2110 results about "State prediction" patented technology

The invention relates to the technical field of bearing fault detection, and discloses a bearing fault detection method and system based on a health state index. The method comprises the following steps: collecting multi-source sensing signals at least comprising a vibration signal, a temperature signal and an acoustic signal during bearing operation; respectively performing time domain feature extraction and frequency domain feature extraction on the multi-source sensing signals, and performing normalized fusion on the extracted time domain features and frequency domain features to generate a multi-dimensional health state index sequence; constructing a long-short-term memory network model based on an attention mechanism, inputting the multi-dimensional health state index sequence into the model for training, and outputting a bearing health state prediction sequence; and calculating a dynamic early warning threshold according to the historical health state prediction sequence, comparing the current prediction value with the dynamic early warning threshold in real time, and generating a fault early warning signal. The method can improve the accuracy of bearing health state evaluation and fault early warning, and is suitable for complex operation conditions.

The invention provides a fatigue life prediction and residual life evaluation method for a high-power heavy-duty gearbox, and belongs to the technical field of intelligent operation and maintenance based on computer data processing. Comprising the following steps: acquiring dynamic data in an operation process, and performing multi-scale decomposition to form multi-source multi-scale data; inputting the multi-source multi-scale data into a designed multi-scale fatigue feature extraction module and a health state prediction module to obtain a multi-scale health index sequence and a health state label; establishing a fatigue damage evolution model, introducing the generated health index sequence for self-adaptive updating, outputting a comprehensive damage value, performing staged evaluation of fatigue degradation to obtain a damage label set, and performing multi-scale health index sequence and health state labels as well as the comprehensive damage value and the damage label set to obtain a multi-scale health index sequence and health state labels; inputting into a designed double-source fusion fatigue life prediction model, and outputting residual life prediction quantity; according to the invention, high-precision prediction and residual life evaluation of the fatigue life of the high-power heavy-duty gearbox are realized.

The invention relates to the technical field of robot anomaly prediction, in particular to a robot anomaly prediction method and system based on multi-dimensional fusion and causal inference. The method comprises the steps of performing multi-scale depth state characterization based on acquired robot multi-joint sensing data, and performing dynamic causal graph fusion based on the multi-scale depth state characterization. Comprising the steps of priori knowledge graph construction based on a kinematics chain, dynamic association attention mechanism construction based on data driving, state fusion of knowledge and attention guidance and global state vector generation. Performing hierarchical space-time dependency prediction based on the fused features, wherein the hierarchical space-time dependency prediction comprises robot joint topological graph construction, spatial dependency dynamic modeling, long-range time evolution prediction and future robot health state prediction; the method shows excellent performance in a plurality of core dimensions such as prediction precision, early warning timeliness and diagnosis interpretability, and has extremely high actual deployment value and engineering popularization potential.

The invention relates to the technical field of equipment health state prediction, in particular to a multi-source sensing driven equipment health prediction method and system. The method comprises the following steps: synchronously acquiring equipment temperature, vibration, current and acoustic data through a multi-source sensor, carrying out denoising and standardization processing, dynamically distributing each signal weight to adapt to an equipment operation stage, generating a high-dimensional dynamic feature vector, and embedding a historical smoothing mechanism to realize continuous updating; performing standardization and nonlinear mapping on the features, constructing a dynamic coupling factor matrix to quantify a cooperative relationship between the features, fusing interaction information and adaptively enhancing abnormal features; three-layer progressive health prediction from a local part, a middle-layer subsystem to global equipment is implemented based on coupling characteristics, a trend consistency verification mechanism is introduced, global and middle-layer prediction differences are quantified through residual errors, weights are adaptively corrected, and the equipment health state evolution trend and the risk level are output. According to the method, the multi-working-condition adaptability, the feature coupling sensitivity and the prediction result reliability are remarkably improved.

