42 results about "Cognitive systems" patented technology

The invention discloses an AI agent design platform decision optimization method based on big data, and particularly relates to the field of artificial intelligence, comprising multi-modal data sensing layer construction, a streaming feature calculation engine, a dynamic index fusion center and an adaptive decision matrix. According to the method, accurate synchronous monitoring of the utilization rate of hardware resources and dynamic collaborative optimization of heterogeneous computing units are achieved, and the resource scheduling efficiency in a complex computing scene is remarkably improved; knowledge system degradation caused by long-term learning is effectively prevented, and the continuous reliability of a cognitive system is ensured. The provided multi-dimensional decision credibility verification system is fused with interpretability penetration analysis, environment coupling modeling and logic drift detection technologies, the limitation of a traditional single credibility index is broken through, the risk prediction and fault-tolerant capability of the decision process is remarkably enhanced, and a full-dimensional safety decision guarantee system is constructed for an intelligent agent.

An upper-body sensorized exoskeleton is configured to interface with external systems, such as FES, advanced cognitive systems, and VR / AR interaction programs, in order to facilitate rehabilitation and assistance of patients affected by upper-limb impairments. The sensorized upper limb exoskeleton delivers anti-gravitational support at the shoulder level, enables upper-limb configuration limits for shoulder and elbow ranges of motion, and reads kinematic data that may be analyzed by users and clinicians. The sensorized upper limb exoskeleton is equipped with electronic processing and communication means that support bidirectional communication between exoskeleton and external systems.

The invention particularly relates to a multi-level self-cognition system based on a large model, and relates to the technical field of large models. A neural symbol world model module; a large language model cognition core module; and a hierarchical decision planning system module. According to the method, deep integration of perception, cognition and decision making is achieved through the hierarchical fusion architecture, and compared with the prior art, the method has remarkable advantages; the multi-modal perception encoder adopts layered encoding and a cross-modal attention mechanism, so that the semantic alignment problem of multi-source perception data is effectively solved, and the understanding ability of the system to a complex scene is greatly improved; according to the neural symbol world model, the neural network and symbol reasoning are combined, the limitation of a pure neural network method in physical modeling is overcome, meanwhile, the calculation complexity of a pure symbol system is avoided, and efficient and accurate environment characterization and prediction are achieved.

The invention provides an industrial cognitive base system based on multi-modal comparative learning and an execution method, and the industrial cognitive base system based on multi-modal comparative learning is characterized in that a system architecture comprising a knowledge graph management module and a cross-modal coding fusion module is constructed; industrial field knowledge is dynamically injected into a multi-modal feature learning process in the form of a structured sub-graph, and a comparative learning mechanism introducing industrial semantic constraints is combined, so that a model obtained by final training can realize semantic alignment of multi-modal data; it can be ensured that generated unified semantic representation and reasoning results strictly conform to predefined industrial logic rules, and therefore the defect that in the prior art, a pure data driving method may generate results violating industrial common knowledge is effectively overcome; and the output reliability, the decision credibility and the practical application value of the industrial cognitive system in key tasks such as fault diagnosis and state monitoring are remarkably improved.

The invention provides a maze robot brain-like cognitive system based on an emotion and memory mechanism, and the system comprises a perception and emotion generation module which obtains the environment state of a robot, and calculates the reflective fear value and cognitive fear value of the robot; the memory module is used for recording short-time memory, working memory and long-time memory of the robot in combination with an anti-thought memory mechanism; the behavior decision-making module is used for comparing the evaluation function based on the cognitive fear value with the energy margin to obtain a behavior decision-making result; the learning module is used for exploring and learning a target node and an energy node to obtain an updated target guide table and an updated energy guide table; and the action selection module is used for updating the target guide table and the energy guide table according to the behavior decision result, and outputting the next action. According to the invention, the survival rate of the robot in the environment learning process can be improved, and long-time learning is realized; and the robot can be prevented from repeatedly exploring in an invalid area, the learning efficiency is improved, the energy consumption is reduced, and self-survival is ensured.

