1670results about "Biological neural network models" patented technology

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a task scheduling execution method, device, equipment and medium based on intention recognition, and the method comprises the steps: fusing an interaction request and a dialogue history library to generate a session context, and carrying out the analysis to obtain an intention recognition result; task planning is carried out based on the intention recognition result and a knowledge base, and a task subitem set and an execution sequence are generated; matching the task subitem set with a resource directory to generate an execution resource list; scheduling the task subitem set by using the execution sequence to generate an ordered task sequence; generating an execution plan in combination with the ordered task sequence and the execution resource list; and triggering execution based on the execution plan, and aggregating output to generate an execution result and a process tracking record. According to the method, automatic connection of session analysis, task planning, resource matching, sequential scheduling and execution triggering is achieved, manual operation splitting among multiple systems is eliminated, and task processing efficiency and execution accuracy are improved.

Embodiments of the present disclosure relate to multi-view LIDAR perception with motion cues for autonomous and semi-autonomous machines and applications. A DNN may be used to detect objects, a navigable space, weather or surface conditions, artifacts, and / or other parts or features of an environment based on multiple views of LIDAR data from multiple time slices. The DNN may include multiple input channels for processing multiple views of sensor data from multiple time slices to provide motion cues, and the extracted features from the different time slices may be geometrically projected from a first 2D view to a second 2D view, combined with features that were extracted from the second 2D view, and applied to a subsequent stage of the DNN. The data generated by the DNN may be provided to the drive stack of an autonomous vehicle or other ego-machine to enable safe planning and control of the vehicle.

The invention discloses a full-closed-loop automatic spraying method, system and device based on geometrical characteristic driving and multi-physical quantity feedback and a storage medium. The method comprises the following steps: performing three-dimensional scanning on a workpiece and extracting geometric features; generating a spraying track and a spraying posture for keeping a topological structure based on the geometrical characteristics; motion, process and coating quality parameters are collected in real time through a sensor, and are processed and fused to form a state vector; inputting the state vector into an intelligent control module, performing optimization by combining model predictive control, fuzzy logic and a neural network, and outputting trajectory and process parameters; a control strategy is updated according to the quality deviation of online detection, and closed-loop correction is carried out; spraying is automatically executed in the closed spraying cabin, and unmanned operation of the whole process is achieved. According to the method, complex curved surface spraying can be completed without manual programming, and the method has the advantages of high flexibility, multi-parameter closed loop, self-adaptive optimization and the like.

The present invention relates to a high temperature and drought composite disaster monitoring and early-warning method and system, belonging to the technical fields of disaster risk assessment and early warning. Internal correlation features and abnormality information of high temperature and drought events are input into a model, multi-time-space scale features of the high temperature and drought events can be identified accurately, high event identification accuracy and space resolution are achieved, the progress can be predicted progressively, the drought and high temperature threshold change can be monitored closely, and fine forecasting and early warning can be performed in different periods, regions and intensities, thereby ensuring that indicators are in the same time scale, avoiding the complication of the high temperature and drought process caused by frequent time and space discontinuities of the indicators in a single point or small region, and ensuring the suitability for any periods of the process.

A method for processing a piece of textual information. The piece of textual information is parsed into a set of plaintext input tokens. Each of the plaintext input tokens is individually transformed using a first binary data transformation, to achieve a set of binary input tokens. Each of the set of binary input tokens is transformed individually or collectively, using an embedding data transformation, into one or several vectorized input tokens. The one or several vectorized input tokens is / are fed to a first neural network. A response is received from the first neural network in the form of one or several vectorized output tokens.

The invention provides a virtual sample generation method and system based on a non-stationary neural network Gaussian process, and the method comprises the steps: obtaining original data, judging the non-stationarity, and judging whether a statistical characteristic changes with an input position or not; if not, a hidden variable model fusing the neural network and the Gaussian process is constructed, and a hidden variable space representing non-stationary distribution is obtained through training; sampling based on hidden variable space probability distribution; using the model to generate = (, Z) (Z, Z)-1X through non-stationary high-dimensional mapping, which is a non-stationary kernel function, Z is a low-dimensional hidden variable, and X is an input variable; checking the consistency of the high-dimensional virtual samples and the original data, and screening a virtual sample set meeting statistical consistency; according to the method, high-quality and high-diversity virtual samples can be generated in small sample and non-stationary scenes.