The invention discloses a battery abnormal data prediction and state evaluation method, and relates to the technical field of battery state prediction, and the method mainly comprises the steps: carrying out the preprocessing of an experiment data set, and obtaining multi-dimensional time series data; a combined feature encoder, a pre-response encoder and a memory analysis module are constructed to realize a battery abnormal data fault prediction model; training the model by using the multi-dimensional time sequence data to obtain a trained model, and predicting the to-be-predicted data to obtain a prediction result; and calculating a reconstruction error between a prediction result and original data, constructing an AUROC evaluation model, and evaluating the battery abnormal data fault prediction model. By implementing the battery abnormal data prediction and state evaluation method provided by the invention, the feature extraction efficiency, the abnormal recognition precision, the detection stability and the generalization ability can be improved.

The invention discloses a power grid energy storage capacity demand determination method and system based on multiple time scales, relates to the technical field of power grid energy storage capacity demand calculation, and aims to solve the problem of inaccurate energy storage capacity demand calculation. By generating multiple scenes, quantitatively screening key scenes and incorporating various uncertain factors, energy storage capacity calculation focuses on high-influence scenes, the coping capacity of the scheme to actual risks is enhanced, decision scientificity is improved, ultra-short-term to long-term multi-time scales are divided, core contradictions of all the scales are captured in a targeted mode, limitation of a single scale is avoided, and energy storage capacity calculation efficiency is improved. According to the method, capacity requirements and equipment distribution are integrated, operation rules are defined, visual documents are generated, multi-scale energy storage cooperative operation is achieved, the stability of a power grid is guaranteed, meanwhile, cost is reduced, scheme landing performance and operation efficiency are improved, multi-time-scale power grid state prediction is carried out based on a dynamic database, and historical rules and real-time data are combined, so that the power grid state prediction efficiency is improved. And the prediction coordination is ensured through multi-dimensional verification.

The invention provides a compressor energy-saving operation control method and system based on reinforcement learning, and belongs to the technical field of compressor control. The method comprises the steps that multi-dimensional data in the operation process of a compressor are collected through a multi-parameter sensor network; preprocessing the multi-dimensional data to obtain target feature data; inputting the target characteristic data into a state prediction model, and predicting an operation parameter prediction value in a future control period; splicing and fusing the target feature data and the operation parameter predicted value, and constructing state representation of the reinforcement learning model; inputting the state representation into a target reinforcement learning model based on a near-end strategy optimization framework to obtain an optimal control action; safety verification is conducted on the optimal control action based on preset compressor safety operation constraints, and an execution instruction is determined; and adjusting operation parameters of the compressor based on the execution instruction. According to the compressor energy-saving operation control method and system based on reinforcement learning, the energy-saving performance and the operation stability of the compressor are improved.

The invention discloses a refrigerated cabinet multi-sensor heterogeneous data fusion method based on artificial intelligence, and the method comprises the following steps: collecting multi-source heterogeneous data in the operation process of a refrigerated cabinet, and constructing a unified event time axis; preprocessing the multi-modal data, and generating a physical enhanced multi-modal tensor in combination with a heat-airflow-electricity model; constructing a disturbance activation vector and a modal mask pattern, and embedding the disturbance activation vector and the modal mask pattern into a multi-modal tensor; inputting the disturbance enhanced input data into an improved Tide model to perform long sequence modeling, and outputting a state prediction sequence; performing prediction residual analysis, constructing a dynamic anomaly scoring function, and fusing a disturbance activation vector and a modal mask graph to generate an anomaly type mark and confidence fusion state representation; sparse gradient uploading, parameter aggregation and structure synchronization are carried out among multiple devices by adopting a federal training framework. According to the method, the multi-modal data fusion precision and the anomaly recognition reliability of the refrigerated cabinet can be improved, and cross-equipment collaborative intelligent optimization is realized.