A device for decision support of a cognitive system based on data originating from a plurality of data sources, the device including a processing unit associated with each data source, each processing unit including an encoding unit configured to determine, from the data from the data source associated with the processing unit, a representation of data in a common representation space by applying a machine learning algorithm to the data, the device further including a data fusion unit configured to determine a representation model of an environment of the cognitive system by combining the data representations determined by the encoding units associated with the plurality of data sources through the application of a data fusion algorithm.

The invention discloses a teaching expansion board, relates to the technical field of electronic information, and aims to solve the technical problems of low teaching scene switching efficiency, waste of hardware resources, difficulty in constructing a complete engineering cognitive system by students, and insufficient system-level design and innovation ability cultivation caused by physical solidification of functional modules and one-board-one-purpose of a traditional teaching expansion board. Comprising an expansion board body. The expansion board main body is connected with a main control chip integrated with a microprocessor, and the expansion board main body is also provided with an actuator module used for realizing common functions, a sensor module used for carrying out detection type teaching demonstration, an interaction module used for carrying out interaction operation and an expansion interface used for carrying out function expansion. According to the invention, through modular replaceable architecture design, rapid module replacement and experiment scene reconstruction are realized, the problem of low scene switching efficiency caused by a traditional'one-board-one-use 'mode is solved, and a single expansion board can cover multiple teaching scenes.

An adaptive interference method based on target characteristics, according to the real-time and changeable working mode characteristics of the cognitive system, establishes a dynamic knowledge base based on individual target characteristics, and establishes a feasible interference effect evaluation method based on individual characteristics, and combines the radiation source characteristics of the target to form a targeted interference strategy and generate a corresponding interference signal, thereby forming an adaptive interference closed loop from reconnaissance to interference and then to evaluation, and continuously optimizes the interference strategy according to the interference effect evaluation, thereby forming an efficient interference suppression method; at the same time, based on the radiation source characteristics of the target individual, accurate tracking interference can be accurately implemented to effectively cope with the changeable working mode of the cognitive system, with the characteristics of strong operability and strong target targeting.

The application discloses a kind of memory hub methods and systems for multi-agent cognitive decision-making, comprising: obtaining the perception information, behavior result or decision context generated in the running process of cognitive system;The perception information is structured modeling, and memory unit with time attribute, access attribute and evolution attribute is generated;The memory unit is stored and managed in life cycle;Before or in the process of cognitive decision generation, according to the current decision context, select relevant historical experience from the memory unit to participate in decision formation;Based on the use effect of memory unit in the decision process, update the hierarchical state or call strategy of the memory unit, and construct the continuously evolving cognitive decision process based on historical experience feedback. Unified management and monitoring can be carried out on memory access behavior, permission configuration, life cycle state and abnormal conditions, and parameter configuration and state viewing can be supported during system operation.

The invention discloses a space-ground integrated electromagnetic spectrum hierarchical semantic situation multi-granularity cognitive system and method, and the system employs a cloud-edge-end hierarchical architecture, and comprises a space-based network, an air-based network and a ground-based network from top to bottom. Cloud computing nodes in the space-based network, edge computing nodes in the air-based network and terminal nodes in the ground-based network are subjected to layered molecular network processing; the method comprises the following steps: establishing a cloud-edge-end layered spectrum semantic situation multi-granularity cognitive mechanism, and collaboratively cognizing a spectrum semantic situation by cognitive granularity gradually coarsened from end to cloud; aiming at heterogeneity of a space-ground integrated network, nodes in air-space-ground network domains are subjected to hierarchical molecular networking based on space breadth, node density and frequency utilization characteristics, and a frequency spectrum hierarchical semantic situation multi-granularity cognitive mechanism in each network domain among the air-space-ground network domains is established. According to the method, the problems of spectrum sensing data explosion and resource shortage of computing nodes and communication links under a traditional spectrum situation sensing architecture are effectively solved.