The invention relates to a high-speed rail station building ventilation control method and system based on a physical information neural network, and the method specifically comprises the following steps: collecting related parameters of a high-speed rail station building, and carrying out the preprocessing; based on an electric field intensity theory, constructing a passenger space-time distribution function according to the preprocessed data to calculate dynamic heat source intensity; constructing a PINN model of 10 layers of MLP, and constructing a loss function according to physical constraints to train the PINN model; according to output of the PINN model, a control target is deduced in combination with ventilation parameters, and then ventilation of the high-speed rail station building is controlled; and comparing a derivation result with an actual condition, if the deviation is too large, performing fine adjustment on the PINN model, and correcting the output derivation result until the deviation is within a specified range. According to the method, related data are collected and converted into dynamic heat source items which can be identified by the PINN model, and the model is trained through physical constraints, so that the accuracy of a derivation result can be improved, and the ventilation control precision and the energy-saving effect of the high-speed rail station building are improved.

The invention provides a CNN-PINN-based gas turbine combustion chamber performance prediction method, belongs to the technical field of combustion chamber combustion, and aims to solve the problems of high resource consumption and high grid dependence of the conventional CFD calculation at present, and the method comprises the steps: S1, collecting historical operation data of a combustion chamber, generating simulation data through numerical simulation, and carrying out the calculation of the simulation data; performing comparative analysis on the simulation data and the collected experimental operation data to prepare a data set; s2, preprocessing the data set; s3, establishing a CNN-PINN neural network prediction model according to the data characteristics and the target; s4, establishing a loss function containing a data error term and a physical information error term; s5, training and evaluating the established model by using the preprocessed data; and S6, performing performance prediction on target data by using the trained model, and performing reverse normalization on a prediction result to obtain prediction data corresponding to multiple targets.

The application relates to the technical field of motor life prediction, and provides a stepping motor life prediction method and system based on multi-physical field coupling, electrical parameters of a stepping motor, multi-physical field data in a running state and a multi-dimensional feature vector corresponding to the multi-physical field data are acquired, a physical damage index corresponding to the stepping motor is determined according to the electrical parameters, the multi-physical field data and a preset physical damage index calculation formula, the multi-dimensional feature vector is input into a trained data-driven model, a data-driven damage index corresponding to the stepping motor is acquired, and a life prediction result of the stepping motor is acquired according to the physical damage index, the data-driven damage index and a preset life prediction formula. In this way, the influence of the multi-physical field data on the life of the stepping motor can be comprehensively considered, the limitation of a single mode in predicting the life of the stepping motor is overcome by fusing the physical damage index and the data-driven damage index, and the accuracy of the life prediction result of the stepping motor is improved.

The invention relates to the technical field of industrial compressed air system intelligent control, in particular to a digital air compression station intelligent control method and system based on AI self-learning, and the system comprises a data collection layer, an AI prediction layer, an optimization control layer and an execution layer. The data acquisition layer realizes accurate acquisition and preprocessing of multi-dimensional data through a three-stage high-precision sensor network and an edge computing node; the AI prediction layer adopts an XGBoost and LSTM mixed model to be combined with an online learning mechanism to realize accurate prediction of the gas consumption demand and the energy efficiency ratio within 1-24 hours; the optimization control layer completes air compressor operation combination optimization and pressure dynamic control based on an improved NSGA-II algorithm and self-adaptive PID adjustment; the execution layer realizes control instruction landing through frequency converter and cluster cooperative scheduling. Meanwhile, an equipment health degree monitoring module, a multi-source data fusion correction module and a waste heat-gas consumption collaborative optimization module are innovatively introduced, and the problems that a traditional air compression station is low in energy efficiency, poor in self-adaption, insufficient in prediction precision and lack of equipment management are solved.

The present invention relates to a device (6) for associate physiological signals from a user (51), said physiological signals being multimodal physiological signals, with commands of a brain-computer interface - BCI - (71) using a trained user-specific machine learning system, and a device for training said a trained user-specific machine learning system. Specifically, the invention features a hardware-agnostic, multimodal BCI powered by generative artificial intelligence, cognitive AI agents, and Riemannian geometry, with reinforcement learning techniques aimed at making the BCI adaptive to each user's cognitive and affective states, and physicality, by translating said physiological and neurophysiological signals into passive and active (mental) commands of connected devices and digital environments.

The invention discloses an oil and gas long-distance pipeline automatic control system fault identification method based on deep learning, and the method comprises the following steps: collecting and preprocessing a time sequence, and generating a standardized input sequence; constructing a physical constraint feature set, and generating a physical observable representation; encoding the valve opening instruction time sequence and the pump station frequency instruction time sequence to form a combined input sequence; obtaining Koopman state representation, and generating a linear evolution state sequence; a residual nonlinear compensation structure is introduced to generate residual features, and a comprehensive state sequence is formed; time scale splitting is carried out, and fast, medium-speed and slow time scale feature sequences are decomposed; identifying anomalies and generating a time scale anomaly feature set; leakage energy loss indexes are calculated and fused to form fault mode drift representation; judging the pipeline state, and outputting a fault recognition result. According to the invention, a deep Koopman operator structure is adopted, accurate identification of pipeline faults is realized, and the method has the advantage of high diagnosis reliability.

