2919 results about "Model parameter" patented technology

The invention discloses a CAD automatic generation system and method based on intelligent model selection and application, and aims at achieving automatic modeling under the multi-modal design requirement. The system comprises a user interaction module for receiving multi-modal input such as natural language, sketch and voice; the intelligent demand analysis module is used for combining an industrial large language model and a product knowledge graph, combining semantic analysis and generating a structured demand; the intelligent model selection calculation module is used for matching the optimal parameter combination and the component list based on a multi-objective optimization algorithm; the CAD automatic generation module calls a parametric modeling engine to generate an editable three-dimensional model; the constraint solving module is used for processing hard constraints and soft constraints in real time and dynamically adjusting model parameters; and the model output and interaction module feeds back a design state and supports user iteration. The system realizes full-process automation from the design intention to the CAD model, improves the design efficiency and accuracy, and is suitable for the fields of mechanical design, intelligent manufacturing and the like.

The invention discloses an image processing method and device and a storage medium. The method comprises the steps of obtaining a to-be-simulated 3D convolution model and training data; decomposing the 3D convolution model into cascading of a 3D space convolution model and a 3D time convolution model to obtain a pseudo 3D cascading convolution model; training a pseudo 3D cascade convolution modelby using the training data, and obtaining parameters of a 3D spatial convolution model and a 3D time convolution model; converting the 3D space convolution model and the 3D time convolution model intoa 2D space convolution model and a 2D time convolution model; setting a feature rearrangement rule for the 2D spatial convolution model and the 2D time convolution model; mapping model parameters ofthe 3D spatial convolution model and the 3D time convolution model into parameters of a 2D spatial convolution model and a 2D time convolution model to obtain a 2D cascaded convolution model; and performing convolution operation on the image by using the 2D spatial convolution model and the 2D time convolution model. By means of the mode, image processing conducted through 3D convolution operationcan be achieved through the 2D convolution model.

A method and device for optimizing comprehensive production of an oil reservoir. The method comprises: on the basis of historical production data, performing regional level recognition, single-well level recognition, and production optimization model recognition, and updating oil reservoir dynamic recognition; on the basis of the historical production data, using the following steps to perform cyclic historical fitting on model parameters of a plurality of production optimization models: performing historical fitting on the model parameters, and performing well-to-well communication relationship calibration on the plurality of production optimization models; determining, on the basis of the updated oil reservoir dynamic recognition, a constraint condition corresponding to each production optimization model, solving an objective function corresponding to each production optimization model, obtaining an optimal decision variable corresponding to each production optimization model, and obtaining a development regulation scheme corresponding to each production optimization model; obtaining a comprehensive development regulation scheme of an oil reservoir; and updating the oil reservoir dynamic recognition on the basis of a regulation effect obtained by real-time monitoring of a development operation.

The invention relates to the technical field of oil and gas field development, and discloses an ultra-large type true triaxial hydraulic fracturing fracture evolution path prediction method and system.The prediction method comprises the steps that a coupling geomechanical model integrating microscopic, macroscopic and wellbore flow scales is constructed; initializing the model and performing crack initial expansion simulation; collecting construction data in real time, and dynamically optimizing model parameters through a data assimilation algorithm; performing fracture evolution advanced prediction by using the updated model; and generating an optimization decision based on the prediction result and feeding back to the construction site. According to the method, fusion of a multi-scale physical mechanism and real-time dynamic prediction is realized, and accurate prediction and active control can be performed on crack expansion under the ultra-large true triaxial condition.

The invention relates to a power distribution network data intelligent analysis method based on data consanguinity and multi-modal fusion learning. The method comprises the following steps: S1, constructing a dynamically evolved data consanguinity topological graph; s2, designing a label-guided graph neural network architecture, embedding historical abnormal knowledge into a graph learning process, and outputting a deep semantic feature vector; s3, constructing a multi-modal fusion analysis framework, performing multi-dimensional feature fusion and data quality analysis, and identifying abnormal nodes; s4, designing a semi-supervised and incremental learning combined mixed training normal form, and performing model training and strategy optimization; and S5, based on the dynamic consanguinity topology constructed in the step S1 and the identified abnormal nodes, constructing a probabilistic reasoning framework, and fusing the model parameters obtained by optimization in the step S4 to realize quality abnormality root positioning and full-link visualization so as to form a complete data intelligent analysis scheme. According to the invention, efficient and accurate management of the topological data quality of the power distribution network is realized.