The invention relates to the technical field of cognitive system application, and particularly discloses an ACT-R-based personnel cognitive behavior modeling method and a cognitive load calculation method.The cognitive load calculation method comprises the steps that on the basis of the relation between module activation time and weight in an ACT-R cognitive system, the instantaneous cognitive load in the task execution process is calculated; during model operation, according to the task process and the activation time of the knowledge base calling module, the instantaneous cognitive load of each module is dynamically calculated; integrating the instantaneous cognitive load in the task execution process and the instantaneous cognitive load of each module to obtain the personnel cognitive load; a weight distribution and error correction mechanism is used to realize personnel cognitive load quantitative evaluation of different task complexity and personnel states; the cognitive load calculation formula embedded in the cognitive model can calculate the cognitive load generated when each module is called, and reference is further provided for personnel cognitive behavior analysis and work performance improvement.

The invention discloses a method for modeling a brain cognitive system based on a deep neural network, belonging to the fields of artificial intelligence and neurobiology. Aiming at the deficiencies of the background and existing research results, this paper proposes a method for modeling a cognitive system based on a deep neural network, aiming to solve the problems that the existing technology models do not conform to the operation mechanism of the biological brain cognitive system, the modeling does not consider the high-dimensional information in the cognitive system, how to train the model using the temporal data set of cognitive tasks containing image information, and there is still a large gap between the model representation inferred by the existing technology and the biological brain representation.

The application discloses a Token marking method and a cognitive system for large language model input data, and belongs to the technical field of artificial intelligence. The application generates corresponding discrete coordinates for each Token input into the large language model, including: a time coordinate for providing a unified time reference; a context coordinate for identifying a dialogue time sequence; a logic coordinate for identifying the sequential position of the Token in a text or instruction, and the logic coordinate is independently numbered within the range defined by the time coordinate or the context coordinate; and a device coordinate for identifying an Internet of Things device, a role or a priority. The coordinates jointly constitute a unified discrete space-time cognitive framework, and provide a structured worldview for the large language model, so that the large language model can perform structured understanding based on multi-dimensional information such as time, context, logical sequence and identity. The application does not change the core architecture of the model, has strong compatibility, and can be widely applied to the fields of multi-round dialogue, intelligent customer service, industrial automation, smart home, Internet of Vehicles and the like.

The invention discloses a role playing model training data generation method and device and a related medium, and the method comprises the steps: obtaining a demand instruction, defining a task type, and constructing a task system through the task type; using the task system to generate simulated dialogue data, and generating an environment information system according to the demand instruction; performing data association on the environment information system and the simulated dialogue data, and performing cognitive adjustment processing based on a pre-constructed cognitive system to obtain adjusted simulated dialogue data; and calling a pre-constructed evaluation system to carry out quality evaluation on the adjusted simulated dialogue data, and generating final role playing model training data. According to the invention, the environment information system is associated with the simulated dialogue data, the associated simulated dialogue data is subjected to cognitive adjustment processing based on the cognitive system, and finally, the quality evaluation is performed, so that the final role playing model training data is output, and therefore, the quality of the training data is improved, and the training efficiency is improved. And the training requirement of a role playing model can be met.

The invention discloses a tensegrity structure autonomous state cognition system in a difficult-to-reach area, the system comprises a state reconstruction module and a man-machine interaction module, the state reconstruction module detects resistance change data of a cable in a tensegrity structure in real time based on a flexible sensor, predicts the real-time state of the structure by using a long-short-term memory network, and obtains the real-time state of the structure; reconstructing a three-dimensional shape and a mechanical behavior of the structure; the man-machine interaction module carries out word segmentation on the text and converts the text into a vector space with context perception so as to capture a semantic interdependence relationship, and a fine tuning data set is established through an aggregation sensor and image features; a pre-trained LLM model is obtained by combining parallel input of text prompt and time sequence coding, and training is carried out based on a fine tuning data set; in the reasoning process, the model uses reconstruction errors as abnormal scores for calculation, the anomalies are distinguished according to statistical distribution of the scores and a frequency threshold value of a specific field, and the anomalies are divided into sensor interference, sensor damage and rod deformation.

