitoring technology. The system includes a multi-source heterogeneous sensing module and an intelligent vehicle-mounted processing terminal. The intelligent vehicle-mounted processing terminal integrates an automotive-grade embedded processor, an in-vehicle touch screen, and an audible and visual alarm. The processor receives and preprocesses sensor data, calculates the dynamic confidence of each sensor in real time, and weights and fuses them to generate environmental obstacle and driver attention information. Simultaneously, it predicts dynamic dangerous movement trajectory sectors based on the vehicle's motion state. The driver's attention is mapped into an attention heatmap, and coupled with the fused obstacle information and dangerous trajectory sectors, a comprehensive risk coefficient for each obstacle is calculated. Visual warnings and audible and visual alarms are output in a graded manner based on the risk coefficient. This invention improves the perception reliability and warning accuracy of engineering vehicles in harsh environments through dynamic confidence weighted fusion and attention-situation coupling risk assessment.

This invention relates to the field of water environment monitoring technology and discloses an online method for measuring water pollutant flux. The method includes using a sensor array module distributed along a water body cross-section to acquire vertical flow velocity profile data, and calculating real-time flow rate by combining cross-sectional geometric parameters. Real-time concentration data of the target pollutant is acquired using either an optical sensor or a pollutant concentration sensor. Dynamic integration is performed by an embedded processor to output the pollutant flux per unit time. An adaptive calibration algorithm is introduced to correct sensor errors using historical data and environmental parameters (temperature, turbidity). This online method and equipment for measuring water pollutant flux can achieve simultaneous measurement of multiple parameters, i.e., simultaneous acquisition of flow velocity, concentration, and environmental parameters, eliminating spatiotemporal difference errors without manual intervention, and enabling continuous measurement.

The present disclosure provides a sea surface temperature prediction method and device, electronic equipment and storage medium, relating to the technical field of ocean science and technology, which comprises: obtaining historical sea surface temperature data of a target sea area; performing image block embedding processing on the historical sea surface temperature data to obtain an embedding vector sequence; performing bidirectional time sequence scanning on the embedding vector sequence to obtain bidirectional time sequence fusion features; extracting memory from the bidirectional time sequence fusion features to obtain time sequence memory features and spatial memory features; fusing the time sequence memory features and the spatial memory features to obtain sea surface temperature spatiotemporal features; and reconstructing the sea surface temperature spatiotemporal features to obtain predicted sea surface temperature data of the target sea area. The present scheme can achieve efficient and high-precision sea surface temperature prediction.

The application discloses an operation device and method based on non-0 starting non-cavity coding. By mapping all-0 coding to a non-zero reference value, using a negative number of inverse code, a full-field coding system without negative zero redundancy and coding cavity is constructed, and the bit-by-bit inversion and carry injection operations required by the traditional complement code are completely eliminated. Full-field coding can realize addition and subtraction operations only by switching the reference mirror image, and addition and subtraction have isomorphism, and the efficiency of multiplication and division operations is close, so that the iteration delay of division is significantly reduced. The application has the advantages of small hardware overhead, high operation efficiency, strong compatibility and expansibility, and is suitable for high-performance operation scenes such as CPU, GPU, AI accelerator, DSP and embedded processor.

The application discloses a six-dimensional force online decoupling processing system integrated with multi-modal monitoring and relates to the technical field of sensors.The system comprises a signal acquisition and conditioning module, a multi-modal monitoring module and an embedded processing core.A physical information enhanced twin forward network is arranged in the embedded processing core, the network parameters of which are determined through training of prior knowledge of physical mechanisms such as the relationship of elastic mechanics, thermal drift characteristics and structural symmetry, the embedded processing core operates a double-loop adaptive decoupling architecture, the fast response loop inputs six original digital signals and multi-modal signals into the twin forward network to output real-time decoupling values in each sampling period, the slow calibration loop updates the state tracking twin network at a lower frequency and iteratively corrects the real-time decoupling values by solving inverse problems, and the confidence assessment unit executes a safety rollback according to the prediction error and output deviation.The application realizes online adaptive decoupling processing of a six-dimensional force sensor.

