1570 results about "Continual learning" patented technology

The invention provides an intranet service quality optimization method and system based on deep reinforcement learning, and relates to the technical field of artificial intelligence, and the method comprises the steps: collecting multi-dimensional performance index data, carrying out the preprocessing, constructing an analytic hierarchy model, outputting the performance score of each layer, integrating the performance score of each layer, and outputting an intranet service quality score. Constructing a simulation environment, adjusting hyper-parameters of the adaptive deep reinforcement learning model, continuously learning in an actual intranet environment, determining multi-dimensional rewards, performing multi-target optimization, and generating a multi-target optimization result; and setting a hierarchical performance index threshold, if a hierarchical optimization strategy is triggered, adjusting a network topology structure, adjusting resource allocation and service configuration, adjusting weights of different indexes in the service quality evaluation model, updating a reward function, solving an optimal service decision, generating an optimal service decision sequence, and executing the optimal service decision sequence.

The invention relates to power distribution network fault processing equipment applied to the technical field of power distribution, and aims to improve the fault prediction, diagnosis and processing capabilities of a power system. The equipment comprises a pre-training module, a neural network architecture module, a data acquisition layer, an edge calculation layer, a core processing layer, a context learning module, a reasoning processing module, an output and interaction layer and a continuous learning and improvement layer, and is characterized in that a large-scale pre-trained Transform model is utilized, and professional knowledge of a power system is combined; according to the method, deep understanding and analysis of a complex power grid state are realized, through multi-source data fusion, edge calculation and central cooperative processing, the equipment can monitor the power grid state in real time, accurately predict potential faults and provide intelligent diagnosis and repair suggestions, and model pre-training and optimization are performed by adopting self-supervised learning and federated learning technologies; using reinforcement learning to optimize a fault response strategy; the interpretability of the model is realized, and the decision transparency is improved.

A computer system implements a unified framework integrating an adaptive elastic funnel (AEF) with a convergent intelligence fabric (CIF) for multi-agent AI collaboration. The system provides a universal multi-modal key-value subsystem for sharing partial computations, implements hybrid placement strategies for dynamic memory management, and incorporates quantum-resistant secure enclaves. The architecture integrates hardware acceleration through GPU-FPGA hybrid caching and neuromorphic processors, applies adaptive energy and thermal management across hardware generations, and implements autonomous flash resource orchestration with multi-dimensional wear management. The system orchestrates tensor workflows using hierarchical scheduling, enables cross-agent collaboration with privacy preservation, and supports continuous learning without catastrophic forgetting. This integration delivers unprecedented computational efficiency and security in high-dimensional decision-making environments while supporting incremental adoption through modular interfaces.

The invention discloses an unmanned aerial vehicle route automatic planning method for electric power inspection, and relates to the technical field of unmanned aerial vehicle inspection. Comprising the following steps: starting an unmanned aerial vehicle, carrying out environment perception initialization, calculating the total cruise mileage, evaluating the interference risk, carrying out real-time obstacle avoidance, dynamically optimizing an inspection route, carrying out energy monitoring management, generating a return flight strategy, recording an inspection task and carrying out adaptive learning. Through dynamic electromagnetic interference modeling, multi-modal fusion perception, self-adaptive risk decision and cloud collaborative learning, the problem of insufficient adaptability of a traditional electric power inspection unmanned aerial vehicle in a complex electromagnetic environment and a dynamic obstacle scene is solved, the safety and the inspection efficiency are improved, the robustness is enhanced, and the method is suitable for popularization and application. Intelligent upgrading is carried out through continuous learning and multi-machine cooperation, the overall operation and maintenance cost of the system is reduced, a high-reliability and full-automatic inspection solution is provided for intelligent power grid construction, and the industrial application value is remarkable.

