16 results about "Pruning algorithm" patented technology

The application provides a multiple layout splitting method, system, medium and program product. First, an initial splitting coloring operation is performed based on a saturation priority strategy to generate an initial layout splitting scheme. Then, the global exploration ability of a tabu search algorithm is utilized, and a preset target function is combined to perform multiple rounds of iteration update operations on an initial conflict node set to generate a multiple layout splitting optimization scheme. For ultimate conflict nodes that cannot be completely eliminated by the tabu search algorithm, a backtracking pruning algorithm is started to maximize the attempt to eliminate all ultimate conflicts to generate a final multiple layout splitting target scheme. The application combines the initial optimization ability of the saturation priority strategy, the global exploration ability of the tabu search algorithm, and the quality guarantee of the backtracking pruning algorithm to realize accurate splitting of multiple layouts, guarantee splitting quality and lithography yield, improve splitting efficiency, reduce computational complexity, and adapt to the needs of multiple layout splitting.

This invention discloses a method for solving power system time-series optimization problems, comprising: constructing a dynamic mathematical model of the power system time-series problem and determining the objective function and constraints; generating an initial solution based on the dynamic mathematical model, performing pruning preprocessing on the initial solution using a fast pruning algorithm to select high-quality initial feasible solutions; generating neighborhood solutions based on the high-quality initial feasible solutions using a simulated annealing algorithm, determining whether to accept the neighborhood solutions according to the Metropolis criterion, and marking the completion of one iteration after the determination; performing pruning feedback and search direction optimization after each iteration until a preset iteration convergence condition is reached, and outputting the optimal solution for power system time-series optimization. This invention can effectively improve the efficiency, accuracy, and global optimality of solving power system time-series optimization problems, and is suitable for the complex requirements of high-voltage and high-efficiency power systems.

This invention provides a continuous subgraph matching method for social networks based on dynamic pruning, belonging to the field of social network behavior detection. The invention proposes a dynamic pruning algorithm that establishes local equivalence theory by calculating the backward neighbors of query points and candidate points, defines subtree-level equivalence criteria in dynamic graph environments, and is applicable to highly isomorphic behavior patterns such as zombie account groups and terminal nodes of propagation chains. The dynamic pruning algorithm skips isomorphic subtree structures to achieve the pruning effect. A query graph decomposition and sorting mechanism is proposed to remove the binding constraint of the matching order on the update edges and candidate sets, actively identifying and prioritizing highly redundant candidate set vertices, further reducing the search space. Ultimately, while ensuring result completeness, it achieves a 16.9-fold performance improvement over RapidFlow and a 100% completion rate for complex queries, providing sub-second decision support for abnormal behavior detection in social networks and promoting the application of dynamic graph analysis in resource-constrained environments such as edge computing.

The application provides a multiple layout splitting method, system, medium and program product. First, an initial splitting coloring operation is performed based on a saturation priority strategy to generate an initial layout splitting scheme. Then, the global exploration ability of a tabu search algorithm is utilized, and a preset target function is combined to perform multiple rounds of iteration update operations on an initial conflict node set to generate a multiple layout splitting optimization scheme. For ultimate conflict nodes that cannot be completely eliminated by the tabu search algorithm, a backtracking pruning algorithm is started to maximize the attempt to eliminate all ultimate conflicts to generate a final multiple layout splitting target scheme. The application combines the initial optimization ability of the saturation priority strategy, the global exploration ability of the tabu search algorithm, and the quality guarantee of the backtracking pruning algorithm to accurately split the multiple layout, guarantee the splitting quality and the lithography yield, improve the splitting efficiency, reduce the calculation complexity, and adapt to the demand for multiple layout splitting.

The application discloses a kind of lightweight collaborative detail enhancement small target detection method, belong to small target detection field, this method includes the improvement to YOLOv11n network model, obtains collaborative detail enhancement SDE-YOLO network model;The improvement includes the network connection structure between each C3K2 feature extraction module of Neck neck network of YOLOv11n network model is replaced as IGB multiscale collaborative optimization module;Each detection head is replaced as SDDH collaborative dynamic detail enhancement detection head;Using layer adaptive amplitude pruning algorithm LAMP to the collaborative detail enhancement SDE-YOLO network model of training completion is restructured, and the collaborative detail enhancement network model after reconstruction is obtained;The collaborative detail enhancement network model after reconstruction is deployed on unmanned aerial vehicle, and using scale adaptive non-maximum suppression algorithm ASD from detection result screening optimal small target detection result.The application solves the problem that the existing unmanned aerial vehicle-oriented method cannot balance target detection accuracy and computational overhead under limited computing power conditions.

