17 results about "Combinatorial search" patented technology

The invention discloses a power system moving target defense method based on power flow betweenness and power flow disturbance, and relates to the field of power systems. Calculating to obtain a power flow betweenness and a power flow disturbance index of the normalized power system line; constructing a comprehensive index based on the power flow betweenness and the power flow disturbance index; and determining a to-be-deployed line set according to a preset device deployment number and the comprehensive index. And changing the line impedance value corresponding to the line set to obtain a target measurement matrix, calculating a measurement value residual error after the line impedance value is changed according to the original measurement matrix and the target measurement matrix, and determining an attack success rate based on a size relationship between the measurement value residual error and a detection threshold value. And deploying power system moving target defense according to a result of the attack success rate. According to the invention, the detection effectiveness and the MTD concealment are balanced. A traditional global combination search problem is converted into a linear complexity sorting problem, rapid solving is achieved, calculation efficiency is improved, and rapid generation of a hidden MTD strategy is achieved.

The embodiment of the invention relates to the technical field of quantum computing and quantum processor resource management, and provides a neutral atom quantum processor virtualization scheduling method based on a capability block, which comprises the following steps: receiving and analyzing a quantum algorithm task, and extracting a structural feature set; converting the structural feature set into an algorithm capability portrait based on a feature and capability mapping rule; obtaining a capability block resource pool comprising a plurality of capability blocks and block capability portraits thereof; carrying out capability matching on the algorithm capability portrait and the block capability portrait, and judging a coverage relation; determining a candidate capability block based on the coverage relationship and performing feasibility judgment; performing cross-region combination search on the candidate capability block set to construct a candidate block combination; performing cost evaluation on each candidate block combination; and selecting a target block combination based on the combination cost, constructing a virtual processor description object and outputting the virtual processor description object to an execution layer. Therefore, implementation and expandability of a complex quantum algorithm on a large-scale neutral atom quantum processor are realized.

The invention discloses an automatic part processing method based on CATIA and Excel. The method comprises the steps that a target part list is determined; a VBA program is utilized in Excel, part identifiers in the list are converted into combined search character strings capable of being recognized by CATIA, and special characters are automatically processed in the process; opening a general assembly in the CATIA, entering an advanced search interface, and importing the combined search character string; searching is executed to select 3D digital models corresponding to all target parts in batches at a time; and carrying out subsequent batch processing operation on the selected part set, wherein the subsequent batch processing operation comprises batch activation, color change, subset copying and creating or exporting. According to the method, whole-process automation from the list to three-dimensional model batch processing is achieved, manual operation of several hours is shortened to several minutes, the accuracy rate is 100%, the design efficiency and the data quality are greatly improved, and meanwhile based on a universal software platform, the implementation cost is low.

The application discloses a three-dimensional human skeleton fitting method and system based on a multi-camera joint association graph and combined search, and relates to the three-dimensional human skeleton fitting method, which comprises the following steps: S1, acquiring images collected by multiple cameras, detecting human two-dimensional joint nodes by using a two-dimensional pose estimation algorithm, and projecting the two-dimensional joint nodes into three-dimensional rays; S2, taking the joint nodes as nodes and the geometric consistency between three-dimensional rays of different cameras as edge weights, constructing a multi-camera joint node association graph; S3, performing group clustering on the cross-camera joint nodes with high geometric consistency on the association graph by using a breadth-first search, and forming a candidate skeleton cluster; S4, reserving the first K joint nodes with the highest geometric consistency with the candidate skeleton cluster in each camera, and obtaining candidate joint node pairs; S5, using a depth-first search (DFS) to enumerate the cross-camera joint node pairs, and screening optimal joint node pairs in combination with a pruning strategy; and S6, performing triangulation processing on the screened optimal joint node pairs, obtaining three-dimensional skeleton points, and outputting a three-dimensional human skeleton structure.

This invention provides a reinforcement learning-based data augmentation method for small target detection. It involves acquiring images of small targets and dividing them into training and validation sets based on image quantity and classification. Hyperparameters of the reinforcement learning controller are set to determine a set of potential image augmentation operations. A two-stage decoupled search is employed: the search space is decomposed into individual parameter search and policy combination search; the individual parameter search space is determined; the individual parameter search is performed; the policy combination search space is determined; within the policy combination search space, the reinforcement learning controller continues the search to obtain the optimal policy combination; and the optimal policy combination is used for data augmentation. This method significantly reduces search time, enabling reinforcement learning to be effectively used in data augmentation, enriching the dataset, and improving the accuracy of small target detection by selecting the optimal policy combination.

The invention discloses an equipment support stability control method based on a swarm intelligence algorithm, and the method comprises the following steps: S1, collecting the operation state data of an equipment support, and constructing a state sensing matrix; s2, constructing a comprehensive similarity matrix, executing clustering division, and constructing an intelligent group structure; s3, constructing a stability control optimization model, and executing control parameter combination search by adopting an improved TSO algorithm to obtain an initial optimization result of the control parameters; s4, executing dynamic correction on the initial optimization result of the control parameters; s5, judging whether the stability control optimization model meets a convergence condition or not, and generating a stability control parameter set; s6, setting control parameters, and updating the state sensing matrix; and S7, storing in an evolutionary memory bank and updating. The stability control precision and response efficiency of the equipment support are improved, and the method has the advantages of being high in adaptability, stable in control effect and high in convergence efficiency.