This invention relates to an online partial discharge detection device and monitoring system. The discharge detection device includes a sensor module for acquiring partial discharge signals; a data acquisition system connected to the sensor module, which includes a multiplexer for multi-channel signal selection and switching, a signal conditioning circuit including a low-noise amplifier and bypass control circuit, an analog-to-digital converter, and an FPGA data acquisition module for receiving digital sampled data and power frequency synchronization signals; a data processing system, which includes an FPGA data processing module with data buffering, trigger control, phase detection, waveform management, and digital filtering functions, and an embedded processor communicating with the FPGA for feature extraction, unsupervised clustering, and supervised classification; and a communication system for transmitting the identification results to a remote monitoring platform. This invention can meet the application requirements of large-scale online monitoring of power equipment.

This invention discloses a fluxgate digital compass and heading reference system for intelligent unmanned aerial vehicles (UAVs), relating to the field of UAV navigation technology. It includes a magnetic field sensing unit, a signal conditioning circuit, and an embedded processing center. The magnetic field sensing unit consists of three orthogonal fluxgate sensors positioned perpendicularly to each other in space. The signal conditioning circuit includes a preamplifier, a bandpass filter, and a phase-sensitive detector connected in sequence. The embedded processing center integrates an error compensation engine and a heading calculation engine. It performs dynamic physical calibration using a built-in temperature sensor and sensitivity temperature coefficient curve, while correcting magnetic field distortion through ellipsoidal fitting logic and hard iron interference correction logic. The heading calculation engine combines accelerometer data with a tilt compensation algorithm to calculate the heading angle. This effectively eliminates the influence of environmental temperature drift, soft iron, and hard iron interference on the measurement results, improving the system's anti-interference capability and heading measurement accuracy under complex dynamic conditions.

This invention discloses a high-resolution timing-programmable burst pulse generator based on an embedded processor, which realizes multi-channel, high-precision, and fully programmable burst pulse output for optical module testing. It combines the advantages of low cost, high integration, low power consumption, and high timing accuracy. It includes an embedded processor module, a host computer configuration module, a USB-IIC bridge communication module, and an embedded main control module. The embedded processor module adopts the GD32E5 series chip and has a built-in SHRTIM timer with a DMA burst transfer unit and multiple pulse output channels. The host computer configuration module establishes data interaction with the embedded main control module through the USB-IIC bridge communication module to provide a visual configuration interface and support users to customize the configuration parameters of each pulse signal. The embedded main control module receives and parses the configuration parameters, stores them in the FLASH parameter storage area, and completes the hardware configuration. The system waits for the rising edge of the external AUX signal to trigger the signal throughout the process. After triggering, it synchronously outputs 6 high-precision pulse signals.

Embodiments of the present disclosure disclose a power generation equipment control method and device, electronic equipment and a computer readable medium. A specific implementation of the method comprises: obtaining a load parameter time sequence matrix and a power generation load prototype matrix; performing variable-by-variable normalization processing on the load parameter time sequence matrix; performing position encoding embedding processing on the normalized time sequence matrix; performing prototype space aggregation and prototype prediction on the position encoding time sequence matrix; performing frequency domain prediction processing on the normalized time sequence matrix; performing inverse transformation processing on the frequency domain prediction matrix; performing adaptive branch fusion and residual correction processing on the prototype prediction time sequence matrix and the frequency domain inverse transformation matrix; performing bias correction and reverse normalization processing on the residual projection matrix; and controlling the power generation equipment to perform a power generation task. The implementation can reduce the calculation resources consumed and reduce waste of power resources.