The invention discloses a CAD automatic modeling method and system based on parameter driving. The method comprises the steps that model parameter information needing modeling is obtained; generating a CAD (Computer Aided Design) model on the basis of parameter analysis, geometric construction and surface recognition schemes; performing reverse reconstruction of the three-dimensional CAD model based on topology analysis, geometric feature extraction, pattern recognition and parameter relation inference; modeling, optimization and feedback evolution are carried out based on model analysis, feature learning and generation of parametric modeling rules; and according to the obtained data information, CAD automatic modeling based on parameter driving is completed. According to the method, CAD automatic modeling based on parameter driving is achieved, the reliability is higher, the accuracy is better, and the efficiency is higher.

The invention provides a multi-modal pre-training model construction method and system for monitoring videos, and the method comprises the steps: automatically constructing a high-quality multi-modal alignment data set through a single-modal description generation model and a large-scale language model, and remarkably reducing the marking cost; a special coding network and a shared projection layer are adopted to realize feature extraction and uniform semantic space alignment of video, audio and text modes; performing dynamic semantic fusion by using a modal collaborative attention mechanism; designing cross-modal contrast learning, mask prediction and time sequence consistency tasks to carry out multi-task pre-training; an external knowledge base is introduced, and the semantic reasoning ability is enhanced through a microretrieval mechanism; and optimizing model parameters by adopting a multi-task joint loss function and an end-to-end training strategy. According to the method, efficient and automatic construction, deep semantic alignment and fusion and intelligent reasoning of knowledge enhancement of monitoring video multi-modal data are realized, and the understanding and generalization ability of the model in a complex scene is effectively improved.

The invention discloses a live-action three-dimensional dynamic change modeling method based on digital twinning. The method comprises the following steps: collecting multi-source data; carrying out space registration and time synchronization processing, and outputting a fusion data set; carrying out three-dimensional modeling, optimizing model parameters, and generating a live-action three-dimensional model; virtual-real mapping is carried out, and generated virtual space data and state parameters thereof are stored; performing differential analysis to generate a change detection result data set; performing dynamic change modeling to generate an updated virtual model; executing geometric accuracy optimization and semantic consistency check, and outputting an optimized virtual model; bidirectional data flow association and dynamic updating are executed, and the live-action three-dimensional dynamic change modeling process is completed. The real scene dynamic updating is realized by adopting the improved Gaussian sputtering modeling, and the method has the advantages of high precision, strong real-time performance and virtual-real synchronization.

The invention discloses a high-precision static aeroelastic model optimization design method based on a model correction technology, and relates to the technical field of aircraft design, and the method comprises the following steps: S1, firstly constructing an initial model, and carrying out statics pre-analysis to verify integrity; s2, executing SOL 101 statics analysis based on the initial model and outputting a physical field result; s3, carrying out consistency analysis in combination with test data and generating a correction decision; s4, screening high-priority correction parameters through local or global sensitivity analysis; s5, correcting model parameters by adopting a mixed algorithm of a gradient method and an agent model, and verifying precision and generalization ability; s6, the corrected model is output as a Nastran file and a reduced-order model in a standardized mode, and a parameter change log is recorded; s7, executing static aeroelastic coupling and flutter analysis, and feeding back a result to drive optimization iteration; s8, constructing a multidisciplinary coupling optimization model in combination with aeroelastic and flutter results to realize collaborative optimization; and S9, finally performing engineering standardization packaging on the optimization model and outputting a verification report.

The invention belongs to the crossing field of environmental engineering and hydrology and hydrodynamics, and particularly relates to a traceability method based on coupling hydrodynamics and a pollutant degradation equation, which comprises the following steps: firstly, acquiring and preprocessing multi-source heterogeneous monitoring data, then selecting a one-dimensional Saint-Venant equation or a two-dimensional shallow water equation according to a water body form to solve a hydrodynamic field, and finally, determining the hydrodynamic field. A multi-component convection-diffusion-reaction equation is coupled to simulate pollutant migration and transformation; an LSTM module is introduced to identify suspected pollution events, pollution source parameters are inverted through a two-channel framework of PDE constraint optimization and Bayesian inference, uncertainty is quantified, model parameters are updated online in combination with data assimilation, and finally the uncertainty is quantified and verified. The method considers traceability precision, efficiency and compliance, supports multiple water bodies and multiple data sources, and is suitable for complex water body pollution traceability.