The invention discloses a liquid cooling charging module health state prediction system, and relates to the technical field of electrochemical detection, and the system comprises a multi-parameter collection unit which collects the pressure and flow data of a cooling liquid, and is provided with an independent collection node at each branch of a double-gun charging system, and provides basic data for monitoring; the dynamic calibration unit communicates with the acquisition unit, compensates data drift based on a temperature-vibration interference model of machine learning training, executes baseline calibration by using an idle period to update a reference value, and guarantees data accuracy; the leakage detection unit is used for calculating a data change rate after compensation, outputting an early warning according to a preset condition and positioning a leakage branch; and the health state evaluation unit fuses the early warning signal and the historical operation data to generate a prediction result. The problems that tiny leakage detection is difficult and a sensor is prone to interference can be solved, manual inspection is reduced, the maintenance cost is reduced, and the requirement of a high-power charging scene is met.

The invention provides an alloy bearing unsteady-state damage detection and evaluation method based on digital twinning. The method comprises the following steps: constructing an alloy bearing multi-physical field digital twinning model; the method comprises the following steps: acquiring operation state data of a bearing under an unstable working condition based on a sensor network, preprocessing the acquired data to obtain multi-source features, and fusing the multi-source features to generate a comprehensive health index; inputting the preprocessed data into a digital twin model for forward simulation, generating a prediction observation vector, comparing the prediction observation vector with a sensing measurement value to generate a residual error, and performing damage state assimilation based on the residual error; and based on the assimilated state, predicting a damage evolution trajectory under an unsteady state working condition, and evaluating probability distribution of residual life. According to the method, the reduced-order proxy model for dynamic mode switching is constructed, the mode basis is automatically expanded and the sub-models are switched when the load suddenly changes or the rotating speed jumps, and the problem of feature drift of a traditional fixed basis function model under variable working conditions is effectively solved.

The invention discloses a water turbine operation state monitoring method based on digital twinning, which comprises the following steps: collecting vibration and pressure signals of limited measuring points of a water turbine in real time to form observation data; establishing a physical constraint Fourier operator network based on the observation data, and obtaining an operation state prediction result; performing data assimilation processing on a prediction result by using the observation data as a constraint condition through an iteration set Kalman smoother; the Fourier operator network is corrected through feedback of an optimization estimation result; and reconstructing vibration and pressure states of unmeasurable positions of the water turbine, realizing state field reconstruction, and judging abnormal positions and types. Accurate monitoring and real-time abnormity early warning of the running state of the water turbine can be achieved, and the monitoring accuracy and stability are improved.

The invention discloses a task scheduling method based on predictable resource state atlas modeling, a platform is oriented to a heterogeneous computing environment, a unified resource state atlas is constructed by collecting multi-dimensional resource state parameters of computing nodes, and performance characteristics and communication topological relations among the nodes are comprehensively described. On the basis, a bidirectional time sequence model and an attention mechanism are fused, and the load trend of each node in a future short time is predicted. The platform constructs a multi-factor scheduling scoring function based on a task feature vector and resource state prediction map, integrates parameters such as resource matching degree, prediction load, communication delay and energy consumption cost, dynamically evaluates the adaptability of tasks and resources, and realizes adaptive scheduling and optimal resource allocation of the tasks. Compared with the prior art, the method has the advantages of being high in resource state predictability, high in task allocation intelligence degree, outstanding in platform evolution capability and the like, and is suitable for intelligent task scheduling application in a large-scale heterogeneous resource environment.

The invention relates to the technical field of computers, and discloses a real-time rendering and interaction method and system for an immersive virtual reality scene. The method comprises the following steps: fusing tuner inertial data and eyeball tracking data, and constructing a prospective state prediction model; generating a predictive focus field in combination with scene visual saliency; synthesizing an anisotropic temporal-spatial resolution graph according to the predicted head angular velocity; gPU variable-rate coloring is driven to realize non-uniform rendering; and re-projection or dynamic fuzzy correction is executed in a self-adaptive manner according to the attitude prediction error before display. According to the technical scheme, the perception delay and the rendering load are remarkably reduced, and the frame rate stability and the visual immersion in a high-dynamic scene are improved.