This invention discloses a voltage transformer fault diagnosis and prediction system, including a data acquisition and preprocessing unit, a signal processing and feature extraction unit, and a fault diagnosis and prediction unit. In the data acquisition and preprocessing unit, the support structure of the diagnostic platform supports a rotating mechanism. The rotating mechanism drives a shielding component to rotate and cover the voltage transformer under test, simultaneously enabling the wiring module to be fitted with terminals and the conductive structure to be in contact with conductive surfaces. Furthermore, the first conductive component of the shielding component is connected to the second conductive component of the diagnostic platform. The data acquisition unit collects multi-modal parameters such as voltage, current, and temperature, and sends them to an embedded processor. An intelligent algorithm engine extracts and fuses fault-sensitive features. The fault diagnosis and prediction unit completes the current fault diagnosis and future fault risk prediction. This invention eliminates the need for manual terminal connection, making operation convenient and detection efficient. It improves the safety of high-voltage operating conditions through shielding protection and power-off linkage, adapts to different equipment specifications, and ensures the reliability of diagnostic predictions.

A life-saving equipment end guidance system and method based on thermal imaging dynamic threshold analysis, comprising a life buoy, a propeller is installed at the rear end of the life buoy; a thermal imaging module and an embedded processing unit are installed on the life buoy, three thermal imaging sensors are respectively installed on the side of the thermal imaging module; the embedded processing unit comprises a temperature quantization FPGA module, a vertical projection positioning module and a motion control module, the temperature quantization FPGA module performs discretization grading on the pixel-level temperature difference based on frame average temperature and dynamic step length, the vertical projection positioning module performs column integral operation on the quantized temperature matrix, and the motion control module generates a control signal according to the azimuth angle to drive the propeller to adjust the heading of the life buoy. The present application overcomes the shortcomings of the prior art, and realizes the rapid identification and positioning of thermal anomaly points by combining thermal imaging technology, dynamic temperature quantization processing of the temperature quantization FPGA module and vertical projection positioning of the vertical projection positioning module.

We provide an AI-powered wearable respiratory monitoring device equipped with multimodal sensing. [Solution] The device 100 includes a flexible band worn on the user's chest and includes a strain sensor 101 for detecting chest expansion and contraction, a bioimpedance sensor 102 for measuring impedance changes during respiration, and a motion compensation sensor 104 for generating motion data. An embedded processing unit 105 receives multimodal sensor signals and performs digital signal processing and in-device artificial intelligence classification to detect respiration patterns and abnormal respiration in real time. A power management unit 106 controls energy distribution within the device, and an output and interface module 103 provides a warning to the user when abnormal respiration is detected. A wireless communication module 107 transmits processed respiration data to an external device via Bluetooth® communication.

This invention discloses a modeling and training method, device, and medium for 12-lead electrocardiogram (ECG) signals. It relates to the fields of medical decision support and deep learning technology. The modeling method includes: acquiring 12-lead ECG signals and constructing a two-dimensional grid input consisting of lead axes and a time axis; performing two-dimensional block embedding processing to generate an embedded feature grid; performing hierarchical windowed self-attention computation within a two-dimensional window, wherein the two-dimensional window covers all lead dimensions in the lead axis direction and performs window shifting operations only along the time axis direction between adjacent layers; limiting the number of downsampling operations in the lead axis direction, and performing downsampling only along the time axis direction after reaching the limit; extracting global representations and outputting prediction results. This invention effectively avoids the irreversible collapse of lead structure information, achieves joint modeling of all leads, and improves the generalization and robustness of the model.

Embodiments of an implantable system and apparatus for providing electrical and optical deep brain stimulation are disclosed. Through the biological introduction of light-responsive opsins to neurons, the system is operable to transform neurons into optogenetically programmable state and history-dependent logical transformers. Embedded processors of the system are operable to stimulate individual neurons using phase-shifted electrical and / or optical signals through a neural interface comprising a plurality of carbon nanotubes, then electrically and / or optically measure the neurons. Due to the parallelism and chemical superposition of the human brain, the implantable system is operable to establish a quantum-like processing ability using its neural interface.