This invention provides a method for analyzing negative Weibo comment data based on a cosine similarity-C4.5 decision tree, comprising: Step S1, preprocessing the original Weibo comment data, identifying whether the comments contain keywords to form a complete descriptive attribute set using the AC automaton algorithm, and labeling the influence category based on the number of comment interactions; Step S2, using the AC automaton algorithm to perform multi-keyword fast matching on the comment text; Step S3, based on the improved C4.5 algorithm, first merging attribute values ​​with similar information entropy using cosine similarity measurement, then calculating the information gain rate of each attribute and recursively constructing a decision tree; Step S4, using a pessimistic error pruning algorithm based on confidence interval estimation to perform post-pruning on the decision tree and extract classification rules. This invention improves retrieval efficiency through the AC automaton, reduces model redundancy through cosine similarity merging, and enhances generalization ability through pessimistic pruning. The three synergistically achieve a simultaneous improvement in classification accuracy, model simplicity, and processing efficiency.

This invention discloses a structure-preserving compression method for aero-engine blade defect detection model, belonging to the field of intelligent detection technology. It includes constructing a teacher model with linear deformable convolution and deformable attention mechanisms; employing a gradient-sensitivity-weighted layer-adaptive amplitude pruning algorithm to calculate the gradient sensitivity of weights to the defect focusing loss function, and combining this with a linear deformable convolution structure decoupling protection mechanism for pruning, generating a sparse student model architecture; constructing an initial student model based on this architecture; and training the model using a classification and regression branch co-distillation strategy to output a lightweight student model and deploy it on a detection terminal. This invention solves the technical challenge of deploying high-precision detection models on resource-constrained terminals, reducing the number of model parameters by 40% while maintaining over 96% mAP, and increasing the inference speed to 68.5 FPS, achieving real-time online high-precision detection of minute defects in aero-engine turbine blades.

The application relates to a model pruning method and device, computer equipment, a storage medium and a program product. The method comprises the following steps: acquiring an initial model set corresponding to a source model to be pruned, each initial model in the initial model set being obtained by pruning the source model by using each initial pruning algorithm; taking the initial model set as a current model set, performing iterative learning, for each iteration process, determining a target model and a reference model from the current model set, training the target model by using the reference model, and updating the current model set by using the trained target model until a target model set is obtained; and determining a model pruned from the source model based on the target model set and a screening condition. The method can improve pruning efficiency while ensuring the performance of the pruned model.

This invention discloses a self-optimizing control system for ice cream production parameters, relating to the field of production control technology. It includes: a working condition construction module, which collects process operating parameters of a continuous freezing section, divides the process into time periods according to different operating stages, and maps the process operating parameters of each time period to the current position of each actuator, forming a working condition page; a state classification module, which extracts freezing state features from the working condition page, classifies the freezing sections, and matches them with historical operation segments, forming a freezing state table and a list of reachable paths; and a sequence optimization module, which uses a working condition transition sequence optimization algorithm to progressively expand the adjustment parameters corresponding to the freezing state table and the list of reachable paths, and uses a sequence forward convergence pruning algorithm to filter out sequences that deviate from the target production zone and disrupt continuous output, forming candidate operation chains. This invention significantly reduces the effective search space and ensures the safety of the adjustment process through the sequence forward convergence pruning algorithm.

The application provides a kind of vehicle networking security audit method, system and storage medium based on feature decoupling and double-way mixed retrieval, method includes: step 1: receiving the natural language audit demand of user;Step 2: the query sentence is parsed using hybrid feature extractor, hard features are extracted from the parsed results by regular engine, and soft features are extracted from the parsed results by semantic classification;Step 3: double-way parallel retrieval strategy of vector retrieval channel and atlas retrieval channel is adopted;Step 4, correlation joint evaluation: the fact context formed by vector retrieval channel and the rule context formed by atlas retrieval channel are correlated and evaluated;Step 5: convert JSON topology into natural language linear sequence, and use adaptive pruning algorithm for dynamic pruning;Step 6: guide large model to execute logical judgment according to inference template, and generate structured security audit response report.The beneficial effect is: the proportion of effective information input into the model is improved.