The invention discloses an infrastructure risk operation optimization management system based on AI driving, and belongs to the field of computer data processing and commercial management, and the system comprises a multi-modal causal twinning construction module which integrates on-site multi-modal data streams to construct a dynamic space-time causal map; the risk evolution deduction module is used for performing anti-fact simulation based on a causal atlas to construct a prospective risk model; the collaborative configuration optimization module is used for solving an optimal collaborative defense strategy according to the risk model; the instruction analysis and digital prescription generation module is used for analyzing the defense strategy into a job digital prescription for a specific risk scene; and the intervention efficiency attribution and evolution correction module performs attribution analysis according to the execution effect of the digital prescription and adaptively updates the causal atlas. According to the method, a comprehensive method of constructing a dynamic causal map for risk deduction, coupling resource constraints for collaborative optimization and performing closed-loop feedback on a correction model is adopted, and active prediction, accurate intervention and continuous learning optimization of capital construction risks can be realized.

The invention relates to the field of business data distribution, financial science and technology and medical health, and discloses a multi-agent cooperative task execution method, which comprises the following steps: receiving a task instruction input by a user, and carrying out task step configuration on the task instruction to obtain a sub-task sequence; judging whether the task information of each subtask meets a preset execution logic standard, and if yes, sending the subtask to an intelligent agent according to an identifier of the intelligent agent corresponding to the subtask; and recording task data generated when each agent executes the subtask, extracting a second keyword in the task data to construct a decision-making dependency relationship graph among the agents, extracting collaborative data of each agent collaborative completion task instruction from the decision-making dependency relationship graph, and optimizing subsequent task allocation by utilizing the collaborative data. According to the method, the cooperation process among the agents is optimized, and continuous learning and improvement are carried out, so that the overall efficiency and reliability of the multiple agents are improved.

The invention discloses a multi-working-condition-oriented equipment fault diagnosis method based on unbalanced learning, and the method comprises the following steps: collecting the operation data of equipment under different working conditions through a plurality of sensors, carrying out the preprocessing, feature extraction and feature selection of the collected original operation data, and obtaining a feature set; processing the feature set by adopting an unbalanced learning technology and a cost-sensitive learning algorithm to obtain a balanced data set; constructing a multi-working-condition fault diagnosis model based on learning, and optimizing model parameters by adopting a method of combining cross validation and grid search according to a balanced data set in a model training process; deploying a multi-working-condition fault diagnosis model obtained through training, carrying out the online diagnosis of equipment faults, and continuously updating and optimizing the model through a continuous learning mechanism. According to the method, the unbalanced learning technology is applied to multi-working-condition fault diagnosis, the accuracy and robustness of diagnosis can be remarkably improved, and real-time monitoring and predictive maintenance of the equipment health state are achieved.

The invention relates to the technical field of data processing, and discloses a data processing method and system for cloud computing and a storage medium. The method comprises the following steps: acquiring preprocessed data through a distributed acquisition node to obtain a structured data packet; performing classification analysis through mixed feature extraction to obtain a safety disaster recovery data set; analyzing the incidence relation through a heterogeneous fusion network to generate a knowledge base; detecting abnormity through a self-adaptive threshold algorithm to form a risk library; determining a recovery scheme through a multi-objective decision to generate an execution log; and a protection strategy is formed through multi-angle comparison and optimization of parameters. According to the method, the data association relationship is mined and analyzed by constructing the multi-level heterogeneous fusion network, the optimal disaster recovery scheme is automatically determined through the multi-target decision process, and meanwhile, a continuous learning optimization mechanism is introduced; the intelligent level of data processing in the cloud computing environment, the timeliness of disaster recovery response and the stability of system operation are remarkably improved.

The invention discloses a user behavior prediction system and method based on multi-modal data fusion, and particularly relates to the field of user behavior prediction, and the system comprises a multi-modal data collection module, a preprocessing and feature extraction module, a cross-modal fusion module, a user behavior prediction module, and a model optimization and feedback module. According to the system, multi-dimensional original data such as visual sense, auditory sense, text, physiological signals and environment context of a user are acquired in real time through a multi-modal data acquisition module; then, deep networks such as ResNet, VGGish and BERT are adopted to extract high-dimensional feature vectors of all modals, and contribution weights of features of different modals are dynamically learned through an attention mechanism; and finally, based on a time sequence model of Transform and LSTM, analyzing fusion features, and outputting probability distribution of future behavior intentions. And parameter joint optimization and continuous learning are realized through a multi-objective loss function and end-to-end back propagation.

