171 results about "Recurrent neural nets" patented technology

The invention discloses a concrete internal defect quantitative detection method and system based on reflected waves, and belongs to the technical field of nondestructive testing. A reflected wave data matrix is obtained through multi-angle excitation and synchronous receiving; calculating energy characteristics of each channel, and constructing an energy response residual field; extracting waveform offset, spectrum jitter and phase change caused by defects by adopting a disturbance comparison algorithm to form a disturbance feature vector set; a defect-response function curved surface is further constructed, a defect topological structure is inversed based on gradient and curvature analysis, and defect geometric parameters are output; and finally, inputting the multi-moment defect parameters into the recurrent neural network, and predicting a defect evolution path and a failure risk. The method has high resolution and trend prediction capability, and is suitable for detection and early warning of concrete structures in bridges, tunnels and nuclear power projects.

The invention provides a hydro meteorology early warning method for an offshore oil and gas platform, and belongs to the technical field of offshore hydro meteorology. Extreme weather events are identified by adopting minimum probability abnormal event identification vectors to match abnormal characteristic parameters, and abnormal signal characteristic parameters are input into an ocean dynamics prediction model to calculate real-time sea condition parameters; calling a multi-temporal-spatial-scale early warning fusion matrix to combine with a wavelet decomposition technology and a recurrent neural network to realize multi-scale information integration, analyzing an environmental parameter change trend through a sea condition jump identification model and triggering an emergency response, dynamically adjusting system parameters according to a stability evaluation index vector, and optimizing prediction precision by adopting an early warning residual value compensation matrix. And finally, multi-level early warning information is generated and a real-time early warning notification is sent to an operator, so that the technical problem of insufficient early warning precision of an offshore oil and gas platform hydro meteorology early warning system in multi-spatio-temporal scale data fusion processing is solved.

The invention relates to a multi-source disturbance perception and self-adaptive regulation and control technology in an electroplating production process, and aims to solve the problem that the process uniformity is difficult to guarantee due to diverse and dynamic changes of disturbance sources in an electroplating process. The invention provides a regulation and control method taking multi-dimensional disturbance sensing signal acquisition, preprocessing, recurrent neural network prediction, spatial distribution modeling and meta-learning reinforcement learning regulation and control as the core. The method comprises the following steps: acquiring various disturbances and auxiliary parameters such as fluid, temperature and current in real time, performing denoising, normalization and time sequence alignment processing, dynamically predicting a disturbance trend by using a recurrent neural network, and combining with process parameters to form high-expression feature input. And training the reinforcement learning regulation and control model by using a meta-learning algorithm to realize rapid adaptive optimization of the process parameters in the new disturbance scene. According to the system, through real-time feedback and dynamic self-optimization, the electroplating uniformity and the production stability are improved, and the robustness and the self-learning ability to a complex disturbance environment are effectively enhanced.

The invention provides a robot social adaptive navigation knowledge learning and migration method and system, and relates to the field of mobile robot navigation. Aiming at the problems that an existing path planner lacks time sequence memory and neglects pedestrian social intent, a man-machine co-fusion scene is constructed, a training set containing an expert demonstration path is made, and a recursive generation model is input; designing a recurrent neural network embedded RRT, generating an RNN-RRT planner, and fusing historical information and pedestrian convergence probability in training; new scene loading training parameters are finely adjusted to realize knowledge migration, loss convergence or output RNN final parameters after reaching a preset round number. According to the method, the path anthropomorphism and generalization ability are improved, and the method is suitable for complex human-computer interaction scenes.

The invention relates to the technical field of computer vision and multi-sensor data fusion, in particular to a multi-sensor image fusion obstacle real-time detection and tracking system, which comprises the following steps of: firstly, fusing data of a camera, a millimeter wave radar and a laser radar, and extracting and fusing multi-modal features; performing multi-target tracking based on a recurrent neural network: generating a target state through space-time modeling, associating a target with a historical track by using an attention mechanism, and maintaining track consistency; and the system performs semantic classification and interaction analysis on the obstacle, predicts the movement track of the obstacle, realizes deep semantic understanding, and finally outputs the identity label and the complete historical track of the obstacle.

