2007 results about "Simulation training" patented technology

The invention relates to the technical field of intelligent building energy-saving control, and discloses a green building equipment intelligent energy-saving control method and system based on artificial intelligence, and the method comprises the steps: collecting building environment data, and carrying out the preprocessing; and simulating the energy consumption of the building equipment by using the digital twin model, and training the baseline neural network model to obtain an initial control strategy. And adopting a genetic algorithm to optimize hyper-parameters of the baseline model, and generating a plurality of candidate energy-saving control strategies. And inputting the optimized strategy into the deep reinforcement learning model, carrying out strategy iteration, calculating the energy-saving efficiency and comfort score after each iteration, screening the optimal strategy through a reward mechanism, and improving the accuracy of energy consumption prediction of the building equipment. The hyper-parameter of the neural network is optimized by using the genetic algorithm, and the intelligent degree of the energy-saving control strategy is improved. On the basis of deep reinforcement learning, the intelligent agent can adaptively adjust an energy-saving strategy under different environmental conditions, and the user comfort is guaranteed while the optimal building energy consumption is ensured.

The invention provides an agent construction method and system based on a large language model, a medium and equipment, and relates to the technical field of decision agent construction. According to the method, an appropriate large language model is selected, a series of prompt project texts are established for the large language model, target codes are generated as required by utilizing the large language model and prompt projects, and class library functions of a network model library, a reinforcement learning algorithm and a simulation training environment can be correspondingly called by running the codes. Finally, auxiliary construction and training of the decision-making agent are realized, and the problems in the existing decision-making agent construction and training process are solved.

The invention discloses a robot reinforcement learning control method and system based on a gating circulation unit, and relates to the technical field of artificial intelligence. Based on a reinforcement learning state machine, a trained student network is deployed to a real quadruped robot in a zero mapping manner, and movement control of the quadruped robot is realized; the training process of the student network is as follows: building a simulation training terrain environment of the robot, and setting different levels of terrain difficulty for each terrain; based on the distillation learning thought, establishing a teacher network and a student network based on a gating circulation unit, constructing a total loss function, and based on the constructed total loss function, fitting a motor joint angle output by the teacher network through an omnibearing simulation training terrain environment, a large-scale parallel training method and the student network; according to the robot reinforcement learning control method and system, robust motion of the robot in an unstructured terrain is realized, and the robot reinforcement learning control method and system have good anti-interference capability and self-recovery capability for a dynamic environment in which traditional control cannot be competent.

The invention belongs to the technical field of robot control, and discloses a high-gear-ratio humanoid robot robust walking control method and system based on potential dynamics self-adaption, and the method comprises the steps: designing a control strategy training frame based on deep reinforcement learning, and carrying out the optimization through an Actor-Critic structure and a PPO algorithm; constructing a potential dynamic adaptive network LDAN, and extracting environment and ontology dynamic parameters through a variational auto-encoder; designing a multi-dimensional reward function; constructing a periodic gait library by using von Mises distribution and motion capture data; gradually introducing terrain disturbance and dynamic change through curriculum type simulation training; the trained strategy and the LDAN module are deployed on the high-gear-ratio driven humanoid robot, the problems that the high-gear-ratio humanoid robot is poor in motion stability, poor in adaptability and insufficient in action expression in a variable environment are solved, and the high-gear-ratio driven humanoid robot has the advantages of being high in robustness, high in natural expressivity and high in migration ability.

The invention relates to the technical field of power system operation training, and discloses a virtual reality-based power grid fault simulation training system and method, and the system comprises a virtual scene modeling module which is used for constructing a three-dimensional model of power grid equipment and an operation environment thereof, and providing a vivid training environment for subsequent modules; and the interaction and operation module is connected with the virtual scene modeling module and realizes interaction between the trainee and the virtual environment by using a VR head-mounted display, a handle and a motion capture device. By adopting a high-precision three-dimensional modeling technology, PBR material rendering and an LOD optimization strategy, high simulation of power grid equipment and an operation environment is realized, students can experience a vivid visual effect and immersion in a virtual reality scene, and meanwhile, by combining a gesture recognition technology realized by motion capture equipment and a convolutional neural network, the real-time simulation of the power grid equipment and the operation environment is realized. Operation actions of trainees can be accurately captured, a natural and visual interaction mode is provided, and the authenticity and operation experience of the training system are greatly improved.