The present invention relates to a privacy-preserving EV charging authorization and billing system, and a method for the same. The proposed system is configured to integrate permissioned blockchain with fully homomorphic encryption. The present invention aims to eliminate plaintext exposure mitigates single point of failure, by performing all authorization and billing computations on encrypted data and recoding transactions immutably, wherein an EV user securely generates encrypted authorization and billing requests using FHE-based public keys. The charging station routes these encrypted requests to the blockchain network, which records immutable encrypted transactions and verifies them via consensus. The FHE computation layer performs secure operations on the encrypted data for authorization and billing, while smart contracts execute automated verification and billing computations, ensuring transparency and auditability.

The invention relates to the technical field of artificial intelligence, and provides an intelligent insurance matching method and system combining a user portrait and a clause knowledge base. The method comprises the following steps: constructing a first portrait of a user based on a user data set and optimizing to generate a second portrait; extracting features from the insurance clause knowledge base in a multi-dimensional manner to construct a product portrait; performing demand coverage risk evaluation on the product portrait based on the second portrait of the user to generate a demand coverage risk matrix; performing guarantee conflict risk evaluation to generate a guarantee conflict risk matrix; introducing an underwriting risk analysis channel, and generating an underwriting risk matrix in combination with the second portrait of the user; and fusing the three types of risk matrixes to perform multi-level optimization matching, and generating an insurance optimization matching graph, so as to solve the technical problem of significant deviation between the recommendation result and the actual risk preference and guarantee demand of the user, realize dynamic and accurate matching of the insurance product and the user demand, reduce the guarantee conflict risk, improve the correlation of the recommendation result, and improve the user experience. And the technical effect of personalized insurance configuration requirements is met.

The invention relates to the technical field of electromagnetic metasurfaces and bionic computing, and discloses a bionic information processing method and system based on an electromagnetic metasurface. The method comprises the following steps: S1, mapping DNA sequence information into a multi-dimensional coding state of the bionic metasurface information processing unit in an electromagnetic adjustable parameter space; s2, realizing co-evolution of the multi-dimensional coding state in space, frequency, phase and amplitude dimensions by dynamically regulating and controlling electromagnetic response parameters of the bionic metasurface information processing unit; and S3, using propagation and interference of electromagnetic waves in the diffraction neural network and a near-field coupling effect between the bionic metasurface information processing units to complete copying, logical operation and information read-write operation of the DNA sequence information. The functions of DNA coding, copying, logic calculation, information reading and writing and the like are uniformly mapped into the programmable electromagnetic metasurface platform, and cooperative processing of information in the aspects of space, electromagnetism and algorithm is achieved.

The invention belongs to the technical field of communication, and particularly relates to a multipath channel DOA estimation method based on a heterogeneous attention double-branch neural network, and the method comprises the steps: constructing a training set, a verification set and a test set through stratified sampling; introducing a multi-head attention mechanism to construct a heterogeneous attention double-branch neural network for realizing intelligent DOA estimation; designing a frequency weighted loss function as a loss function of neural network model training; a dynamic attention mechanism and an early stop mechanism are designed in a model training process to prevent an overfitting phenomenon. According to the method, the problem of serious DOA estimation model overfitting in a small sample and data non-uniform real acquisition data scene is effectively relieved, and the DOA estimation accuracy is improved.

The invention discloses a river and lake water bloom prediction method and system based on an integrated diffusion learning model, and the method comprises the steps: carrying out the preprocessing of an original water quality time sequence according to the water quality monitoring data of a target river and lake region, and constructing a standardized water quality time sequence input data set; disturbing the data set based on a conditional diffusion generation model, and constructing a plurality of initial condition diversified input disturbance sets; a plurality of input disturbances of the disturbance set are sent into a deep neural network prediction model for parallel prediction, so that a plurality of algal bloom prediction orbits are formed to jointly form an integrated prediction result under disturbance driving; and statistical analysis and fusion processing are carried out to form prediction result distribution with uncertainty quantification capability, and visual display is carried out. According to the method, the uncertainty of the prediction result can be quantitatively described while the prediction precision is kept, and a more reliable decision basis can still be provided for water environment scheduling, ecological early warning and emergency response especially under extreme hydrological conditions such as flood and drought.