The invention relates to the technical field of photoelectric pod target tracking, and discloses a multi-mode target automatic identification tracking method and system based on a photoelectric pod. A multi-modal data acquisition module of the system integrates visible light, infrared thermal imaging and laser ranging sensors based on a photoelectric pod, acquires target multi-modal sensing data and completes time-space synchronous calibration; the dynamic feature extraction module adopts a multi-scale convolutional neural network to process multi-modal data, generates a target multi-level feature map and performs significance region labeling; the adaptive tracking decision module deploys a recurrent neural network to predict a target motion trajectory according to a labeling result, and generates a tracking control instruction through an optimization algorithm; the execution control module drives the photoelectric pod holder mechanism and adjusts the orientation of the sensor to realize target tracking; the feedback optimization module monitors the tracking state in real time, calculates a tracking deviation index, and generates a strategy adjustment parameter to dynamically optimize the tracking strategy.

The invention discloses a low-altitude large-speed small-radius maneuvering control method for an unmanned aerial vehicle. The method comprises the steps of data acquisition and preprocessing, establishment of an accurate unmanned aerial vehicle model, design of a model prediction controller, sensor fusion, real-time control and the like. Flight state data is collected and preprocessed through various sensors, dynamics and kinematics models considering aerodynamic nonlinearity are established, a model prediction controller is designed by using a recurrent neural network and a particle swarm optimization algorithm, sensor fusion is performed by using an extended Kalman filtering algorithm, and a model prediction model is established. Accurate control of the attitude and track of the unmanned aerial vehicle is realized. Meanwhile, the functions of environment perception and obstacle avoidance, fault diagnosis and fault-tolerant control, real-time controller parameter adjustment, target optimization and the like are added, the maneuvering performance, adaptability and safety of the unmanned aerial vehicle are improved, and the unmanned aerial vehicle can be widely applied to the fields of military reconnaissance, logistics distribution, film and television shooting and the like.

The invention belongs to the technical field of aircraft control, particularly relates to a range extender control system for an unmanned aerial vehicle and a controller, and aims to solve the problems of limited endurance, poor flight stability and the like caused by discontinuous energy supply and power response lag. The system comprises a range extender power source module, an electric energy conversion and distribution module, a flight state sensing module, a load power prediction module, a multi-target optimization scheduling module and a closed-loop feedback execution module. Through flight state real-time perception and load power prediction based on a recurrent neural network, an optimal power generation instruction is generated in combination with multi-target optimization scheduling, and precise rotating speed tracking is realized through adaptive PID control. A fuel consumption, power supply smooth switching and battery health combined cost function is introduced, dynamic weighting is carried out according to task types, and the cruising ability and the system robustness are improved; the system has the functions of fault emergency response, power battery hot plug and model online updating, and safety and maintainability are enhanced.

The invention discloses a time-varying and time-lag multi-CSTR system adaptive control method and system based on a neural network, and the method comprises the steps: collecting key operation parameters of a multi-CSTR system in real time, constructing a time-varying and time-lag dynamic model through a recurrent neural network, and updating a time-lag parameter estimation value. And designing a hierarchical neural network structure comprising an upper global coupling model and a lower local compensation model, designing an adaptive algorithm based on a stability theory to adjust the weight of the neural network, and generating a control instruction to drive an execution mechanism. The system correspondingly comprises a data acquisition layer, a time-varying time-delay estimation module and the like. According to the scheme, the problems of time varying, time lag and coupling of the multi-CSTR system are solved, and stable operation of the system is guaranteed.