The invention discloses an industrial AI assistant cross-modal interaction method based on a dynamic knowledge graph, and the method comprises the following steps: collecting texts, voices, images and multi-source sensor data in an industrial system, carrying out the modal recognition, feature extraction and time alignment, and generating an event feature set and a state feature set with timestamps. Modeling a time dependency relationship between event types by constructing a multivariable Hawkes model, and outputting an event trigger sequence and trigger strength; and in combination with a neural controlled differential equation model, guiding the state to evolve along with time and jump at a specific moment to form a state evolution trajectory. Performing fusion coding on the event and the state, constructing a dynamic knowledge graph with a causal structure and semantic continuity, and generating interactive output based on context reasoning; and after system feedback is received, the triggering strength and the state track are updated, and continuous evolution of the knowledge graph and reverse optimization of model parameters are achieved.

The invention discloses a video monitoring abnormal behavior real-time detection method based on a graph neural network, and the method comprises the following steps: collecting a video frame sequence, extracting a detection frame, a key point and an optical flow feature, generating a node feature matrix, and constructing a dynamic graph structure; establishing a dynamic graph neural network model based on EvolveGCN, and updating a convolution weight by using a gating circulation unit; calculating event intensity and change rate according to the motion abrupt change signal, generating a time delay parameter and adjusting a weight modeling step length; performing low-rank decomposition and spectral radius projection on the convolution weight matrix, and adjusting a spectral constraint threshold according to an abnormal score; inputting a weight matrix to generate graph branch and hypergraph branch embedded representation; exchanging topology correction information based on a mutual generation mechanism and updating model parameters; and inputting the dynamic graph structure and the node feature matrix in real-time reasoning, calculating an abnormal score and outputting a detection result. According to the invention, adaptive evolution and high-precision anomaly detection of dynamic graph modeling are realized.

The invention provides a magnesium alloy electric drive main shell mold machining precision control method and system, and belongs to the field of precision numerical control machining. The method comprises the steps that dynamic milling force, key point temperature and position signals of all shafts are synchronously collected through a multi-source sensor; a recursive least square method is adopted to identify the dynamic rigidity and damping coefficient of the tool-workpiece system on line, and the thermal deformation drift distance is calculated in combination with a temperature signal; predicting a three-dimensional deformation error vector in real time based on the dynamic model, thermal deformation and geometric errors, and constructing a space compensation field; macro motion compensation is achieved by modifying numerical control program coordinate points, and meanwhile high-frequency micro-amplitude compensation is conducted through a spindle tail end fast tool servo system. And establishing an adaptive disturbance observer to perform online estimation on unmodeled disturbance and feed-forward compensation, and realizing online self-tuning of model parameters in combination with in-situ measurement feedback. Real-time sensing, full-band dynamic compensation and closed-loop self-adaptive control of multi-source errors are achieved, and the machining precision and long-term stability of the mold are remarkably improved.

The invention discloses an oil reservoir comprehensive production optimization method and device, and the method comprises the steps: carrying out the region level recognition, single well level recognition and production optimization model recognition according to the production historical data, and updating the oil reservoir dynamic recognition; based on production historical data, performing cyclic historical fitting on model parameters of the various production optimization models by adopting the following steps: performing historical fitting on the model parameters, and performing inter-well communication relation calibration on the various production optimization models; determining a constraint condition corresponding to each production optimization model according to the updated reservoir dynamic cognition, solving a target function corresponding to each production optimization model, obtaining an optimal decision variable corresponding to each production optimization model, and obtaining a development regulation scheme corresponding to each production optimization model; an oil reservoir comprehensive development regulation and control scheme is obtained; and according to the regulation and control effect obtained through real-time monitoring of the development operation, oil reservoir dynamic understanding is updated. According to the invention, the efficiency and precision of oil reservoir production optimization can be improved.

The invention discloses an environmental data processing method and system based on ocean engineering, and relates to the technical field of data processing, and the method comprises the steps: receiving an original observation data flow through a multi-source data preprocessing module, and carrying out the dynamic noise filtering and abnormal value adaptive detection; fusing the multi-source heterogeneous data through a multi-scale data fusion module, and embedding the fused multi-source heterogeneous data into a marine kinetic equation as a soft constraint; non-linear evolution features are extracted from the fusion data through a feature extraction and state representation module, and a high-dimensional environment state vector is constructed; real-time prediction of model parameters is executed through online learning and an inference engine; and performing uncertainty propagation calculation on the processing flow through a confidence evaluation module and generating a final environment state report. According to the method, the adaptive capacity of data preprocessing can be remarkably improved, the physical consistency of multi-source data fusion is improved, the nonlinear evolution law of ocean phenomena is accurately captured, and continuous online optimization and edge side low-delay response of model parameters are achieved.