The invention discloses an electric connector fault prediction method, and the method comprises the steps: synchronously obtaining multiple types of time series data of a contact area of a target electric connector in a preset time window based on a preset data collection device, and the time series data comprise temperature spatial and temporal distribution time series data, contact resistance transient time series data and a surface image sequence; performing feature analysis on the multi-source time sequence data by using a preset time sequence coupling sign extraction mechanism to obtain a multi-source phase change feature vector of the target electric connector in the current time window; inputting the multi-source phase change feature vector into a pre-trained electric connector health state prediction model, and outputting a prediction result of the health state of the target connector at the current moment, including a health state index; and performing differential early warning based on the predicted health state index of the target connector. Therefore, early and quantitative accurate prediction of the state of the electric connector is realized.

The invention discloses an electric vehicle charging monitoring method and device based on a heterogeneous feature fusion network, and the method comprises the steps: S01, collecting various time domain feature parameters and high-frequency key current signals of an electric vehicle battery in different charging states, and building a sample database; step S02, constructing an electric vehicle charging state prediction model, wherein the model comprises a time-frequency feature fusion network, a CNN-LSTM feature fusion module and a predicted value output module which are connected in sequence; step S04, the edge device receives the time domain characteristic parameters and the high-frequency key current of the tested electric vehicle battery in real time, and calls the deployed electric model for prediction to obtain a charging state prediction value; and S05, judging the abnormal charging state of the detected electric vehicle according to the state prediction value. The method has the advantages of being simple in implementation method, low in cost, high in monitoring precision, fast in early warning response, easy to apply on a large scale and the like.

The invention discloses a continuous charging production line intelligent management and control strategy and framework, and the framework further comprises a multi-physics field coupling modeling module which carries out the physical mechanism modeling and state prediction of a production process through a heat-fluid-solid three-field coupling model on the basis of a conventional data collection and control module; the intelligent twin module adopts a mode of dynamically fusing a mechanism model and a data-driven residual model to improve the precision and robustness of system state estimation; the scheduling optimization module is used for realizing dynamic adaptive scheduling of AGV paths and production tasks based on a multi-target deep reinforcement learning strategy; and the multi-layer safety causal reasoning module realizes advanced prediction and active prevention and control of safety risks through a causal map, fuzzy logic and a linkage response mechanism. According to the method, the problem that safety, quality and efficiency are difficult to collaboratively optimize in the prior art is solved, and the predictability, safety, scheduling efficiency and adaptive capacity of the production line are remarkably improved.

The invention belongs to the technical field of satellite orbit prediction, and provides a low-orbit satellite orbit state prediction method based on artificial intelligence, which comprises the following steps of: dividing step lengths of a predicted orbit according to a time window, calculating a resultant acceleration of each step length, performing stress input on an initial position and speed, and obtaining a nominal orbit state sequence; forming an augmented state vector with aerodynamic synthesis parameters; inputting a time sequence neural network model, and performing rolling updating on the augmented state according to the output; obtaining an original standard deviation through mapping by using uncertainty measurement, calculating a standardized residual error, solving a right side weighted quantile according to a normalized time weight, comparing the right side weighted quantile with a preset threshold value, and carrying out adaptive deviation adjustment; and calculating risk measurement by using the acquired data, comparing the risk measurement with a set threshold value and a priority rule, automatically executing a judgment action on the risk measurement, recording a trigger reason and the judgment action, and updating parameters in the time sequence neural network model.

The invention discloses a numerically-controlled machine tool fault prediction system, and the system comprises a data collection and preprocessing module which is used for collecting and preprocessing multi-modal data of a numerically-controlled machine tool; the data set construction module is used for constructing a cross-modal index and generating a multi-modal data set; the feature extraction module is used for extracting continuous type, periodic type, environment type and image type modal features to generate a multi-modal feature set; the feature processing module is used for forming a sample-level multi-modal feature set; the multi-modal fusion modeling module is used for inputting the sample-level multi-modal feature set into the improved MGPR model and outputting fusion feature representation; the state prediction module is used for inputting the fusion feature representation into an HMM model and outputting a prediction result and a health index; and the early warning generation module is used for generating and storing fault early warning information entries, realizing fusion and modeling of multi-source heterogeneous signals, and improving the accuracy and robustness of fault prediction of the numerical control machine tool.