The invention relates to the technical field of digital network broadcasting, in particular to a multifunctional integrated digital network broadcasting system based on an intelligent algorithm. Comprising a multi-modal data acquisition module, a broadcast content generation module, a broadcast scheduling decision module, an abnormal interference suppression module and a terminal collaborative feedback module. The multi-modal data acquisition module collects multi-source information of a coverage area in real time. And the broadcast content generation module performs intelligent clustering analysis based on the information to generate structured content. And the broadcast scheduling decision module adopts a graph attention scheduling algorithm to optimize a scheduling strategy. And the abnormal interference suppression module identifies an interference source by using a residual convolutional network, and repairs an interfered signal through a recurrent neural network. And the terminal collaborative feedback module executes playing and data feedback, and dynamically adjusts receiving parameters and pushing strategies by means of federal reinforcement learning. The intelligent level, the adaptability, the user satisfaction and the overall service efficiency of the digital network broadcasting system are remarkably improved.

A recurrent neural network includes a plurality of n damped harmonic oscillators (DHOi), each of the n damped harmonic oscillators being one cell (nci) of the neural network, an input unit (IU) that receives and inputs time-series input data (S (t)), a recurrent connection unit (RCU) that includes, for each of the cells (nci), at least one connection (wi,j) between the input / output node (IOi) of the corresponding cell (nci) and the input / output node (IOj) of at least another one of the cells (ncj) for transmitting the resulting damped harmonic oscillation (hi) output from the input / output node of the corresponding cell (nci) to the input / output node of the another one of the cells (ncj).

Apparatus and methods detect sleep staging events. The apparatus (100) may be configured to obtain a facial biopotential signal measured between two electrodes connected to a user's face, which may be configured to form a transverse-ocular measurement vector. The biopotential signal may be measured by a biopotential measurement device comprising the two electrodes. The apparatus (100, 200) may be configured to derive, from the facial biopotential signal, a plurality of biosignals suggestive of sleep staging events. The apparatus may be configured to classify individual segments of the plurality of biosignals as belonging to one of a plurality of sleep stating events. The classification may involve using a trained machine learning model, such as a recurrent neural network, to predict sleep staging events for different segments or epochs of the plurality of biosignals.

The invention provides a three-coordinate measuring machine adaptive dynamic error compensation method, and relates to the field of three-coordinate measuring machines, and the method comprises the steps: obtaining multi-source state data of a three-coordinate measuring machine, and calculating a real-time dynamic error, the state data comprising a motion state, a dynamic response and environment disturbance data; based on the real-time dynamic error, using a recurrent neural network to construct a virtual measuring machine model, and performing offline training to obtain an initial error prediction model; and acquiring real-time error feedback data, and performing fine adjustment on the initial error prediction model by using an incremental learning algorithm and a sliding time window mechanism to obtain a prediction model capable of dynamically evolving and aging adaptive parameters. The method is used for overcoming the defect that in the prior art, a linear model or a fixed compensation parameter is difficult to accurately describe and compensate all dynamic errors sometimes.

The invention relates to the technical field of coal-fired unit energy flow analysis and prediction, in particular to a coal-fired unit boiler energy flow analysis and prediction method under rapid variable load, which specifically comprises the following steps of: firstly, combining a thermodynamic physical framework with a recurrent neural network, and representing a function relationship between a time-varying parameter and a state variable in a mechanism model through a neural unit; rapid fluctuation of energy in the boiler is accurately captured; secondly, a multi-mode switching strategy based on an attention mechanism is introduced, flexible switching is carried out between a strong coupling state and a weak coupling state, and the prediction precision under high-speed load change is guaranteed; and finally, introducing an error compensation model, and dynamically correcting a hybrid model prediction residual error. According to the method, physical constraints of mechanism modeling, multi-mode switching of an attention mechanism and an error compensation model are integrated and fused, so that the problems that most of hybrid modeling studies in the prior art mainly focus on steady-state or finite transient prediction, and studies on energy flow dynamic characteristics under rapid load change are still relatively deficient are solved.

The invention discloses a multi-voice separation method based on a lightweight dual-path Transform network, and the method comprises the steps: collecting audio multi-voice data, and carrying out the preprocessing of the data, and forming a data set; the method comprises the following steps: constructing a dual-path Transform network model DPTNet, and introducing a recurrent neural network to optimize the dual-path Transform network model DPTNet; and training the dual-path network model DPTNet, and performing engineering deployment based on the trained model. The method is beneficial to obtaining higher-quality audio fingerprint recognition capability, sound source separation capability and voice enhancement function, can be used for tracking and positioning the position of a sound source, helps positioning and tracking related applications, can be expanded to the medical field, can be used for heart sound segmentation, namely, recognition of specific signals of the heart, and can be applied to the field of medical science. The method helps to diagnose cardiovascular and other medical problems, and has technical innovation and practical application value.