The invention provides a scene reproduction and augmented reality simulation training method for floating installation. The method comprises the following steps: acquiring an external environment and an equipment state in real time; acquiring operation behavior data of the trainee in the training process, performing timestamp matching on the operation behavior data and the dynamic simulation scene, constructing a synchronous model, calculating a key performance index based on the synchronized data to quantify the operation effect of the trainee, and obtaining an operation performance record; when it is detected that the student fails to operate, a failure scene of the student is reconstructed based on currently collected operation behavior data and a dynamic simulation scene, and a failure path is analyzed by using a reinforcement learning model; generating an optimized floating track and an optimized posture adjustment path according to the failure scene of the reconstructed student and the optimization suggestion; and the trainee completes the floating support installation operation according to the optimized path displayed by the augmented reality equipment. According to the invention, the efficiency and effect of floating support installation training can be greatly improved, and a more efficient and intelligent solution is provided for the ocean engineering field.

The invention provides a PICC puncture virtual simulation training system and method constructed based on XR. A nursing skill simulation training platform and an intelligent training room based on augmented reality (XR) combine a simulation medical technology, an artificial intelligence large model, XR intelligent hardware, intelligent agent application, space calculation, computer vision and a mainstream technology of Internet of Things in scheme planning and design. According to the PICC puncture virtual simulation training platform, the ground practice, standardization, high virtual simulation and adaptive application of PICC puncture virtual simulation training are realized, and the intelligent development process and teaching concept of PICC puncture teaching are improved. According to the PICC puncture teaching system, intelligent virtual tutors, puncture teaching holographic three-dimensional sharing, virtual reality fusion technologies and the like are provided, an interactive teaching mode can be brought to PICC puncture teaching, the interest of students in teaching and clear operation guidance are improved, the theory and practice are fused, and the clinical cooperation ability is improved.

The invention provides a storage environment deep reinforcement learning multi-AGV conflict-free path planning method, and the method comprises the steps: creating a virtual storage simulation training environment, and extracting virtual storage simulation training environment information from the virtual storage simulation training environment; determining a state observation value code of each AGV according to the virtual storage simulation training environment information; setting an action space of the AGV, and constructing a reward function of deep reinforcement learning; and constructing a distributed conflict-free multi-AGV path planner introducing a dynamic grouping mechanism, carrying out iterative training on the distributed conflict-free multi-AGV path planner through a deep reinforcement learning network according to the reward function and the state observation value code and the action space of each AGV, updating network parameters of the distributed conflict-free multi-AGV path planner, and carrying out path planning on the distributed conflict-free multi-AGV path planner according to the network parameters of the distributed conflict-free multi-AGV path planner. A trained distributed conflict-free multi-AGV path planner is obtained; and carrying out conflict-free path planning on a plurality of AGVs in an actual storage environment by using the trained distributed conflict-free multi-AGV path planner to obtain a conflict-free path of each AGV.

The invention discloses a deep reinforcement learning quadruped robot motion control method and system based on constraint rewards. The method comprises the following steps: establishing a simulation training environment of quadruped robot deep reinforcement learning; determining a reward function, a domain randomization parameter and a cost constraint function of the simulation training environment; based on the robot information and the first simulation environment information, training the initial strategy network model through a reward function and a cost constraint function in a simulation training environment to obtain a trained strategy network model; establishing a reasoning test environment, and deploying the trained strategy network model to the reasoning test environment for model reasoning test optimization to obtain a target strategy network model; and deploying the target strategy network model into the quadruped robot to perform motion control on the quadruped robot. According to the method, the difference between a simulation training environment and a real environment is reduced, so that the quadruped robot is stably controlled by the target strategy network model in the real environment.

The invention relates to the technical field of image processing, and discloses a virtual operation analogue simulation method and system based on a mixed reality technology. The method comprises the steps that three-dimensional space information is collected, feature extraction and registration are conducted on a real operation environment and video data, and mixed reality mapping and monitoring data are generated; performing medical image registration processing based on mixed reality mapping to obtain virtual scene data; processing instrument sensing data according to the mapping data and the monitoring parameters to obtain interaction data; performing surgical process simulation according to the virtual scene and the interaction data; performing operation trajectory analysis and identification based on the simulation data, and generating evaluation and early warning information; and analyzing the evaluation and early warning information, and outputting an analysis report. Through multi-source data fusion and a real-time evaluation feedback mechanism, the authenticity and accuracy of operation simulation training are effectively improved.