The application discloses an AI intelligent child camera system. The AI intelligent child camera system comprises a cognitive understanding engine, an association memory engine, a creation generation engine and a learning evolution engine. The cognitive understanding engine is used for performing preliminary identification and deep identification on a first target photo by using a dynamically switched end-cloud collaborative cognitive system; the association memory engine is used for associatively storing a new photo with an existing event; the creation generation engine is used for acquiring a child voice instruction and a corresponding second target photo, generating an animation short film corresponding to the child voice instruction from the second target photo, and generating an artistic image from a third target photo based on a decomposable flow matching; and the learning evolution engine is used for adjusting an AI model in the AI intelligent child camera system on the camera by using an orthogonal fine-tuning technology. An adaptive intelligent image interaction system capable of deeply understanding an individual world, intelligently reconstructing image narratives, actively cooperating with content creation, and growing together with children can be constructed.

The application provides a social emotional quotient quantitative criterion system and an AGI safety alignment method based on a four-dimensional discrete space-time cognitive system, which is a unified bottom layer base of general artificial intelligence (AGI) and is suitable for large language models, brain-like cognition, autonomous intelligent agents, brain-computer interfaces and all things interconnection systems. The application allocates a unique four-dimensional coordinate (T, C, L, D) to each Token / neural information element in the model, and constructs a full-link structured cognitive framework that is traceable, addressable, calibratable and constrained. The system adopts a double-bottom parallel architecture: bottom layer A: pure mathematical causal reasoning kernel, reasoning is completely free and cannot be intervened; bottom layer B: quantitative safety criterion bottom layer, only scoring, no reasoning, no pollution, no modification. The application realizes AGI safety alignment, anti-disguise, anti-evasion and anti-autonomous expansion from the root of the architecture by three-domain hard isolation (silicon-based endogenous domain, carbon-based interaction domain and external game domain), double-layer verification (output text + reasoning chain), execution gating (only lock actions, not thoughts), brain-computer interface default ban and cancellation of AI refusal right, and is a necessary bottom layer system of the next generation of strong cognitive AI.

The present invention discloses a method for optimizing the maximum sum rate resources of an IRS-assisted cognitive SWIPT system. With maximizing the end-to-end sum rate of the IRS-assisted cognitive SWIPT system as the optimization goal, the cognitive SWIPT receiver adopts nonlinear energy collection. The multi-resource allocation problem is modeled as a nonlinear non-convex optimization problem, which is mutually coupled with the subcarrier, power, and IRS reflection phase shift vector. Therefore, an alternating optimization method is adopted to fix the IRS phase shift vector, and the Lagrange dual transformation and subgradient method are used to obtain the optimal power and subcarrier set for information decoding and nonlinear energy collection. Then, the IRS reflection phase shift vector is obtained by a continuous convex approximation method, thereby maximizing the end-to-end sum rate of the IRS-assisted cognitive SWIPT system. The present invention effectively allocates multiple resources such as subcarriers, power, and IRS reflection phase shift vectors for information decoding and nonlinear energy collection in the cognitive SWIPT system, and maximizes the end-to-end sum rate under the actual SWIPT nonlinear energy collection model.

This invention relates to the fields of basic research in cognitive neuroscience, large-scale artificial intelligence model architecture, brain-computer interfaces, and educational cognitive assessment. Specifically, it relates to a unified four-dimensional discrete-time spatiotemporal modeling method for basic neural information elements of the brain (neural impulses, neuronal cluster firing, and sensory neural signals). This method is based on a homologous and unified underlying theoretical framework and is an engineering embodiment of a unified four-dimensional discrete-time spatiotemporal cognitive system. It shares a completely homologous, isomorphic, and unified mathematical framework with the inventor's previously submitted "A Token Tagging Method and Cognitive System for Input Data of a Large Language Model." The only difference is that the AI ​​patent processes computational information elements (Tokens), while this patent processes biological information elements (neural impulses, sensory signals, and neuronal firing events). Together, they constitute a unified underlying architecture covering biological cognition and artificial intelligence.

The invention provides a model construction method and system for a communication interference cognitive system. The method comprises the following steps: constructing a communication interference cognitive reference set containing interference scene association features and interference response reference features; then, interference feature correlation modeling is carried out based on the communication interference cognitive reference set to form a correlation map; implementing cognitive model structure adaptation according to the correlation map to generate an initial interference cognitive model; carrying out interference scene adaptive iterative optimization on the initial model to form an optimized target interference cognitive model; and finally, outputting the target model for interference identification and response processing of the communication interference cognitive system. According to the method, the interference characteristics are comprehensively captured, the incidence relation is deeply mined, the model structure is continuously optimized, and the accuracy and adaptability of a communication interference cognitive system in a complex interference environment are effectively improved.