The invention discloses an application method and device of a battery model management iteration system based on federated learning and block chain verification, and relates to the field of block chain energy storage management, and the method comprises the steps: deploying an intelligent contract on a block chain to publish a federated learning task, and initializing a global model; each participant uses a private data local training model, encrypts and uploads the private data local training model to a central coordination server for security aggregation; submitting the aggregated global model to a block chain verification layer, and automatically evaluating the performance and recording a result by the smart contract; the verified model is deployed to an actual scene for battery state prediction; collecting actual operation data of the battery through an oracle machine feedback network, and feeding back the data to a chain; and the intelligent contract incentive engine automatically re-evaluates the performance of the model according to feedback data, distributes rewards, and iteratively optimizes a loop. According to the method, the problem that data islands exist in key parameter prediction models such as the state of health and the remaining service life of the battery is solved, and the credibility and the sustainable evolution ability of the prediction models are enhanced.

The invention provides an equipment operation state prediction method and system based on digital twinning. The method comprises the following steps: firstly, constructing a multi-physics field digital twinning model comprising an equipment independent model and a coupling relation between equipment; a target stress event that causes damage to the device is identified by processing multi-modal sensor data in real time. Once identified, simulation deduction is performed in the twin model, the simulation simulates the propagation, transformation and accumulation process of the stress in the device network based on the dynamically adjusted stress conduction weight, and calculates the accumulated stress of each associated device. And finally, converting the accumulated stress into an equivalent aging increment of each device in combination with a preset damage model, updating a health state baseline of each device, and generating a prediction report of a future associated aging risk. According to the method, the conduction and accumulation effects of the stress in the equipment network are simulated, so that the spanning from monomer monitoring to system-level associated aging prediction is realized, and the accuracy of equipment state prediction is greatly improved.

The invention discloses a new energy access region power grid balance scheduling method based on artificial intelligence, and relates to the technical field of power grid scheduling, and the method comprises the steps: data collection and preprocessing: obtaining and processing power grid multi-source data and micro-scale meteorological data; meteorological feature coding: extracting dynamic meteorological features by using a long short-term memory network; constructing a dynamic space-time hypergraph and embedding nodes, and generating node dynamic embedding in combination with a graph convolutional network; based on power grid state prediction and pre-fault analysis of causal intervention, accurate prediction and fault identification are realized; and generating and executing a pre-fault scheduling strategy, and generating and executing an optimization strategy through reinforcement learning, so that the method can realize the transformation of the power grid from response type recovery to prospective self-healing, improves the toughness, reliability and economy of the power grid in an extreme scene, and is suitable for the power grid balance scheduling of a new energy access region.

The invention discloses a bridge girder erection machine structure health intelligent management system and an intelligent management method thereof. The system comprises a state data acquisition unit, a data preprocessing module, a structure state modeling module, a state prediction and residual life estimation module, an early warning and control module, a self-learning and optimization module and a system integration module. The method comprises sensor deployment, data preprocessing, structural state modeling, state prediction and residual life estimation, early warning and control, self-learning and optimization and system integration. Structural state monitoring data are collected in real time through a sensor, a structural state model is constructed after preprocessing, a future state is predicted in combination with time sequence prediction and a mechanism model, and the remaining safe life and risk points are estimated. And the system triggers multi-stage early warning according to a prediction result, executes an active maintenance control strategy, and continuously improves a prediction model through a self-learning and optimization module. According to the invention, real-time health management, remote operation and full-life-cycle risk control of the bridge girder erection machine structure can be realized.