The invention relates to the technical field of data processing, in particular to a multi-source data fusion method based on a railway station building and a city, which comprises the following steps: acquiring an IFC format data stream and a CityGML data stream; performing space-time correlation integration on the two types of data streams to generate a data set with engineering constraints; based on the geometric topological relation and the construction stage characteristic parameters, an improved ant colony optimization algorithm is adopted for coordinate conversion; constructing a railway component and city element mapping rule base and carrying out semantic mapping; performing multi-working-condition collision detection by adopting a parallel universe simulation algorithm based on the construction event identifier; carrying out collaborative decision analysis by adopting a hierarchical response mechanism based on conflict types and severity; a verification environment variable set is injected into a preset digital twin platform, and verification parameters are dynamically adjusted by adopting a particle swarm optimization algorithm. According to the method, the problem of multi-source data collaborative fusion of the railway station building and the city is solved, and the accuracy of data processing and the robustness of the system are improved.

A computer-implemented system for improving radiation dose reconstruction in adaptive real-time radiotherapy, comprising a data acquisition module for capturing treatment data in real time during radiotherapy, a preprocessing module for preprocessing the captured data, a deep-learning dose reconstruction engine for reconstructing the delivered radiation dose distribution using a trained deep-learning model, a dose evaluation and comparison module for comparing the reconstructed dose distribution with a planned dose distribution, and an adaptive decision support module to assist in adjusting radiotherapy treatment parameters based on the comparison results.

Systems, methods, and apparatuses for providing dynamic, prioritized spectrum utilization management. The system includes at least one monitoring sensor, at least one data analysis engine, at least one application, a semantic engine, a programmable rules and policy editor, a tip and cue server, and / or a control panel. The tip and cue server is operable utilize the environmental awareness from the data processed by the at least one data analysis engine in combination with additional information to create actionable data.

Described is a system for feeding pet avatars by accessing a particular geographic location for a first user; accessing map data for the particular geographic location; identifying that a pet avatar for a second user is associated with the particular geographic location; causing display, on a first computing device of the first user, of: the particular geographic location using the map data; the pet avatar; and an interface element configured to feed the pet avatar; receiving a user selection of the interface element from the first user; causing display, on the first computing device for the first user, of an indication of the pet avatar being fed corresponding to the interface element; and transmitting a notification to a second computing device of the second user including an indication that the pet avatar is being fed corresponding to the interface element.

There is provided a system and method of runtime defect detection in a semiconductor specimen. The method includes obtaining a plurality of runtime images acquired for a plurality of dies on the specimen, feeding the plurality of runtime images to a plurality of input channels of a neural network (NN) in an input order, wherein the NN is previously trained in a training phase, and processing, by the NN, the plurality of runtime images simultaneously, to obtain a plurality of defect maps, each corresponding to a respective runtime image and indicating probabilities of defect candidate presence thereof. Each given runtime image is processed as a target image using remaining images in the plurality of runtime images as reference images of the target image, and the defect map of the target image remains invariant, irrespective of changes to the input order.

The invention discloses a large language model training-oriented FP8 quantization noise compensation method and system, and relates to the technical field of computer large language model data noise compensation. The method comprises the following steps that: a processor quantizes original data input into a memory according to a designed FP8 format to obtain quantized data representation; and obtaining quantization noise according to the original data and quantization data representation stored in the memory, and constructing a quantization noise training model by taking the quantization noise as a random variable. And designing a noise compensation mechanism by taking the parameters of the quantization noise training model and the compensation model stored in the memory as compensation units. And the processor determines an embedding point of the noise compensation mechanism embedded large language model in a compensation unit training process according to the compensation granularity demand, so that the original data is input into the large language model embedded with the noise compensation mechanism through the processor for iterative learning, and FP8 compensation data is output. By adopting the method, the hardware calculation speed can be increased and the memory access frequency can be reduced in the big language model data processing process.

The invention discloses a mental stress assessment and intelligent dredging method and system, and belongs to the technical field of digital medical treatment and health informatics. Multi-modal physiological signals are continuously collected through the wearable device, medical diagnosis level pressure state evaluation is carried out based on the personalized physiological baseline, and personalized pressure indexes and levels are generated; secondly, when it is diagnosed that the pressure level exceeds the standard, the system serves as an intelligent decision support system, the environment and schedule information after privacy protection processing is fused, and an optimal grooming action is dynamically selected from a predefined intervention action library and executed; finally, the system serves as a continuous learning system, the intervention efficiency is evaluated in real time according to feedback data of the pressure index after execution, the decision model is updated, and collaborative self-optimization of the diagnosis strategy and the intervention strategy is achieved. According to the method, medical diagnosis, personalized treatment decision and adaptive learning are integrated, and the systematicness, accuracy and intelligent level of mental stress related health problem management are improved.