Examples relate to the field of radio frequency (RF) signal processing such as classifying RF signals with low latency. The method involves receiving portions of an RF signal, transforming these portions into a time-resolved frequency representation using a continuous wavelet transform, and processing this representation with a recurrent neural network. The neural network modifies a neural network state incrementally to generate a classification output, which may include modulation classification, signal-to-noise ratio (SNR) classification, or jamming detection. The system achieves sub-millisecond inference latency through techniques such as model quantization and batch size optimization. Principal uses include real-time RF signal analysis and jamming detection, with applications in communication systems and environmental monitoring. The RF signal may be received from a quantum RF sensor based on Rydberg atoms, enabling broad frequency range detection.

Computer-implemented methods and systems provide a persistent fixed-size recurrent memory that supports long or unbounded sequence processing with substantially constant compute and bounded memory. A recurrent model maintains a matrix state updated per chunk from key, value, gate, and optional control signals; a gated error between value and a key-weighted state yields a rank-1 proposal. Chunk proposals are reconciled coordinate-wise by an order-invariant convex combiner; solitary proposals pass unchanged. Queries can be answered from proposals without materializing a full updated state. Multiple asynchronous agents emit sparse updates with overwrite strengths that are merged using sparse convex or max selection and optional optimizer preprocessing. A feedback projection writes higher layer activations into lower layer state to consolidate durable skills and reduce forgetting. The approach enables durable retention, parallel inference-time adaptation, scalable linear-attention approximation, and persistent fixed memory use over arbitrarily long sequences.

The invention belongs to the technical field of data processing, and particularly discloses an intelligent bellyband for household respiratory rehabilitation and a management system.The intelligent bellyband and management system comprises the following steps that thoracic and abdominal movement data are collected through a flexible sensor array, and the thoracic and abdominal movement data comprise thoracic expansion amplitude, abdominal pressure change and blood oxygen saturation information; constructing an initial multi-mode signal set; according to the initial multi-mode signal set, a recurrent neural network is adopted to extract time sequence features, and dynamic feature representation is obtained and used for follow-up fusion processing; performing priority evaluation on respiratory training quality from the dynamic feature representation, and determining key signal subsets to identify potential deviations; if the key signal subset meets a triggering condition, video auxiliary analysis is carried out; the invention aims to solve the problems of multi-mode signal complexity and inaccurate real-time feedback in respiration data acquisition and analysis.

The invention discloses a composite wing unmanned aerial vehicle intelligent obstacle avoidance method and system based on deep learning, and the method comprises the steps: responding to an obstacle avoidance task signal, collecting multi-modal environment perception data, constructing a lightweight multi-modal fusion neural network model which comprises an encoder and a decoder, and carrying out the recognition of the multi-modal environment perception data through the encoder and the decoder; fusing the image sequence, the three-dimensional point cloud data and the circumferential sensing data through an encoder to obtain a fused feature tensor, decoding the fused feature tensor through a decoder, outputting a risk result based on an obstacle, inputting the risk result based on the obstacle and flight state parameters into a recurrent neural network controller together, and outputting the flight state parameters. And outputting the unmanned aerial vehicle flight control instruction vector, thereby obtaining an unmanned aerial vehicle control signal, controlling the unmanned aerial vehicle to avoid the obstacle, and obtaining an obstacle avoidance task completion result. The invention provides a composite wing unmanned aerial vehicle intelligent obstacle avoidance technology capable of realizing deep fusion of multi-mode perception, end-to-end intelligent decision and platform customization control.