The invention relates to the technical field of robot training. The robot self-adaptive training method based on reinforcement learning comprises the steps that task sub-target information is generated through a high-level strategy network, the task sub-target information is input into a low-level execution network, an action control instruction is generated according to the task sub-target information, interaction feedback information is collected in the execution process, and the action control instruction is sent to a robot through a robot. Calculating a reward value according to the interaction feedback information, carrying out association processing on the reward value and the scene complexity parameter, executing a dynamic reward shaping operation, generating an adjusted reward signal, generating a strategy model optimized by meta-learning based on the adjusted reward signal, loading the strategy model in a simulation environment, and carrying out dynamic reward shaping. A target strategy model optimized through simulation training is generated, the target strategy model is loaded to the robot, and the robot is controlled to execute task operation in the actual interaction scene. The method has the effect of realizing adaptive task learning of the robot in a multi-interaction scene.

The invention provides a virtual reality simulation training method, system and equipment for wind power plant operation and maintenance personnel and a medium. The virtual reality simulation training method comprises the following steps: step 1, constructing a virtual wind power plant; 2, realizing man-machine interaction between the user and the virtual wind power plant; 3, generating a virtual coach with professional knowledge and training experience, and guiding the operation of the user to realize simulation training of the operation and maintenance personnel of the wind power plant; according to the method, the application potential of emerging technologies such as virtual reality and artificial intelligence in the field of wind power operation and maintenance training is fully exerted, and the method has great theoretical significance and practical value in the aspects of improving industry talent quality, reducing training cost, promoting green and low-carbon development and the like.

The invention discloses a snake-shaped robot control method based on reinforcement learning. The snake-shaped robot control method comprises the steps that a complex terrain simulation scene comprising walls, cubic obstacles and low channels is built; modeling is conducted on the structure, joints and kinetic parameters of the snake-shaped robot, and a snake-shaped robot simulation model is obtained; defining a state space and an action space of the snakelike robot, and designing a diversified reward function; constructing an action network and a value network by adopting a multi-layer perceptron (MLP) structure; and performing multi-round simulation training on the snakelike robot by applying a reinforcement learning algorithm, so that the robot learns a parameter set for making an optimization decision under different scenes, thereby realizing optimization control of the snakelike robot and reaching a target point. According to the method, visual information does not need to be depended on, efficient and stable motion control of the snake-shaped robot in a complex environment can be achieved through autonomous learning, and the crossing capacity of the snake-shaped robot in a low and narrow space is improved.

The invention relates to the technical field of safety simulation training drilling, in particular to an intelligent assessment and evaluation method for mine accident emergency risk avoiding, which comprises the following steps of: S1, acquiring multi-dimensional data in an emergency risk avoiding capability drilling process of a student; s2, performing data preprocessing on the multi-dimensional data, and constructing an evaluation data set; s3, constructing an emergency risk avoiding capability evaluation system, wherein the system comprises evaluation items and distribution weights of the evaluation items; s4, establishing a weight evaluation model, and dynamically adjusting the distribution weight of the evaluation item in combination with the environmental data; and S5, scoring is completed according to the emergency risk avoiding capability evaluation system, and a personalized evaluation report is generated. According to the invention, comprehensive, objective and real-time evaluation of the emergency risk avoiding capability of the miner is realized, and the influence of human factors on the evaluation result can be reduced; the method can flexibly adapt to different environments and scenes through dynamic adjustment of the evaluation model, provides personalized evaluation reports and improvement suggestions, and effectively assists students in improving the emergency risk avoiding capability.

The invention relates to the technical field of robot control methods, in particular to a simulation-to-reality migration-oriented robot gait optimization method, which comprises the following steps of: acquiring input characteristics of a robot; constructing a gait control model by adopting a near-end strategy optimization reinforcement learning algorithm; training the gait control model in parallel in a physical simulation environment; the input feature matrix is input into the trained gait control model, a joint target position control instruction is generated, and the robot is controlled to execute gait actions; executing domain randomization operation by applying simulation to a reality checking tool; the simulation parameters are adjusted according to the difference parameters, and optimized gait control model parameters are output. The uncertainty of the real environment is covered by applying simulation to reality troubleshooting tool execution domain randomization operation, the systematized difference positioning process is combined, the simulation and real machine difference parameters are recognized, and the accuracy of the gait control model is improved. And action malformation, poor stability or hardware damage of the gait strategy of simulation training during real machine deployment can be prevented.