The invention relates to a data processing method, device and equipment based on an artificial intelligence technology and a data model. The data processing method comprises the following steps: acquiring historical process parameter data of original production equipment and real-time process parameter data of novel equipment; constructing an association network model reflecting the mutual influence strength of the parameters; carrying out partial correlation analysis and cross correlation analysis on the real-time process parameters of the novel equipment and the historical process parameters of the original production equipment; performing incremental updating on the process parameter prediction model; integrating the real-time process parameter characteristic representation of the novel equipment into the association network model; applying the optimal process parameter configuration set to an actual production process; and adaptability knowledge introduced by novel equipment is continuously accumulated and perfected. According to the method, a set of dynamically optimized closed-loop control mechanism is formed through continuous learning and knowledge accumulation, technological parameter changes caused by introduction of novel equipment can be effectively coped with, and continuous optimization and quality improvement of the production process are achieved.

The invention relates to the technical field of unmanned aerial vehicle remote sensing and artificial intelligence crossing, in particular to an intelligent monitoring system and method based on multi-modal remote sensing data and deep learning, and the system comprises an unmanned aerial vehicle cluster networking subsystem, a mixed feature matching subsystem and a multi-modal fusion and continuous learning subsystem. The method comprises the following steps: constructing an unmanned aerial vehicle cluster carrying a multispectral sensor and a laser radar LiDAR, and carrying out wireless networking among a plurality of unmanned aerial vehicles to realize sharing of acquired images; feature point extraction is carried out on collected images of different time phases, the extracted feature points are input into the generative adversarial network, and the feature points are matched; and receiving the matched collected images, dynamically fusing data of visible light, infrared and other multi-modal images through a space-time attention mechanism, and realizing high-precision target recognition and dynamic environment self-adaption in a small sample scene. According to the method, unmanned aerial vehicle multi-source remote sensing data acquisition, feature fusion and deep reinforcement learning are combined, and the method is used for intelligently monitoring a dynamic environment.

The invention discloses a green building intelligent lighting and energy collaborative optimization system based on multi-source data fusion. According to the system, firstly, multi-source information such as indoor and outdoor illumination, weather, personnel occupation and electricity price is subjected to weighted fusion through dynamic confidence, and an accurate real-time environment state is constructed; then, model predictive control is combined with a multi-objective optimization strategy, and a sunshade device, intelligent lighting and a heating ventilation air conditioner are cooperatively adjusted, so that the comprehensive energy consumption or operation cost is minimized on the premise that the indoor illumination and temperature comfort requirements are met; besides, the self-adaptive module continuously learns and adjusts the thermal characteristics and the use mode of the building through online model parameter correction and occupancy probability prediction, and it is ensured that the energy-saving effect is stable for a long time; the system drives each execution device through a standard building self-control interface, is easy to integrate in a newly built or reformed project, and can significantly improve the energy utilization efficiency and the indoor environment quality.

The invention belongs to the technical field of aircraft navigation, guidance and control, and provides a satellite exploration control system and method based on deep reinforcement learning, and the system comprises a sensing layer, a decision layer, and an execution layer. The sensing layer is used for integrating equipment data and performing primary processing; the decision-making layer is used for deploying a deep reinforcement learning model and regularly carrying out data training and model updating; the execution layer is used for controlling satellite attitude and orbit adjustment and performing autonomous fault diagnosis and recovery operation; according to the method, effective control under different environments and emergencies is realized through continuous learning and optimization, the adaptive capacity and the autonomous control capacity are relatively high, and an efficient and accurate control strategy can be realized in a complex and changeable space environment through gradual optimization and iteration, so that the method is suitable for large-scale popularization and application. And the autonomous operation capability and the task execution efficiency of the satellite are greatly improved.