The invention provides a mining motor explosion-proof housing airtightness and internal pressure fault monitoring method, and relates to the technical field of mechanical equipment monitoring, and the method comprises the steps: arranging a piezoelectric sensing array at a sealing key point of an explosion-proof housing, obtaining a vibration signal generated in a motor operation process, and converting the vibration signal into a stress distribution matrix through a piezoelectric effect; and analyzing the stress distribution matrix by using singular value decomposition, and identifying the spatial distribution of the stress abnormal region. Acoustic tracer gas is injected into the shell, acoustic response signals of the piezoelectric sensor in different frequency bands are collected, an acoustic characteristic coefficient group is extracted through wavelet transform, and the mapping relation between the characteristic coefficient group and a leakage path is established. And generating a fault probability density function by using a recurrent neural network in combination with the stress anomaly region and the time sequence change rule of the feature coefficient group, and determining the space coordinates of the leakage point. According to the invention, the real-time evaluation of the airtight state of the explosion-proof housing and the accurate positioning of the leakage point can be realized, and the reliability and accuracy of the positioning result are effectively improved.

The invention discloses a smart home control method based on voice recognition and context awareness. The method comprises the following steps: S1, collecting voice signals and environment state information; s2, performing preprocessing on the voice signal; s3, inputting the speech feature sequence into a speech recognition model based on a recurrent neural network transduction structure for recognition processing, and outputting a semantic text sequence; s4, constructing a two-stage trigger matching strategy, and if a matching result exists, determining that the request is a voice wake-up request; s5, executing situation classification, and outputting a current situation type label; and S6, according to the current situation type label, generating different instructions. According to the invention, by fusing voice recognition and environment perception, high-accuracy smart home control suitable for an aged scene is realized, and the system has timely response and personalized interaction capabilities.

The invention discloses a new energy station high-altitude dangerous operation state monitoring method based on multi-source data, and belongs to the field of new energy station high-altitude operation. The method comprises the following steps: acquiring worker information and wearing data of a safety device, and acquiring operation authority of a worker according to the worker information and the wearing data of the safety device; when the operation authority of the worker meets a set condition, collecting state data of the current worker and current operation environment data, and performing fusion processing to obtain data after fusion processing; performing prediction processing on the data after fusion processing by adopting a recurrent neural network fused with an attention mechanism to obtain an operation danger value of the worker; and judging whether the operation danger value of the worker meets a set condition or not, and if yes, giving an alarm. According to the method, the recurrent neural network state prediction model fused with the attention mechanism is adopted for data fusion processing, comprehensive evaluation of dangerous state monitoring is formed, and the safety of workers is guaranteed in real time.

The invention relates to the technical field of industrial automatic measurement and control communication, and particularly discloses multi-protocol fusion intelligent measurement and control host equipment, which comprises a data acquisition module, a health assessment module, a thermal simulation module and a cooperative control module. The equipment adopts a dual-core hardware architecture with a bottom plate and a top plate, integrates seven communication interfaces, realizes multi-protocol parallel processing through a hardware accelerator and a FreeRTOS system, the protocol switching time does not exceed 10ms, and the health assessment module adopts an online incremental learning recurrent neural network to calculate the health state index of the super capacitor module in real time, so that the super capacitor module can be used as the health state index of the super capacitor module. The thermal simulation module predicts the thermal runaway risk level based on an embedded finite volume method, and the cooperative control module dynamically adjusts the rotating speed of a fan and the switching frequency of a power supply by adopting a Nash equilibrium optimization algorithm, so that early performance attenuation recognition, thermal runaway risk prediction and intelligent cooperative control of the super-capacitor module in an extreme environment are realized. And the reliability and the safety of the system are obviously improved.

The invention relates to the technical field of intelligent question answering, and discloses an enterprise regulation intelligent question answering system and method based on semantic understanding, and the system comprises a data collection layer which is used for obtaining a multi-level complex regulation query statement input by a user; the semantic understanding layer is used for performing word segmentation, dependency syntax analysis and semantic role labeling on a query statement; and the knowledge reasoning layer is used for carrying out multi-step evidence fusion and answer verification through a logical reasoning engine. The method comprises the following steps: S1, acquiring a multi-level complex regulation query statement input by a user; s2, performing word segmentation, dependency syntax analysis and semantic role labeling on the query statement; and S3, analyzing the nested structure statement by using a hierarchical recurrent neural network. By introducing a multi-level semantic understanding mechanism and combining dependency syntactic analysis, semantic role labeling and a level recurrent neural network, deep structure analysis of the composite query statement is achieved, and the real intention which is not shown by a user can be accurately recognized.