The invention discloses a data security policy automatic adjustment system and method. The system comprises a data acquisition module (DCM), a data analysis module (DAE), a security policy adjustment module (SPAM), a monitoring feedback module (MFM) and a training data module (TDR). The DCM collects key data such as user access records, network traffic and system logs. The DAE analyzes these data and generates security event alerts and risk assessments in combination with a risk knowledge base. And the SPAM formulates or updates security parameters according to the evaluation result, and generates and pushes a new security policy to the cloud environment. And the MFM monitors network flow and user behaviors, compares the network flow and the user behaviors with a preset baseline, and feeds back when abnormity is found. The TDR uses historical data to carry out scene simulation training, a data analysis model is optimized, and the safety performance of the system is improved.

The invention relates to the technical field of automatic driving, and discloses an explanatory enhancement method and device for an automatic driving scene, and the method comprises the steps: obtaining a driving scene label and sensor performance data corresponding to the driving scene label, and determining a plurality of weight combinations based on the driving scene label and the sensor performance data; performing experimental simulation training based on the sensor performance data and the plurality of weight combinations to generate experimental record data; performing anti-fact simulation on the experimental record data to generate anti-fact simulation record data; and finally, based on the experimental record data and the anti-fact simulation record data, performing causal model verification optimization to obtain a sensor weight distribution strategy in the automatic driving scene. According to the method, dynamic evaluation and optimization of sensor performance data in different driving scenes are realized, the problem of'black box 'in an automatic driving system is solved, and interpretability and transparency of a sensor fusion process are realized.

The invention discloses a quadruped robot motion control method based on adaptive deep reinforcement learning. The method comprises the following steps: S1, determining a network model, a composite reward function, a state space and a bionic action generation mechanism of a simulation training environment; the state space provides environment information input, the network model processes the input information and generates a decision, the bionic action mechanism executes a specific decision behavior, and the composite reward function evaluates a behavior effect and optimizes a decision direction; s2, constructing a simulation training environment of the quadruped robot, wherein the simulation environment comprises quadruped robot model information and simulation environment information; s3, training the network model by using a deep reinforcement learning algorithm based on robot model information and simulated environment information to obtain a trained motion control strategy; and S4, verifying the feasibility of utilizing the trained motion control strategy by controlling the motion of the quadruped robot in a real environment. According to the invention, the self-adaptive capability to a complex environment is obviously improved.

The invention belongs to the technical field of new energy automobile teaching, and particularly relates to a VR interactive new energy automobile high-voltage electrical principle teaching method, system and equipment, and the method comprises the steps: obtaining the physical parameter data and dynamic circuit simulation data of a new energy automobile high-voltage electrical system; constructing a high-precision 3D model according to the physical parameter data, inputting dynamic circuit simulation data into the model, and generating interactive principle annotations and particle effect animations; based on a gesture recognition model and a new energy automobile VR interaction technology, circuit state changes in the particle effect animation are obtained, a new energy automobile high-voltage system principle and a signal flow are displayed, a practical operation simulation training scene is constructed, student operation data are recorded according to the practical operation simulation training scene, an AI evaluation model is trained, and an evaluation report is obtained. And adjusting interaction feedback parameters of the training scene, and meanwhile, generating a personalized learning report of the trainee. Therefore, the problems of safety risk, teaching lag, abstraction obstacle and the like in the prior art are solved.

The invention discloses a dexterous palm internal operation control method based on reinforcement learning and real-time pose feedback, and the method comprises the steps: building an under-actuated dexterous palm internal operation experiment scene, and carrying out the data collection in a simulation environment; aiming at an operation task in the under-actuated dexterous palm, building a reinforcement learning algorithm framework, and designing a refined state space, an action space and a reward function according to task requirements and structural characteristics of the under-actuated dexterous palm; an IsaacLab environment is configured to realize efficient interaction between a reinforcement learning algorithm and a physical simulation platform; simulation training is carried out in the environment, and an optimal strategy is learned; different task scenes are set, pure state input driving decisions are adopted, and the autonomous adaptability and generalization control ability of the under-actuated dexterous hand are improved. The problems that a current robot underactuated dexterous hand is insufficient in control flexibility in a complex environment and is difficult to adapt to an unstructured environment, and reinforcement learning training results are difficult to migrate from simulation to reality can be solved.