The invention belongs to the field of network security, and particularly discloses a network defense agent system based on a large language model. Through the design of the sensing layer, the decision analysis layer and the action execution layer, comprehensive protection of network threats is realized. The sensing layer is responsible for collecting original information from multiple channels and converting the original information into standardized data; the decision analysis layer performs modeling and threat reasoning on attack behaviors, evaluates a risk level and predicts subsequent actions; and the action execution layer specifically executes defense operation according to the defense strategy scheme output by the decision analysis layer. In addition, the application also constructs a data set oriented to attack and defense confrontation, records a complete attack sequence, defense response and effect evaluation thereof, and provides a reliable basis for continuous learning of defense agents. Experimental results show that the framework provided by the invention is superior to the traditional method in the aspects of attack detection accuracy, attack chain identification and defense strategy generation, and has stronger adaptability and real-time response capability.

The invention relates to the technical field of program exception positioning, in particular to an AI-based data interface program exception fault pre-positioning algorithm, which comprises the following steps: S1, data stream pre-analysis; s2, capturing an abnormal signal; s3, performing abnormal clustering and feature extraction; s4, interface behavior modeling is carried out; s5, abnormal traceability positioning; s6, generating a self-adaptive repair suggestion; s7, performing feedback circulation and continuous learning; and S8, performing visualization and report generation. Through the steps of data stream classification, key field extraction, behavior modeling and the like, the system can accurately capture abnormal signals in mass data, meanwhile, an abnormal mode is visually displayed through clustering analysis and a dimension reduction technology, a reliable foundation is laid for subsequent positioning and repairing, and in addition, the method has the advantages of being high in practicability and easy to popularize. The abnormal source is quickly positioned by adopting methods such as topological sorting and causal network reasoning, the false alarm rate and the missing report rate in the positioning process are effectively reduced by combining a sliding window and time sequence analysis, and the problem solving efficiency is greatly improved.

Aiming at the problems of insufficient language understanding depth, weak personalized dialogue generation ability, lack of continuous learning and long-term user state modeling and the like in the current emotion recognition and psychological intervention technology, the invention provides a user emotion recognition and psychological intervention method combined with a large language model (LLM). According to the method, the potential emotional state is identified by analyzing free text information input by a user by utilizing the powerful capabilities of a large language model in the aspects of natural language understanding, emotional modeling and text generation; constructing a multi-round dialogue context, and reasoning a psychological change trend of the user; in combination with a psychological knowledge base, personalized and mild psychological intervention dialogue content with a dredging effect is generated. The system supports recognition and classification of various emotional states such as depression, anxiety and alonity, is suitable for various interaction scenes (such as APPs, webpages and social robots), and can greatly improve the precision of emotion recognition and the timeliness and effectiveness of psychological intervention. The emotion recognition and psychological intervention method based on the large language model provides solid technical support for constructing an intelligent, continuous and personalized psychological health management system, and has wide application prospects and profound social significance.

The invention discloses a talent matching method and system based on artificial intelligence, and aims to improve human resource configuration efficiency and decision intelligence. The method comprises the following steps: collecting talent and demand data in multiple channels, especially unstructured communication data including interview records and work communication records; processing data through technologies such as multi-mode resume analysis, extracting features and fusing the features; an enterprise talent knowledge base which is used for continuous learning and dynamic maintenance based on a system operation result and multi-source feedback and integrates structured and unstructured data is constructed; based on the natural language query of the user, providing intelligent question answering and decision support by using a retrieval enhancement generation model connected with the specific enterprise talent knowledge base; an advanced deep learning algorithm is adopted, a knowledge base is combined to carry out man-post matching and generate recommendation of context perception, and self-optimization of a matching strategy is realized through mechanisms such as reinforcement learning and the like. According to the invention, accurate and dynamic talent matching and intelligent decision making can be realized.