A method, computer program product, and computer system of artificial intelligence (AI) intervention to detect and mitigate an abnormality associated with an object moving during performance of a process. A trained recurrent neural network (RNN) determines, from sensor data, a probability (Pr1) that the abnormality existed at a first time, where Pr1 exceeds a threshold T1 and in response, an alternative generative adversarial network (AGAN) determines a probability (Pa1) that the abnormality existed at the first time. A score S1, which is computed as a function of Pr1 and Pa1, exceeds T1 and in response the RNN and the AGAN determines, from the senso data, a probability (Pr2) and a probability (Pa2), respectively, that the abnormality existed at a second time. A score S2, which is computed as a function of at least Pr2, Pa2, and (S1-T1), exceeds a threshold (T2) and in response, the abnormality is mitigated.

A hearing aid selectively prioritises audio data (eg. one nearby speaker of several in a noisy environment) using visual data from a camera and contextual data (eg. lip movements, gaze, noise levels) input to a real time multi-modal data enhancement model comprising a single-headed transformer artificial neural network with self-attention followed by fully connected layers. The model may switch to a lightweight convolutional recurrent neural network.

The present invention discloses a method and device for real-time analysis of student status based on deep learning. The method obtains and preprocesses student image data within the classroom, performs multi-scale face detection and facial feature extraction on the preprocessed image data, extracts expression features based on standardized facial feature data and fuses them with attention indicators and posture features, constructs a time series feature sequence and applies a heavy-tailed recurrent neural network for time series modeling, establishes evaluation criteria, and adjusts evaluation parameters through adaptive thresholds to ultimately obtain student status assessment results. By utilizing a heavy-tailed recurrent neural network and a slow transition mechanism to low-dimensional chaos, the present invention can accurately capture subtle changes and long-term trends in student status, addressing the technical issues of traditional student status monitoring methods, such as poor real-time performance, limited coverage, and insufficient personalization.

A system includes a machine-learning model implemented on a data processing apparatus, which features a parallel processor with a memory hierarchy. The machine-learning model is a recurrent neural network (RNN) with a multi-head architecture, comprising multiple sub- vectors that process parallel data streams. The RNN's weight matrix is structured as a block-diagonal matrix, allowing for parallel processing of the sub-vectors. A fused computational kernel executes an entire time-series processing loop for the multi-head RNN, maintaining the weight matrix blocks in on-chip memory and performing matrix multiplications and element-wise operations for each sub-vector in a single kernel execution.

The present disclosure relates to methods and systems for object tracking, for example for object detection and grid segmentation using recurrent neural networks. A computer implemented method for object tracking comprises the following steps carried out by computer hardware components: providing random values as a hidden state of a trained neural network for an initial time step, wherein the hidden state represents an encoding of sensor data acquired over consecutive time steps in a grid structure, wherein the hidden state further represents an offset indicating a movement of the object between the consecutive time steps; iteratively determining an updated hidden state by processing a present hidden state and present sensor data using the trained neural network; and determining object tracking information based on the updated hidden state.

Two-pass automatic speech recognition (ASR) models can be used to perform streaming on-device ASR to generate a text representation of an utterance captured in audio data. Various implementations include a first-pass portion of the ASR model used to generate streaming candidate recognition(s) of an utterance captured in audio data. For example, the first-pass portion can include a recurrent neural network transformer (RNN-T) decoder. Various implementations include a second-pass portion of the ASR model used to revise the streaming candidate recognition(s) of the utterance and generate a text representation of the utterance. For example, the second-pass portion can include a listen attend spell (LAS) decoder. Various implementations include a shared encoder shared between the RNN-T decoder and the LAS decoder.