The invention provides an ultrahigh bridge tower safety regulation and control method and system based on a concrete time-varying effect, and relates to the technical field of bridge tower regulation and control systems. Aiming at the problems that in the prior art, the prediction accuracy of long-term internal force redistribution caused by concrete shrinkage and creep is insufficient, the coupling effect of multiple risk factors and bridge tower concrete shrinkage and creep is insufficiently considered, so that the prediction accuracy of the state of a bridge tower is low, and a maintenance regulation and control guidance basis is lacked, multi-source heterogeneous data are fused, and the prediction accuracy of the state of the bridge tower is improved. Concrete shrinkage and creep, dynamic ice impact, static ice pressure, concrete freeze thawing and ice wedging influence indexes are calculated respectively, a comprehensive risk assessment model based on machine learning is adopted for fusion, and a comprehensive risk report containing risk levels and regulation and control suggestions is generated; according to the method, the accuracy of single risk factor assessment is remarkably improved, the comprehensive prediction precision of complex coupling risks is improved, and the decision robustness and maintenance efficiency of the system in a multi-risk concurrent complex scene are enhanced.

The invention belongs to the technical field of software system optimization and intelligent operation and maintenance, and discloses a shared platform autonomous upgrading method based on a graph neural network, and the method comprises the following specific steps: S1, component state perception and feature coding; s2, constructing a heterogeneous dynamic evolution diagram; s3, dynamic graph neural network embedding modeling; s4, state prediction and conflict evaluation; s5, constructing an upgrade game diagram and strategy reasoning; s6, executing an upgrading plan and real-time feedback acquisition; and S7, graph structure reconstruction and strategy closed-loop optimization. According to the method, the heterogeneous dynamic evolution diagram is constructed, the time sequence label and the state quantification model are introduced, real-time sensing and standardized modeling can be carried out on the running state of each service component in the sharing platform, and compared with a traditional upgrading method depending on static configuration and a predefined template, the upgrading efficiency is greatly improved. According to the mechanism, inductive expression of multi-dimensional features such as node resource occupation, access frequency and response fluctuation is completed in the stage before upgrading.

The invention relates to the field of lithium battery health state prediction, in particular to a lithium battery health state prediction method based on TKAN and PxLSTM. According to the scheme, the method comprises the steps of collecting multi-dimensional time sequence data, including voltage, current and temperature, of a lithium battery, forming an input sequence by the multi-dimensional time sequence data, and performing denoising and standardization processing on the data; hilbert-Huang transform is applied to the input sequence, the input sequence is decomposed into a plurality of intrinsic mode functions and residual errors, the time-frequency domain joint information entropy of each intrinsic mode function is calculated, the first k intrinsic mode functions and corresponding entropy values are selected according to entropy value sorting, and a compact feature vector is constructed in combination with the residual errors; the coded compact feature vectors are input into a TKAN module and a PxLSTM module at the same time; and performing weighted combination on the first hidden state sequence and the second hidden state sequence through a fusion weight to obtain a fusion feature, and outputting a lithium battery health state prediction result according to the fusion feature. The method is suitable for battery health state prediction.

The invention discloses a power distribution network transient characteristic prediction method based on supervised learning, and relates to the technical field of power distribution network state prediction, and the method comprises the steps: collecting historical operation data through a power distribution network monitoring system, carrying out the data preprocessing, and obtaining standardized multi-dimensional time series data; carrying out transient feature extraction, constructing a high-dimensional feature set, and carrying out feature dimension reduction according to a transient event tag to generate a feature subset; inputting the feature subset into a mixed supervised learning model of a gradient boosting decision tree GBDT and a long short-term memory network LSTM for joint training to obtain a transient feature prediction result; and calculating a root-mean-square error according to the transient characteristic prediction result and the real-time monitoring observation value of the power distribution network, and dynamically adjusting hyper-parameters of the supervised learning model based on a Bayesian optimization algorithm. According to the method, the detection accuracy can be improved, the calculation complexity can be reduced, and the discrimination capability and the time sequence prediction capability of the model are considered.