A system for identifying and preventing Advanced Persistent Threats (APTs) using big data analytics, including: a data collection module [100] configured to collect real-time data from network devices, endpoints, security applications and external threat data feeds; a data preprocessing module [101] configured to normalize and filter the collected data; a big data analytics engine [102] that uses both supervised and unsupervised machine learning models and is configured to detect anomalous behaviors and known attack patterns from the preprocessed data; a classification module [103] configured to categorize detected threats based on identified behaviors and characteristics of APTs; a real-time threat response module [104] configured to automatically initiate countermeasures based on the classified threat, the countermeasures including isolating compromised systems, blocking malicious traffic and adjusting network security configurations; a continuous learning module [105] configured to update the machine learning models using feedback from past threats and external threat data to ensure adaptability to evolving APT strategies; and a management dashboard [106] configured to provide security administrators with real-time alerts, visualizations and control over system configurations and responses.

The invention discloses a warehouse level code problem solving system based on a large language model, and relates to the technical field of software engineering, a source code file is extracted from a code warehouse, a static analysis tool is utilized to analyze codes, a warehouse level code graph is constructed, and then code entities and dependency relationships thereof are analyzed; based on the warehouse-level code graph, combining historical problem solving data, and constructing a dynamic evolution knowledge graph through continuous learning; through a context retrieval module, according to the context information of the current problem, dynamically retrieving related solution strategies and generating corresponding code patches, complementing or reviewing suggestions; the validity of the generated patch is verified through static analysis, unit testing and integration testing, and a problem solving strategy is optimized according to feedback of a developer. The problem solving efficiency and accuracy can be improved, the development cost is reduced, the solving process is gradually optimized through an intelligent feedback mechanism, a closed-loop feedback mechanism is formed, and therefore the increasingly complex software development requirements are met.

The invention relates to the field of civil engineering and artificial intelligence, in particular to a transfer learning optimization system and method for predicting the early-age strength of concrete, and the system comprises a multi-scale data sensing module, a physical information constrained deep neural network module, a formula adaptive transfer learning module, a Bayesian optimization prediction module and a federated learning feedback module. The whole process from data collection to model optimization is achieved, through the system, the concrete strength prediction errors of the extremely early age and the standard age are remarkably reduced to + / -5% and + / -3% respectively, meanwhile, the number of concrete test pieces for testing is reduced by 85%, the material and labor cost is greatly saved, and the method not only improves the prediction precision, but also reduces the construction cost. And through continuous learning and feedback, the prediction model is continuously optimized, and an efficient and economical concrete strength prediction solution is provided for actual engineering.

The invention discloses a tough city resource allocation method and system based on a block chain and edge computing, and relates to the technical field of city resource allocation, a federal protocol is utilized to exchange and shear a gradient abstract synchronization model version, then a prediction abstract containing fingerprints is written into a permission chain, and a time sequence is solidified through Byzantine consensus; an on-chain smart contract executes adaptive threshold evaluation based on sliding window prediction to generate a deployment intention with a verifiable path, the intention is sent to an execution subsystem through a side chain message bus in a low-delay mode, and cross-department resource rearrangement is completed through rolling optimization of an adaptive decision driver; an execution receipt is mapped into an incremental patch and written back to a local state tensor, a node trust weight is dynamically adjusted, and when the coverage degree and the weight rising range meet a triggering strategy, differential shearing training is started, and a new structure fingerprint and a model signature are generated; and efficient, credible and dynamic allocation and sustainable learning synchronous evolution of urban public resources in an emergency scene can be ensured.