The invention discloses a small sample self-learning accurate identification method based on distillation knowledge migration. The method comprises the following steps: S1, extracting deep semantic features of a source domain and shallow features of a small number of samples of a target domain, and calculating a mapping matrix; s2, calculating an entropy difference distillation excitation function based on the initial alignment features; s3, executing domain knowledge distillation and generating staged distillation representation; s4, constructing a composite fitness function and initializing a parameter population; s5, performing iterative optimization by adopting a variable step size dynamic feedback compression strategy; s6, loading the optimal parameters and performing coupling alignment with the historical distillation representation; and S7, performing combined fine adjustment on the distillation weight and the model parameters through self-learning feedback. According to the method, through adaptive knowledge distillation and dynamic optimization feedback closed loop, high-precision and adaptive identification under extremely few labeled samples is realized, the generalization ability of the model is remarkably improved, and overfitting is effectively inhibited.

The invention relates to the technical field of industrial automation control, and discloses a water-based paint production and processing equipment control system, which comprises the following steps: acquiring stator current and rotor angular velocity, and reconstructing total load torque by using a sliding mode algorithm; extracting an input power low-frequency component to construct a heat dissipation slow manifold, and calculating a thermal drift disturbance component and an orthogonal residual component through orthogonal projection; on the basis of a two-component generation superposition instruction closed-loop adjustment driving mechanism, a double-time-scale orthogonal projection mechanism is utilized, non-structural thermal drift interference is automatically stripped under the condition that heat transfer parameters of equipment do not need to be predicted, the problem of control divergence caused by model parameter mismatch is solved, and decoupling control over the real rheological state of materials is achieved.

The invention relates to an intelligent traffic flow prediction analysis method based on artificial intelligence, and the method comprises the steps: collecting and fusing traffic flow, environmental factors and event information according to traffic levels, and achieving the standardization and automatic clustering preprocessing of multi-level space-time attributes through regional factor labels; and then, expressing a multi-dimensional structure and a dynamic attribute of each node by using regional factor vectorization, dynamically modeling a spatial node heterogeneous adjacency relationship in combination with a self-organizing graph neural network, introducing a cross-level dynamic attention mechanism to perform weighted fusion on multiple spatial and temporal features, and outputting multi-granularity traffic prediction through a hierarchical fusion decoding network. And the model is combined with actual feedback to realize self-adaptive optimization of the area factors and model parameters. The method has the advantages that high-precision prediction of the traffic flow under multiple scales of roads, blocks, cities and the like is achieved, the self-learning and self-adaptive capacity for heterogeneous information, emergencies and spatial dynamic changes is improved, and hierarchical decision making and flow management are supported.

The invention discloses a bearing variable working condition fault diagnosis method fusing model migration and feature migration learning, and the method comprises the steps: processing bearing vibration signals of a source domain and a target domain through wavelet transform, and extracting a time-frequency diagram; expanding the two-dimensional time-frequency graph data set by using DCGAN, and balancing the number of the two-dimensional time-frequency graph data set; then, model parameter migration is adopted, AlexNet network parameters pre-trained in a source domain are migrated, a migrated AlexNet network is constructed, and depth features are extracted; then, a domain adaptation method based on improved migration joint matching is provided, multiple strategies are fused, and a low-dimensional feature space with small distribution difference and good discrimination performance is obtained; and finally, on the basis of a labeled source domain feature data training model after domain adaptation, realizing identification and classification of unlabeled target domain feature data. The method is ideal in diagnosis performance and high in accuracy under variable working conditions and data imbalance, domain data distribution difference can be reduced by improving the migration joint matching method, and feature discrimination performance and fault diagnosis accuracy are improved.