The application provides an industrial cognitive base system based on multi-modal contrast learning and an execution method, wherein the industrial cognitive base system based on multi-modal contrast learning constructs a system architecture comprising a knowledge graph management module and a cross-modal encoding fusion module, dynamically injects industrial field knowledge in the form of a structured sub-graph into a multi-modal feature learning process, and combines a contrast learning mechanism with the introduction of industrial semantic constraints, so that the model obtained through final training can not only realize semantic alignment of multi-modal data, but also ensure that the unified semantic representation and reasoning results generated by the model strictly comply with pre-defined industrial logic rules, thereby effectively overcoming the defects that the simple data-driven method in the prior art may produce results contrary to industrial common sense, and significantly improving the output reliability, decision confidence and practical application value of the industrial cognitive system in key tasks such as fault diagnosis and state monitoring.

The invention provides a public opinion information retrieval method based on bidirectional enhancement of a big language model and a knowledge graph. The method comprises the following steps: constructing a bidirectional enhancement mechanism, enhancing the big language model through the knowledge graph, and enhancing the knowledge graph through the big language model to obtain an enhanced big language model and the enhanced knowledge graph; acquiring real-time public opinion data in real time, and dynamically constructing a knowledge graph by adopting the bidirectional enhancement mechanism according to the real-time public opinion data; personalized retrieval and sorting are carried out, retrieval calculation is carried out on the knowledge graph and the content in the public opinion data retrieval pool according to user retrieval content, a retrieval result is obtained through sorting, and public opinion information retrieval based on bidirectional enhancement of the large language model and the knowledge graph is completed. According to the method, the crossing from a static knowledge base to a dynamic cognitive system can be realized, and second-level response knowledge reserve and deduction capability are provided for decision making; the qualitative change from a retrieval result to decision insight is realized, the user is effectively enabled, and the user stickiness is improved.

The invention relates to the technical field of environment intelligent interaction, in particular to an environment intelligent interaction cognition system and method based on data driving, and the method comprises the steps: in a qualitative cognition stage, firstly carrying out the dimension reduction and structural classification of high-dimensional data through employing a PAM clustering algorithm, determining the optimal clustering number through a contour coefficient method, and achieving the preliminary organization of the data; then, an FP-Growth algorithm is used for mining association rules among features, an interpretable knowledge rule base is constructed, in the quantitative cognition stage, a qualitative analysis result is mapped back to original data, a POS-BP neural network model is adopted, the initial weight of a neural network is automatically optimized through a particle swarm optimization (PSO) algorithm, the convergence speed and prediction precision of the model are remarkably improved, and the prediction accuracy of the model is improved. And finally, accurate prediction of key indexes (quality, energy consumption, cost and the like) in the manufacturing process is realized through the optimized BP neural network. According to the method, effective mining of massive, high-dimensional and low-quality data in the continuous manufacturing industry is realized through a divide-and-conquer strategy.

This invention discloses a cognitive decision-making method and system based on multidimensional situation coding, belonging to the field of artificial intelligence and cognitive computing technology. The method includes: acquiring input data and converting it into an event sequence, thereby generating a six-dimensional external polar field; driving a six-dimensional energy vector to converge from an initial neutral state to a stable state under the action of an energy evolution equation; binarizing the converged energy vector and matching it with a situation coding library to determine the current situation type; generating a structured description and action suggestions for the current situation; selecting a target situation type and planning an evolution path based on situation value weights; and outputting a cognitive decision report. This invention uses events as the native cognitive unit, which can be characters, words, phrases, sentences, or any combination thereof. Through a six-dimensional complete situation space and an energy dynamics emergence mechanism, it provides cognitive systems with endogenous interpretability, a unified situation coding framework, autonomous planning capabilities based on intrinsic value, and online self-evolution capabilities. It can be widely applied to intelligent decision support, complex system situation assessment, and other scenarios.