The invention relates to the field of electroencephalogram signal processing, and particularly discloses a multi-mode electroencephalogram characteristic cognitive ability evaluation system based on a neural network, comprising: an acquisition module used for synchronously acquiring whole brain signals and eye movement data during pilot simulation training; the data preprocessing module is used for primarily processing the received whole brain signals and the eye movement data, monitoring the primarily processed data in real time by utilizing a preset monitoring mechanism, and triggering to generate interaction test starting signals if preset symbolic fluctuation occurs in the whole brain signals or the eye movement data; and the test interaction sub-module is used for generating arithmetic test questions according to a preset rule after receiving the interaction test starting signal, and synchronously recording answer correlation parameters of the pilot. By adopting the technical scheme of the invention, the deep-level reaction mechanism of the brain under the stimulation of the special complex cognitive task of flight training can be fully considered, and the cognitive load change of the pilot in the simulated training can be comprehensively and accurately reflected.

The invention discloses an AI-based network security risk identification and management and control system, which is characterized by comprising network security equipment and a server, the network security equipment is used for monitoring flow data flowing into / out of the server, and the AI-based network security risk identification and management and control system also comprises an interface conversion module which is used for connecting different network security equipment, original operation logs of different network security devices are collected; data collection: collecting multi-source data from multiple channels of network traffic, server logs, application program logs and user behavior records; preprocessing the data; preprocessing the data, removing noise data, filling missing values, and extracting useful features from the original data; simulation training: training the model to identify normal and abnormal behaviors through machine learning by utilizing supervised learning, unsupervised learning or semi-supervised learning methods; the deep neural network is used for modeling a complex data mode, and the detection precision is improved.

Methods for co-designing transformer-accelerator pairs are provided. The methods may include using a transformer embedding to generate a computational graph and a transformer model. The methods may include running the computational graph through a surrogate model and outputting accuracy data of the surrogate model. The methods may include using an accelerator embedding and the transformer model to simulate training and inference tasks and outputting hardware performance data of the transformer model. The methods may include sending the hardware performance data (such as latency, energy leakage, dynamic energy, and chip area, which may be optimizable performance parameters) and model accuracy data to a co-design optimizer. The methods may include generating an output transformer-accelerator or a transformer-edge-device pair from the co-design optimizer. The transformer model and accelerator embedding may be the output transformer-accelerator or a transformer-edge-device pair.

The invention discloses a cutter wear state on-line monitoring method based on a physically guided deep learning network, which combines a physical law and a deep learning technology in a cutter wear degradation process, and comprises the following steps: 1, collecting processing experiment data and constructing an actual training data set; 2, the physical model expands training data samples and constructs a simulation training data set; 3, adding a periodic feature extraction unit, a monotone increasing unit and a physical guidance module into the deep learning network, and constructing a physical guidance deep learning tool wear state online monitoring network; 4, constructing a physically guided deep learning loss function; and 5, pre-training the model by using simulation experiment data, and then performing fine tuning optimization on the model by using a small amount of actual data to obtain an optimal tool wear state monitoring network so as to predict the current tool wear. The method is high in adaptability, good in robustness and capable of accurately monitoring the state of the cutter in a complex and changeable machining environment.

The invention discloses a simulation training method and system for a minimally invasive surgery in the obstetrics and gynecology department, and relates to the technical field of virtual reality, and the method comprises the following steps: obtaining related medical data on a patient in the obstetrics and gynecology department, and carrying out the preprocessing of a medical image; the processed two-dimensional image is converted into a three-dimensional model, an operation scene is constructed, and the anatomical model of the patient in the obstetrics and gynecology department and the scene are fused; generating a plurality of operation planning schemes, calculating an operation difficulty coefficient Nd of each operation scheme, and determining an optimal operation scheme according to the size of the operation difficulty coefficient Nd; establishing an interactive interface between the doctor and the virtual operation scene, and recording various indexes of the doctor in the simulated operation process; after the simulated operation is finished, the operation is analyzed, an operation analysis report is generated, and the simulated operation is evaluated and optimized; through virtual reality minimally invasive surgery simulation, the misoperation risk in a real surgery can be reduced, so that the operation error rate in the surgery is reduced.