The invention provides a long-span bridge rotation in-place control method based on a neural network and a computer system, and the method comprises the steps: carrying out the spatial-temporal feature decoupling through collecting a rotation motion data sequence generated in a long-span bridge rotation construction process, and obtaining a rotation posture evolution feature and an environment interference coupling feature; then calling a pre-trained swivel dynamics prediction network to carry out joint dynamics modeling on the two to generate a state prediction sequence of a swivel process, then carrying out deviation comparison on the state prediction sequence and a preset swivel target trajectory sequence, and generating a swivel deviation risk assessment matrix containing a position deviation value and an attitude deviation value; and finally, based on the rotation deviation risk assessment matrix, driving the rotation control parameter optimization network to carry out compensation calculation, generating a rotation in-place control instruction set containing a traction force adjustment coefficient and a rotation speed correction value, and transmitting the rotation in-place control instruction set to a rotation driving system. According to the invention, the accuracy and reliability of rotation in-place control of the large-span bridge can be improved.

The invention discloses a method for confirming the connection relation of switch type equipment based on the topological state of a power distribution network. The method comprises the following steps: carrying out feature extraction on original telemetering data of the power distribution network to obtain telemetering features; performing feature extraction on the original remote signaling data of the power distribution network to obtain remote signaling features; fusing the telemetering features and the telesignaling features to generate fused features, and predicting the switch state of the switch-type equipment according to the fused features to obtain a switch state prediction value; generating a switch state prediction sequence according to the switch state prediction value to calculate the topology state of the power distribution network; and based on the topological state of the power distribution network, judging the connection relation of the switch-type equipment. According to the method, the switch connection relation is objectively deduced through extraction of remote measurement and remote signaling features and fusion prediction and topological state calculation. Data support is provided for power grid planning, and layout and power supply problems are avoided; the problem of fuzzy connection of the old power grid is solved without relying on ledgers or manpower; the method is suitable for complex scenes, and efficiency and accuracy are improved.

The invention discloses an equipment state prediction method and device, equipment and a readable storage medium, and is applied to the technical field of internet equipment management, and the method comprises the steps: determining a feature vector corresponding to each type of equipment data based on a data modal characteristic and business demand coupling principle, and carrying out the alignment of each feature vector, and obtaining an equipment state feature vector; uploading the equipment state feature vector to a server, so that the server updates an equipment state prediction model by utilizing cross-equipment group knowledge on the premise of not sharing equipment data; predicting by utilizing a latest equipment state prediction model based on the equipment state feature vector, and determining an equipment abnormal state; the latest equipment state prediction model is a model for global updating and local incremental updating of the edge node based on the server. According to the method, the information of each mode is deeply associated, global updating and local incremental updating are carried out on the model based on the server, the latest model is used for prediction, and the accuracy of state prediction is remarkably improved.

The invention relates to a computing power resource dynamic scheduling method, device and equipment based on deep reinforcement learning and a medium thereof, and the method comprises the steps: constructing a joint state vector through real-time fusion of a network layer channel state and computing layer node load data, and driving a strategy network to generate transmission parameters and resource allocation actions of cooperative control; the code modulation parameters of the wireless transmission module and the computing resource proportion of the target node are synchronously configured in the execution layer, and dynamic task scheduling in the channel decay environment is achieved; a multi-target reward mechanism is designed to couple transmission bit error rate penalty, resource utilization efficiency and task timeliness evaluation indexes, and a reinforcement learning agent is guided to balance communication stability and computing power demand conflicts; according to the method, strategy network parameters are optimized through time difference errors, closed-loop feedback is formed in combination with channel state prediction and node load updating, the problems of network and calculation layer splitting decision, insufficient dynamic adaptability and multi-target optimization imbalance in the prior art are effectively solved, and the task scheduling success rate in the time-varying wireless environment is improved.