The invention discloses a precast beam field production scheduling analysis management method and system based on a large model, and the method comprises the steps: combing precast beam production technological processes, constructing a beam field Internet of Things sensing network, and collecting production information in real time; defining a general engineering data format, establishing a unified standardized beam field production knowledge base, and performing multi-modal data fusion on production elements; analyzing feasibility constraint rule conditions, describing a precast beam production scheduling problem, and establishing a beam field self-organizing scheduling decision mechanism; defining production state characteristics, establishing a beam field scheduling decision process model, and training a large scheduling model based on a PPO algorithm; and developing a beam field scheduling analysis management platform, automatically generating a production scheduling scheme in combination with a real-time production state, and realizing dynamic management of beam field production scheduling. The problems that in the prior art, the resource utilization efficiency is low, scheduling decision making depends on artificial experience, and continuous learning and self-optimization capacity is lacked are solved, and the intelligent level and the management effect of the precast beam field are improved.

The invention relates to the technical field of computer systems based on specific calculation models, and discloses a blood oxygen change monitoring algorithm fused with dynamic Bayesian modeling, which comprises the following steps: quantizing signal uncertainty by calculating entropy of pulse waveform harmonic energy distribution; and on the basis of the entropy value, driving a dynamic Bayesian network to carry out confidence coefficient evaluation and weighting processing on the blood oxygen estimation model, and finally outputting a decision pair containing a blood oxygen estimation value and the confidence coefficient thereof. According to the method, the signal uncertainty is converted into a computable entropy index, so that the system can autonomously distinguish real physiological changes and measurement noise, the problem of misjudgment caused by motion interference in traditional blood oxygen monitoring is avoided, meanwhile, non-inductive personalized calibration is achieved by utilizing the continuous learning ability of the Bayesian network, and the accuracy of the system is improved. And the reliability and the practicability of the medical wearable equipment are remarkably improved.

The invention discloses an AI-driven robot working environment safety sensing method, and relates to the technical field of artificial intelligence, and the method comprises the steps: collecting the visual, laser radar, force sense and environment parameter data of a working environment in real time through a multi-mode sensor; performing feature extraction and semantic understanding on the acquired data by using a deep learning algorithm to generate multi-dimensional perception information of the environment; based on a preset security policy model, in combination with the multi-dimensional perception information, predicting and classifying potential risks; according to a risk classification result, a behavior strategy of the robot is dynamically adjusted, and active safety decision is realized; and through the continuous learning module, the deep learning model and the security policy model are optimized online by using newly collected data so as to adapt to environmental changes.

The invention provides an airfoil profile adaptive design and optimization method based on reinforcement learning and a conditional generative adversarial network, and belongs to the field of airfoil profile optimization design. The method aims at achieving the two core functions of self-adaptive optimal airfoil generation and rapid optimization. According to the first function, through interaction of a reinforcement learning agent and a conditional generative adversarial network generator, generation strategies for generating optimal airfoil profile potential variables under different design conditions are continuously learned, so that the airfoil profile with the optimal aerodynamic performance is adaptively generated, intelligent and adaptive design of the airfoil profile is achieved, and the optimization period is remarkably shortened. Secondly, a self-adaptive agent model based on meta-learning is introduced, the model not only establishes a mapping relation among design conditions, aerodynamic performance and potential variables and directly predicts the optimal potential variables of a specific airfoil profile, but also can quickly adapt to an updated database through simple adjustment; the agent model predicts the optimal potential variable corresponding to the specific airfoil profile for the self-adaptive generation framework, so that the agent can quickly optimize the specific airfoil profile under the condition of fine adjustment.

The invention relates to a long-range task-oriented VLA model method for a humanoid robot, which comprises the following steps of: S01, analyzing a natural language instruction through a space-time semantic analyzer to generate an atomic operation sequence with a space-time dependency relationship; s02, maintaining a task state machine by using a dynamic memory network, and tracking the task execution progress in real time; s03, integrating vision, language and sensor data through a multi-modal perception fusion engine; s04, calling a predefined action primitive based on an adaptive execution system and optimizing a motion track; and S05, performing online updating and optimization on the model through a continuous learning mechanism. According to the method, a natural language instruction is analyzed into a structured task sequence with space-time dependence through a space-time semantic analyzer, an execution sequence and preconditions are defined, the semantic understanding ability and the structuring degree of task planning are improved, and task decomposition and replanning in a dynamic environment are supported; according to the method, the LSTM and the knowledge graph are combined, and the task state is maintained in real time.