This invention discloses a distributed fiber optic temperature measurement real-time noise reduction system and method based on high-speed USB acquisition and embedded AI processing. It belongs to the field of distributed fiber optic sensing technology. The system includes: a high-speed USB acquisition card, used to acquire backscattered signals in the distributed fiber optic temperature measurement system in a single acquisition at a sampling rate of not less than 250 MSPS, generating raw data frames; and an embedded AI processor, connected to the high-speed USB acquisition card via a USB 3.2 Gen 2 interface, used to receive the raw data frames; the high-speed USB acquisition card and the embedded AI processor use a double-buffered or ring-buffered mechanism for data transmission. This invention effectively overcomes the shortcomings of existing DTS systems, which rely on a large amount of accumulated averaging, resulting in long measurement times and low spatial resolution.

This invention provides a state encoder and encoding method for unmanned systems resistant to spoofing attacks, relating to the field of intelligent unmanned systems technology. The unmanned system state encoder includes an analog multiplexer and an embedded processor; the executable computer program in the embedded processor includes a state data sampling module, a differential quantization encoding module, a data encryption encoding module, an error control encoding module, and a data sequence generation module connected in sequence. For unmanned systems such as drones, unmanned vehicles, unmanned ships, and robots, differential quantization encoding, data encryption encoding, and error control encoding techniques are employed to improve the success rate of detecting network spoofing attacks, reduce or eliminate the impact of network spoofing attacks, enhance the security of data transmission in data communication networks, and thus ensure the effective control of the unmanned system.

The application provides an embedded digital cultural exhibition management and control method and system. The method pre-establishes cultural relic detail information model on the embedded processing terminal of the exhibition equipment for the cultural relic object to be displayed. In the exhibition process, the audience position information and the audience line of sight direction information of the audience are obtained, the audience position information and the audience line of sight direction information are mapped to the display coordinate system of the display interface or the cultural relic surface to obtain the audience attention area, the attention degree index of each interest area is determined according to the spatial overlapping relationship between the audience attention area and each interest area in the cultural relic detail information model and the audience staying time, the target interest area to be recommended is determined according to the attention degree index of each interest area, finally, the exhibition equipment is controlled to perform local detail magnification display in the display interface or the virtual camera view according to the target interest area to be recommended, and the closed loop association between the audience line of sight, the staying time and the cultural relic detail structured knowledge in the exhibition process is realized.

The application discloses an AC sampling system resisting damping oscillation wave interference, and belongs to the technical field of data sampling of a relay protection device.The system comprises sequentially connected collecting modules, a hardware filtering module, an analog-digital conversion module and an embedded processing module; wherein the collecting modules input collected AC signals containing damping oscillation wave interference to the hardware filtering module; the hardware filtering module filters high-frequency components of the AC signals, obtains filtered signals and inputs the filtered signals to the analog-digital conversion module; the analog-digital conversion module converts the filtered signals into digital filtered signals and inputs the digital filtered signals to the embedded processing module; and the embedded processing module extracts, classifies and conditions damping oscillation wave interference of the digital filtered signals, so that the damping oscillation wave interference in the AC signals is removed, and the problem that the AC sampling system is weak in resisting oscillation wave and the oscillation wave is not completely suppressed is solved.

This application relates to the field of tubing inspection technology, specifically disclosing a tubing corrosion detection device and method using a non-uniform motion plunger integrated with a magnetic flux sensor. The tubing corrosion detection device includes a plunger body, a magnetic flux detection module, a motion sensing module, a centering buffer module, and a data processing and transmission module. The magnetic flux detection module is distributed around the outer periphery of the plunger body. The magnetic flux detection module includes a magnetization unit and a magnetic flux sensor array. The magnetization unit is configured to generate a magnetic field to magnetize the tubing wall. The magnetic flux sensor array is configured to collect magnetic flux leakage signals at corrosion defects in the tubing. The data processing and transmission module has a built-in embedded processor configured to perform non-uniform signal compensation. Through the coordinated operation of its modules, this device can achieve online detection of tubing corrosion under non-uniform plunger motion of 3-5 m / s, without production shutdown, improving detection efficiency by 2-3 times.