The application discloses a power distribution network fault point positioning method and system fusing a CNN-LSTM-KAN network and graph theory, and belongs to the technical field of power system fault detection. The method comprises the following steps: disassembling a power distribution network into multiple double-end structure power distribution line sections; using a CNN-LSTM-KAN hybrid neural network to preliminarily position faults of each section, wherein the CNN extracts time-frequency diagram spatial features of a traveling wave, the LSTM captures time sequence dependency, and the KAN enhances nonlinear mapping capability; constructing a minimum spanning tree model of the power distribution network, and combining a Teager energy operator to calculate traveling wave energy of each terminal; and based on the energy sequence, using a pruning algorithm to remove false fault points, and gradually locking a real fault section. The application overcomes the shortcomings of high time synchronization accuracy requirement and high hardware cost of a traditional traveling wave method, and has higher positioning accuracy, robustness and generalization capability in complex scenes such as high-resistance faults and low initial phase angles.

The application discloses a mechanical arm path planning method based on an improved fast extended random tree and belongs to the technical field of industrial robot control. According to a set safety distance, a collision-free six-axis joint value sequence set from a starting point to an end point is quickly obtained. In initial path searching, a space preset tree strategy is used to guarantee the connection between an explored mechanical arm pose and initial and final mechanical arm poses and to make different poses as possible as to spread in space; in a path searching algorithm, a bidirectional search tree algorithm, a target bias, a greedy strategy and a variable step length strategy are used to accelerate algorithm convergence; a loop pruning algorithm is used in path optimization to greatly reduce a redundant path and to guarantee the local optimality of the path in an iteration process. The method improves the path planning efficiency of the mechanical arm in a three-dimensional space, solves the problems of long kinematics solving time of the mechanical arm in a multi-dimensional space and a redundant path and realizes a collision-free motion task in a joint space.

This invention discloses a rate-distortion optimization pruning method and system for dynamic anomaly detection in the Internet of Things (IoT). It utilizes a sliding window to extract real-time time-series samples and proposes a distribution difference quantification method based on the Kolmogorov-Smirnov test. By calculating the statistical distance between the current data and the historical baseline distribution, it identifies concept drift in real time and automatically triggers pruning and reconfiguration signals, ensuring the continuous effectiveness of the detection system in dynamic environments. In the strategy optimization stage, rate-distortion theory is introduced, and a rate-distortion objective function integrating model computational overhead, feature fidelity loss, and gradient sensitivity is constructed. The optimal pruning rate under the current distribution is obtained through optimization calculation, achieving a precise mathematical trade-off between model size and detection performance. A channel-level structured pruning algorithm is adopted to remove redundant feature channels according to the optimal pruning rate, generating a lightweight anomaly detection model with a more compact physical structure. This invention not only reduces the inference latency and memory consumption at the edge of the model but also improves the adaptability and robustness of the detection system to dynamically changing environments, possessing high engineering practical value.

The application discloses a farmland water, fertilizer and pesticide collaborative intelligent prescription map generation method based on multi-source data fusion, constructs an inter-class confusion matrix through K-means clustering and an SVM multi-classifier, accurately locates a high-risk area of the same spectrum of foreign matters, and clearly targets a specific object for subsequent processing; a band weight self-adaptive optimizer is designed by fusing a genetic algorithm, a classification accuracy and an inter-class separation degree are taken as targets, and the band weight is self-adaptively adjusted; a teacher-student double-branch feature distillation neural network with heterogeneous parameters is constructed, a composite loss function of knowledge distillation and intra-class covariance loss is designed, robust features are extracted, and spectral confusion is inhibited from the source; a node evaluation function of comprehensive sample consistency, information gain and execution cost is designed, a decision tree is simplified through a branch and bound pruning algorithm, and a decision path is dynamically and efficiently selected; Mahalanobis distance is introduced, low-confidence decisions are corrected in combination with K-nearest neighbor weighted median, a prescription map with confidence annotations is generated, and the robustness and credibility of the decisions are improved.

This invention belongs to the field of image processing and computer vision technology, and discloses a lightweight real-time detection model, method and system for power equipment with blurred targets in complex backgrounds. The lightweight real-time detection model for power equipment with blurred targets in complex backgrounds includes the following core modules: a lightweight context-guided dual-branch module HCG-Block; a dual-stream spatial attention fusion module DSAF; a high-frequency cross-layer edge enhancement module HFCE²; a lightweight windmill-shaped convolution module PSConv; a spatial channel mixing and recombination module GSConv; and an adaptive layer global pruning algorithm. It has excellent detection performance in complex scenes. Through the global context modeling of HCG-Block, the spatial attention denoising of DSAF and the high-frequency edge enhancement mechanism of HFCE², the model can still accurately extract insulator defect features in complex scenes such as fog, strong light, rain, snow and motion blur. The small target detection accuracy (AP_S) is improved by 4.2% compared with the baseline model.