This invention discloses a cognitive architecture for human-computer interaction and robot-to-robot communication based on a four-dimensional discrete spatiotemporal cognitive system, belonging to the field of artificial intelligence human-computer interaction and robot-to-robot communication technology. The four-dimensional discrete spatiotemporal cognitive system is embedded in a large language model and a robot control system, including a time coordinate axis, a context coordinate axis, a logical coordinate axis, and a device coordinate axis. The method includes: the cognitive system automatically generating four-dimensional discrete coordinates for each token, where the time coordinate is based on real-time time, the context coordinate is based on the dialogue sequence number, the logical coordinate is based on the token's sequential position in the interaction information, and the device coordinate is based on the device's globally unique identifier; in the attention mechanism of the large language model and the robot control system, the attention weight is adjusted according to the coordinate differences between tokens. This invention uses the four-dimensional discrete spatiotemporal cognitive system as a fixed underlying architecture for human-computer interaction and robot-to-robot communication, effectively solving problems such as context confusion, timing control, and multi-device collaboration in multi-turn dialogues during human-computer interaction and robot-to-robot communication.

A mechanism is provided in a data processing system comprising a processor and a memory. The memory comprises instructions which are executed by the processor to cause the processor to be specifically configured to implement a recognizer module for detecting user input in a cognitive environment. The recognizer module receives sensor signals from at least one wearable device being worn by a user. The recognizer module analyzes the sensor signals using a machine learning model to determine at least one user input indicator describing user input activity of the user. The recognizer module communicates the at least one user input indicator to a cognitive system executing within the cognitive environment. The cognitive system performs at least one cognitive action based on the at least one user input indicator.

The present application relates to a kind of intelligent reflecting surface IRS assisted cognitive wireless power transfer communication SWIPT system beamforming method, belong to wireless communication field.The method includes: S1: the IRS assisted cognitive SWIPT system scene transmission model is established.S2: with the minimum transmission power of secondary user transmitter as objective function, considering the minimum signal-to-noise ratio constraint of secondary user, power splitting coefficient and minimum energy collection constraint, intelligent reflecting surface phase shift constraint and the interference noise ratio threshold constraint of primary user, based on channel uncertainty, a joint optimization secondary user transmitter beam design, secondary user power splitting coefficient and intelligent reflecting surface phase shift design resource allocation model is established.S3: the non-convex problem with infinite constraints is converted into finite-dimensional deterministic convex optimization problem.S4: it is solved by converting into semi-definite programming problem.S5: for IRS phase shift design subproblem, penalty concave-convex process method is used to solve.

This invention relates to an incremental ontology-based cognitive system and method for intelligent agents with multi-dimensional ontology. The invention includes an ontology motion perception module, an ontology cognitive evaluation module, a memory pool module, an ontology knowledge mining module, and an ontology cognitive training module. The ontology motion perception module is used to obtain multimodal data; the ontology cognitive evaluation module is used to obtain ontology cognitive quantization error using the multimodal data; the memory pool module is used for data storage; the ontology knowledge mining module is used to establish an ontology knowledge base using multimodal data mining; and the ontology cognitive training module is used to construct an ontology cognitive model and train the ontology cognitive model under the guidance of the ontology cognitive quantization error and the ontology knowledge base. Compared with existing technologies, this invention has advantages such as multi-dimensional feature representation and perception of the ontology, improved accuracy of cognitive data, and reduced overall error.

The present invention relates to data mining and intelligent information processing technology, and in particular, to a personalized recommendation method based on a language-value binary concept lattice that can resolve recommendation interpretation ambiguity and cold start issues and can handle different data types. The method is performed in the following steps: data acquisition and preprocessing, collecting data information in multiple data formats, and determining whether the type of fuzzy data information collected by the computer is uniform. If not, the data type is converted into a language-value binary tuple: language-value binary concepts for user sets and item sets are constructed; language-value binary evaluation data for product decision attribute sets is obtained from users; a cognitive system for a training data set is constructed; a sufficient knowledge base and a fuzzy object language knowledge pseudo-lattice for the training data set are constructed; a necessary knowledge base and a fuzzy object language knowledge pseudo-lattice for the training data set are constructed; rules are extracted, selection and judgment are performed, and recommendation results are obtained.