The invention discloses a non-independent identically distributed data asynchronous federal learning method based on an improved aggregation algorithm. The method comprises the steps that a server initializes a global model and issues the global model to all clients; and the client performs local training on the received global model by using local data, and uploads the model and model parameters to the server after training is completed. Then, the server adjusts a model lag degree based on a client data volume proportion, calculates model difference consistency, client historical contribution stability, old degree penalty of the client model and cosine similarity of the client model and the global model based on parameters of the client model and the current global model, and generates an asynchronous federal aggregation factor accordingly; and updating the global model parameters to generate a new global model. And finally, testing the global model by the server, and judging whether the learning process is stopped or not. According to the method, fair and effective model aggregation can be realized, and the model convergence stability and the final model detection precision are improved.

The invention provides a multi-step automatic testing method based on machine vision, and the method comprises the steps: capturing an original image sequence of a test scene through a camera, extracting the contour features of a test object in an image through an edge detection algorithm, and obtaining a preliminary positioning coordinate; if the dynamic position change trend exceeds a preset threshold value, adjusting an image enhancement parameter to suppress background noise, and obtaining an enhanced target image; feature matching is carried out through the enhanced target image, a robust identifier such as a texture mode is extracted, and an accurate three-dimensional position coordinate is obtained; according to the accurate three-dimensional position coordinates, calculating an execution deviation value of the current step, and judging whether the deviation value is within an allowable range or not; if the deviation value is within the allowable range, a control instruction sequence is generated and transmitted to the mechanical arm, and a step execution confirmation signal is obtained; comparing a confirmation signal with a next image sequence through the steps, updating positioning model parameters, and determining a continuous adjustment scheme of the whole test process.

The invention discloses a brain multi-modal multi-sequence data registration method and device based on deep learning. The method comprises the steps of performing first iteration processing on a target image and a first moving image to obtain a first deformation field, wherein the first iteration processing comprises image alignment constraint processing of the target image, smooth constraint processing of the first deformation field, and area alignment constraint processing of a tumor area in the image; registering the first moving image to the target image based on the first deformation field to obtain a second moving image; performing second iteration processing on the target image and the second moving image to obtain a second deformation field; and registering the second moving image to the target image based on the second deformation field to obtain a registered moving image. In the process of performing unsupervised training on the deep learning model, after multiple times of iterative processing, model parameters are updated based on image alignment constraint loss with a target image, deformation field smooth constraint loss and region alignment constraint loss of a tumor region in the image.

The invention relates to the technical field of data center dynamic modeling, and discloses a twin model dynamic modeling method for a data center DCIM platform, which comprises the following steps: constructing a discrete state space model containing a thermal coupling matrix and a system matrix, collecting real-time power and temperature time sequence data, and calculating a cross-correlation function to lock hot air dynamic transmission lag time; calculating cut-off frequency based on physical attributes of the cabinet and decomposing data into high and low frequency components by using a complementary filter; according to the method, the model parameters are made to return to a physical source through a frequency domain decoupling mechanism, the problem of aliasing of airflow coupling and structural thermal inertia parameters in a traditional single-scale identification method is solved, and the method is suitable for large-scale identification. And the physical authenticity and prediction robustness of the twin model under a complex working condition are improved.

A method for prioritizing predictive model data streams includes receiving, by a first device, a plurality of predictive model data streams. Each predictive model data stream includes a set of model parameters for a corresponding predictive model. Each predictive model is trained to predict future data values of a data source. The method includes prioritizing, by the first device, priorities to each of the plurality of predictive model data streams. The method includes selecting at least one of the predictive model data streams based on a corresponding priority. The method includes parameterizing, by the first device, a predictive model using the set of model parameters included in the selected predictive model data stream. The method includes predicting, by the first device, future data values of the data source using the parameterized predictive model.

The invention discloses an oral cavity data monitoring and early warning method and system based on deep learning, and the method comprises the steps: solving a problem that the focus recognition is inaccurate because the image collection process of an oral cavity endoscope of a patient is affected by various environment and equipment parameters; according to the method, synchronous binding of collected images and equipment parameters and intelligent image preprocessing are proposed, historical time sequence images and variation trend features are combined, lesion space and time features are extracted through combination of a lightweight convolutional neural network and a long-short term memory network, feature weighted reasoning is realized by using a context awareness attention fusion network, and the focus space and time features are extracted. According to the method, the model parameters and the risk threshold value can be dynamically adjusted, the focus recognition accuracy and the early warning individuation are improved, and clinical grading treatment and intelligent health management are effectively assisted.