The invention discloses a log abnormal behavior detection method and system based on periodic pattern mining and incremental learning, and relates to the technical field of system abnormality monitoring. The method comprises the steps of obtaining original log data and executing preprocessing operation; main periodic frequency components are extracted through time frequency analysis to form a periodic set, a periodic stable part sequence and a transient change part sequence are divided, and a periodic modeling mechanism and a transient modeling mechanism are used for modeling; a joint prediction model is constructed, Monte Carlo Dropout is introduced to estimate uncertainty, and Bayesian weighting is adopted to generate a prediction value; abnormity is judged through a periodic residual error, a transient residual error and an overall residual error, and an abnormal causal path is analyzed and identified in combination with transfer entropy; a memory sample driven playback and distillation mechanism is adopted to execute incremental training, and modeling structure parameters are dynamically adjusted. The method has the capabilities of periodic rule modeling, unsteady behavior expression, prediction fusion, abnormal causal identification and continuous model learning.

The invention discloses an intelligent service dialogue method and system fusing a generative large model and a knowledge graph, and the system comprises a language model module which builds a statement reply model based on BERT and carries out the training through a question and answer data set; the data extraction module is used for generating candidate item sets and dividing frequent item sets and non-frequent item sets; the knowledge graph module is used for generating association rules and constructing a first knowledge graph and a second knowledge graph; the mixed retrieval module is used for performing cross retrieval on the two knowledge maps to generate a path; the personalized user module is used for determining a demand intention and an emotional tendency value; the statement generation module is used for generating a reply statement by taking the emotional tendency value as a dynamic weight and combining a user demand and a retrieval path; and the habit perception module is used for constructing a habit function and continuously learning habit characteristics of the user through a reinforcement learning optimization model. The system not only can provide accurate and personalized dialogue service for the user, but also can better adapt to the use habit of the user, improves the user experience, and has better interpretability.

The invention belongs to the technical field of fault diagnosis, and particularly discloses a numerical control system fault diagnosis method and system based on knowledge injection, and the method comprises the steps: carrying out the fault diagnosis based on an RAG frame, the RAG frame comprises a knowledge base and a large model, the knowledge base is a knowledge graph constructed based on the fault diagnosis historical data of a numerical control system, and the large model is a knowledge graph constructed based on the fault diagnosis historical data of the numerical control system; retrieving a fault path related to the user question from the knowledge base, and then comprehensively retrieving the fault path and the user question by a large model to generate a fault diagnosis result; after each fault diagnosis, judging whether a user problem is solved, and if the user problem is solved, directly adding a fault path into a feedback database; if not, adding the problem fed back by the engineer into a feedback database in combination with the fault path and the feedback knowledge of the engineer; and performing knowledge injection on the large model through the data in the feedback database, and updating the knowledge graph. According to the method, continuous learning optimization of a large model and a knowledge base can be realized, and the fault diagnosis accuracy is improved.

The invention discloses an industrial time sequence prediction method based on adaptive continuous learning. The method comprises the following steps: firstly, dividing a non-stationary industrial time series data set to obtain a plurality of domains with the maximum distribution difference; different time domains are then modeled in sequence, and an improved empirical playback (DER + +) method is used to avoid catastrophic forgetting of previously accumulated knowledge. Meanwhile, a soft sample buffer area is introduced to promote memory and learning of key modes in the current field. And finally, the time-sensitive activation function TimeRelu enables the time convolutional network (TCN) to have a time evolution property, and the generalization ability of the prediction model is enhanced. According to the method, the continuous learning normal form is introduced into the time sequence prediction task, the limitations of huge resource overhead of traditional cumulative training, disastrous forgetting of an incremental learning mode and the like are overcome, and the method has theoretical and practical significance on industrial time sequence prediction.