The invention relates to the technical field of flight simulation training, in particular to an air traffic control response system based on a flight simulator and a response training method. The method comprises the steps that a voice input module receives input voice audio data and sends the voice audio data to a voice recognition module; the voice recognition module recognizes and converts the voice audio data into voice text data; the instruction understanding module processes the voice text data to generate structured data corresponding to air-ground communication and sends the structured data to the instruction text generation module; the instruction text generation module generates a corresponding air traffic control instruction text according to the structured data; the voice synthesis module converts the air traffic control instruction text into corresponding voice audio data and sends the voice audio data to the voice output module; and the voice output module outputs the generated voice audio data. According to the invention, the recitation condition of the student can be intelligently evaluated, the workload of a teacher is effectively reduced, and the teacher can concentrate on guiding the flight operation of the student.

The invention discloses a flight space directional training auxiliary system and method, and the system comprises a flight task simulation module; the meteorological environment simulation module is used for simulating a meteorological scene corresponding to a flight task, and simulating the meteorological scene by establishing a multi-dimensional meteorological model and combining meteorological parameters; the meteorological environment parameter control module comprises a time step length control module which adopts a preset time step length to input meteorological parameter observation data collected in real time to the meteorological parameter setting module; the specific meteorological event control module is used for transmitting meteorological parameters of a specific meteorological event area to the meteorological parameter setting module when the airplane triggers a specific meteorological event; and the training scheme generation module is used for generating a proper training scheme according to the historical training information of the flight personnel. According to the system and the method, a real and personalized simulation training environment can be provided for the flight personnel, and the professional skills and the capability of coping with various flight conditions of the flight personnel are effectively improved.

The invention provides an automatic history fitting method for gas drive numerical reservoir simulation, which comprises the following steps of: 1, establishing a gas drive numerical simulation mathematical model, and correcting the gas drive numerical simulation mathematical model by using a gas drive PVT phase state fitting experiment; 2, solving the corrected gas drive numerical simulation mathematical model obtained in the step 1, performing parameter analysis, and screening main control factors influencing accumulated oil production; 3, generating a numerical simulation training sample set based on an orthogonal experiment, and obtaining the main control factor parameter initial value constraint screened out in the step 2 through a machine learning algorithm; step 4, establishing a multi-parameter fitting objective function for realizing historical fitting; and step 5, using the main control factor parameter initial value constraint obtained in the step 3, performing iterative calculation through a projection gradient method, fitting the target function established in the step 4, and terminating when a convergence condition is satisfied. According to the method provided by the invention, the constraint parameter values can be adaptively calculated, automatic historical fitting is completed, and the definition of the fitting parameter range does not depend on empirical and experimental determination any more.

The invention provides a scene reconstruction method for an automatic driving simulation test, electronic equipment and a product, and relates to the technical field of computer vision. According to the specific implementation scheme, semantic segmentation is carried out on original image data, and long-shot image information and close-shot image information are determined; wherein the original image data comprises multiple frames of continuous images; constructing a texture map based on the long-shot image information, and sampling in the texture map to obtain a long-shot image; generating a close-shot spot cloud corresponding to the close-shot image information, and performing three-dimensional Gaussian modeling on each point in the close-shot spot cloud to obtain a close-shot three-dimensional Gaussian ellipsoid set; and scene reconstruction is carried out based on the close-shot three-dimensional Gaussian ellipsoid set and the long-shot image to obtain a multi-frame reconstruction scene graph in sequential arrangement. According to the technical scheme of the invention, the method can improve the estimation accuracy of the close-range depth value, achieves the more accurate reconstruction of a scene, and improves the simulation training effect of an automatic driving simulation system.

The invention discloses an interactive control method for virtual simulation training of electric power operation and maintenance, and particularly relates to the technical field of electric power operation and maintenance simulation, which comprises the following steps of: acquiring operation behavior data of students, performing real-time capability grading on the students by utilizing a multi-dimensional capability evaluation model, and performing real-time capability grading on the students; a difference adaptation parameter model is constructed in combination with the difference between the virtual device and the target actual device, training content with corresponding difficulty is matched from a preset training task library, a knowledge graph of students is constructed, and a personalized training path is generated through a path recommendation engine. Iteratively updating the training difficulty and the recommended path according to the newly generated behavior data and the performance of the trainee in the aspect of the difference adaptive capacity, constructing a difference adaptive parameter model, bringing the difference into a training task adjustment and capacity evaluation mechanism, dynamically adjusting the training task parameters to fit the actual equipment complexity, and evaluating the performance of the trainee in the aspect of the difference adaptive capacity. And a migration barrier between virtual training and real operation and maintenance is broken through.