The invention discloses a distribution-micro collaborative operation simulation optimization method, which belongs to the technical field of simulation optimization, and comprises the steps of preprocessing grid-connected point voltage data, tie line power data and communication time delay data, constructing a power distribution network power flow physical network following a Kirchhoff's law, outputting a source load power prediction curve by using a long short-term memory network algorithm, and calculating the distribution-micro collaborative operation according to the source load power prediction curve. And a distribution-micro collaborative simulation optimization model is obtained based on residual error rolling correction tie line impedance parameters, a delay penalty term is set in a target function in combination with the preprocessed communication delay data, a power regulation instruction is obtained by using a particle swarm optimization algorithm, and a dynamic simulation video stream is generated. According to the invention, through rolling correction of the tie line impedance parameters and setting of the delay penalty term positively correlated with the time delay, the problem of control failure caused by physical deviation caused by model parameter solidification and communication time delay accumulation is solved, and the defects of voltage deviation calculation distortion and inaccurate network loss evaluation are eliminated. And the simulation precision and the operation stability of distribution-micro cooperation are improved.

The invention relates to the technical field of power dispatching, and discloses a knowledge distillation-based low-resource power big language big model training method, system and device and a medium, and the method comprises the steps: obtaining accident case data of the power industry, carrying out the problem construction and task setting, introducing a quality evaluation mechanism, carrying out the refusal sampling through a language model, and carrying out the training of a low-resource power big language big model. Generating a distillation data set for model distillation; introducing a LoRA module into the student model for fine tuning, constructing multi-source heterogeneous fine tuning data, and setting a training strategy to optimize the performance of the model; and performing training by adopting reinforcement learning, introducing language consistency rewards until the reinforcement learning achieves convergence on the reasoning task, and generating a final language large model. According to the method, through knowledge distillation and reinforcement learning, the deep knowledge and the reasoning ability of the super-large model are successfully migrated to the small model, so that the parameter quantity of the finally deployed model is greatly reduced, and the computing power resource required by reasoning is sharply reduced.

The invention relates to a building structure health monitoring method and system based on machine learning, and the method specifically comprises the following steps: firstly, building a target building three-dimensional numerical model through finite element simulation, generating a simulation signal, injecting Gaussian white noise, and adjusting model parameters to form a data set with health category labels; performing data enhancement by combining adaptive wavelet denoising with dynamic normalization, and extracting and enhancing high-resolution time-frequency features through adaptive window short-time Fourier transform and adaptive frequency band enhancement; then, a neural network model fusing structure physical prior guidance and multi-scale space-time interaction is constructed, a feature matrix is modulated, fused and coded to obtain a refined feature vector, and damage state probability distribution output is achieved; and training is carried out by using a feature consistency and prediction smoothness regularization term constraint model, and finally, the trained model is deployed, so that building structure health state evaluation and safety early warning are realized, the monitoring accuracy and reliability are improved, and effective technical support is provided for building safety guarantee.

The invention provides a structure perception multi-view city representation learning method and system with coordinated fusion and alignment, and a structure perception multi-view city representation learning model with coordinated fusion and alignment is established on the basis of target city POI and taxi travel data. The method further comprises the steps that multi-view data are constructed, original data are reconstructed through a sparse auto-encoder module, and specific representation of each view is extracted; executing an enhanced specific view representation module to enhance the expression ability of the specific view representation by modeling beneficial inter-view interaction; a cross-view converter module is adopted, and a multi-view fusion process is optimized by means of cross-view converter module region similarity; executing a structure perception multi-view contrast learning module, and enhancing the consistency between the consensus representation and the specific view representation; and optimizing model parameters, executing a soft Lagrange constraint-based training strategy, and solving the problem of a suboptimal solution caused by gradient conflicts in joint learning.

The invention discloses a complex network disintegration method based on evolution deep reinforcement learning. According to the method, an encoder-decoder model fusing a graph convolutional neural network and a deep Q network is constructed, and is used for efficiently extracting importance features of nodes in a complex network and realizing dynamic decision-making of a node disassembling sequence according to the importance features. In order to optimize model parameters and improve search capability, an evolutionary algorithm is introduced to perform global exploration on the model parameters, and the problem that a directional optimization strategy is easy to fall into local optimum is avoided. Meanwhile, deep mining is carried out on an evolution result in combination with a reinforcement learning strategy, the overall optimization process is accelerated, and advantage complementation of parameter evolution and strategy learning is achieved. Experimental results show that the method significantly improves the efficiency and precision of network disassembly while maintaining the robustness of the model, and has good practical value and wide application prospects.