The application relates to the technical field of liquid detection, in particular to a multi-parameter liquid detection experiment device based on a Michelson interferometer, which comprises an interferometer main body, a signal acquisition module, a data processing unit and a multi-parameter sensor assembly. The interferometer main body is provided with an adjusting groove, a sliding support and a beam splitter seat, the angle of a beam splitter lens is adjustable, a wide-spectrum light source and a photoelectric detector are symmetrically arranged, the signal acquisition module is connected with the photoelectric detector through an optical fiber, the data processing unit adopts an embedded processor, and the multi-parameter sensor assembly comprises temperature, concentration and liquid level sensors and respectively detects liquid related parameters. The application can realize synchronous high-precision detection of liquid refractive index, temperature, concentration and liquid level, has the characteristics of compact structure, simple operation and high automation degree, and is suitable for teaching experiments and scientific research applications.

This utility model belongs to the field of railway dispatching technology, specifically relating to an automatic railway dispatching supervision device. It includes a main unit housing, with an embedded processing motherboard fixedly mounted on the upper side of the housing. A dispatching instruction receiving module is fixedly connected to one side of the embedded processing motherboard, and a dispatching center server is located on one side of the main unit housing. The dispatching instruction receiving module has a built-in Ethernet communication module, and is connected to the dispatching center server via the Ethernet communication module. A device feedback acquisition module is fixedly connected to the side of the embedded processing motherboard furthest from the dispatching instruction receiving module. The entire system forms a structured supervision process from dispatching instruction issuance to device status feedback, automatic closed-loop identification, and abnormal timeout alerts. It can promptly alert dispatchers to timeout events, effectively solving problems such as human error in judgment, response delays, and information gaps in dispatching execution, thus improving the visualization, response transparency, and operational safety of dispatching instructions.

The application discloses an intelligent electrically fused pipe fitting with a leakage monitoring function, a monitoring system and a leakage monitoring method. The electrically fused pipe fitting is provided with a monitoring part. A monitoring sensor is installed in the monitoring part. A detection head of the sensor is exposed and flush with the inner wall of the monitoring part. The detection head comprises two electrodes, namely a contact a and a contact b. The monitoring sensor is connected with the monitoring system and is used for monitoring the leakage condition of a butt joint pipeline in the electrically fused pipe fitting in real time. A monitoring signal is sent to an embedded processor of a cloud platform layer through a communication unit for analysis and judgment. According to the analysis result, an alarm instruction or a maintenance instruction is sent to a person in charge, so that the leaking pipe fitting is maintained or the monitoring sensor is replaced and maintained. The application has the advantages of simple structure, high reliability, low-cost water immersion sensor for the core monitoring element, high compatibility of the overall structure with the traditional electrically fused pipe fitting, small change of the production process, limited cost increase and easy industrialization.

The application provides a system for issuing internet LED display content based on FPGA, relates to the field of LED display screen advertising media, and comprises at least one LED display screen, a cloud server in communication connection with the at least one LED display screen, an LED display screen mainboard arranged on the LED display screen, an embedded processor and an FPGA chip arranged on the LED display screen mainboard, and an internet communication module and a display driving communication module arranged in the embedded processor. The application inspects the last stage of display from the FPGA display driving stage, realizes real-time monitoring of system desktop display data of the embedded processor and safety handshake of the FPGA chip through communication data encryption handshake, guarantees the uniqueness of the condition of legal data display, and once display abnormality is found, the FPGA chip quickly closes display output, so that the safety of LED display screen display content is achieved. Meanwhile, the key is issued through the internet, so that the encryption of each LED display screen and the internet account permission hook are guaranteed.