The invention discloses a BMS (Battery Management System) fault prediction method and system in the technical field of battery management system testing, which are used for solving the technical problem that random crash is difficult to position and reproduce in BMS operation. According to the method, the electrical, thermodynamic and RTOS task scheduling operation states of a BMS are synchronized in real time by constructing a multi-physics field coupling digital twinborn model, and feature extraction and fault diagnosis are performed based on simulation process parameters. And carrying out accelerated reproduction on the fault in the multi-physics field coupling digital twinborn model, recording a key parameter combination causing the BMS crash, carrying out verification on an actual BMS hardware platform, and positioning a crash reason and a critical condition. According to the method, real-time data acquisition, digital twin simulation and test verification are combined, and full-process analysis from data acquisition to fault reproduction can be completely realized.

Methods and systems are disclosed for predictive classification of a raw mass spectrum of a sample, the spectrum resulting from mass spectrometry analysis. The predictive classification is performed by a trained large-scale spectral model, including a self-supervised large-scale base spectral model trained with unidentified sample spectra from a plurality of sources and fine-tuned with a task-specific model trained with identified spectra.

A method of facilitating a skill learning process or improving performance of a task, comprising: determining a brainwave pattern reflecting neuronal activity of a skilled subject while engaged in a respective skill or task; processing the determined brainwave pattern with at least one automated processor; and subjecting a subject training in the respective skill or task to brain entrainment by a stimulus selected from the group consisting of one or more of a sensory excitation, a peripheral excitation, a transcranial excitation, and a deep brain stimulation, dependent on the processed temporal pattern extracted from brainwaves reflecting neuronal activity of the skilled subject.

Systems and methods for integrating generative artificial intelligence (AI) capabilities within Software as a Service (SaaS) platforms. One of the computer-implemented methods are for querying a generative AI model about structured data in a SaaS environment, enabling users to interact with and manipulate data through AI-assisted interfaces. One of the systems maintains a generative AI agent configured to interact with SaaS platform data as a virtual team member, capable of understanding context and nuances of project data. Also described are methods for color-context aware data analysis, generation of interactive elements in messaging sessions, and cross-application generative AI agent interactions triggered by user mentions. Also described is facilitating the creation of custom SaaS platform products by combining functionalities from existing products using generative AI. The systems and methods represent advancement in AI-driven SaaS customization and data analysis.

The invention belongs to the field of dam flood discharge safety, and provides a dam flood discharge safety protection method and system based on earthquake response analysis, and the method comprises the steps: S1, obtaining multi-source environment and structure data in real time; s2, preprocessing the multi-source environment and structure data to form a unified time sequence multi-dimensional data set; s3, simulating and analyzing the structural dynamic response of the dam and the flood discharge gate based on the unified time sequence multi-dimensional data set, and evaluating potential damage; s4, comprehensive risk assessment is carried out by combining the earthquake response analysis and damage assessment result, the hydro meteorological data and the flood discharge gate operation state data, and early warning information is generated; s5, generating a dynamically adaptive flood discharge gate opening and closing strategy according to the comprehensive risk assessment result and the early warning information; s6, executing the opening and closing strategy of the flood discharge gate, and monitoring the actual operation state of the flood discharge gate in real time; and carrying out comparative analysis on the actual running state of the flood discharge gate and the prediction strategy, and carrying out updating and optimization according to an analysis result.

The invention discloses a power plant unit early warning method based on a cloud platform, and relates to the technical field of power plant unit early warning, and the method comprises the steps: collecting high-frequency dynamic data and low-frequency steady-state data of a power plant unit; performing spectral analysis on the high-frequency dynamic data, extracting high-frequency sub-signal features, extracting low-frequency trend features from the low-frequency steady-state data, and performing dimensionality reduction through principal component analysis to form a multi-level feature set; the method comprises the following steps: extracting a multi-level feature set before normal operation and fault from a historical database, labeling according to a fault type, generating an association rule and constructing a rule base, dynamically updating a rule weight based on rule prediction accuracy, generating a mode class by adopting an ART neural network, and updating a class center or creating a new mode class after real-time feature clustering; the rules in the rule base are divided into three levels according to the confidence coefficient and the fault emergency degree, the matching degree and the early warning index of the real-time features and the rules are calculated, the hierarchical response is triggered according to the early warning index, and the accuracy, the real-time performance and the comprehensiveness of fault early warning are remarkably improved.