Methods, systems, and computer readable media related to generating a combined search query based on search parameters of a current search query of a user and search parameters of one or more previously submitted search quer(ies) of the user that are determined to be of the same line of inquiry as the current search query. Two or more search queries may be determined to share a line of inquiry when it is determined that they are within a threshold level of semantic similarity to one another. Once a shared line of inquiry has been identified and a combined search query generated, users may interact with the search parameters and / or the search results to update the search parameters of the combined search query.

The invention relates to the field of machine learning and image processing, provides an effective parameter fine-tuning algorithm based on adapter module combinatorial search and a coarse and fine hypernet training strategy, and solves the problem that the existing effective parameter fine-tuning algorithm introduces a large amount of parameters and is difficult to obtain optimal balance of the parameters and precision. The method comprises the following steps of: constructing a coarse super-network architecture which comprises a plurality of adapter types and uses an attention module to guide importance; updating the weight of a coarse hypernet adapter and the weight of an attention module based on the downstream data set, and executing channel pruning based on the precision difference of the data set before and after removal on the Backbone weight; obtaining adapter type selection according to the attention module amplitude value and generating an initial fine hypernet weight; optimizing the fine hypernet weight based on a weight entanglement strategy; performing channel pruning and adapter pruning based on a random mask and an amplitude value mask; and according to the fine super-network optimization weight, searching a hidden layer dimension combination with optimal precision. The method has the advantage that balance of parameter quantity and precision is obtained.

The invention discloses a microfilter combination automatic search heterogenous remote sensing image change detection method and system, and the method comprises the steps: S1, carrying out the feature fusion processing of two obtained heterogenous remote sensing images, and obtaining a fusion feature; s2, according to the fusion features, a convolution kernel in a discriminator is dynamically synthesized by adopting a microfilter combinatorial search strategy, and parameters of a generator and the discriminator are optimized through adversarial training; and S3, supervising the change atlas output by the generator based on geometric structure collaborative supervision loss to obtain a final change detection result. According to the method, the common texture and structural features in the optical and SAR images can be effectively extracted, the problems of large radiation difference and much noise of different-source images are solved, and the omission ratio and the false detection rate are remarkably reduced. Meanwhile, the pixel precision, the marginal definition and the global structure authenticity are coordinated, so that the detected change area is clear in boundary and complete in structure, and common salt and pepper noise and boundary blur phenomena are effectively eliminated.

The invention relates to the technical field of power system operation and emergency scheduling, and provides a power transmission network toughness collaborative optimization method and system for unconventional events, and the method comprises the steps: constructing a concurrent triple scene, and unifying parameter calibers; inputting the constructed disturbance scene and power flow data, continuing to establish a feasible region and completing sequential decision modeling; performing deep Q network training under feasible region gating; performing combined search on the emergency tie line at an execution end, if any constraint is violated, regarding the scheme as infeasible, and after beam search is completed, obtaining an optimal scheme set and a corresponding power flow feasible solution; and according to a mode of gradually executing and gradually checking the power flow, immediately solving and recording the alternating current power flow after the action of each step falls to the ground, and uniformly checking and outputting a result on a fixed test set. A physically feasible and landing toughness improvement scheme can be generated in an emergency window, communication recovery and pressure drop loss are realized, overloading is reduced, and rapid command decision and afterward recovery tracing are supported.

The application discloses a method and device for searching an object, electronic equipment and a storage medium, and is applied to the technical field of data mining. The method comprises the following steps: receiving a search request for an object to be searched, wherein the search request carries search indication information, identification information of the object to be searched and a request triggering moment; the search indication information is any one of content search information, content group search information and content and content group search information; the search indication information is used for indicating single search and / or combined search of multiple unstructured data; a target search result is determined based on target search rules corresponding to the search indication information and associated information of the object to be searched; and the target search result comprises at least two kinds of unstructured data associated with the object to be searched and determined from a preset database. It can be seen that the application can obtain at least two kinds of unstructured data, such as graphic text, video and information, associated with the object to be searched.

The invention discloses a three-dimensional human skeleton fitting method and system based on a multi-camera association graph and combinatorial search, and the method comprises the steps: S1, obtaining an image collected by multiple cameras, detecting a two-dimensional joint point of a human body through a two-dimensional posture estimation algorithm, and carrying out the back projection of the two-dimensional joint point into a three-dimensional ray; s2, constructing a multi-camera joint point association graph by taking the joint points as nodes and geometric consistency among three-dimensional rays of different cameras as edge weights; s3, clustering cross-camera articulation points with relatively high geometric consistency on the association graph through breadth-first search to form candidate skeleton clusters; s4, reserving the first K articulation points with the highest geometric consistency with the candidate skeleton cluster in each camera to obtain candidate articulation point pairs; s5, enumerating cross-camera joint point pairs by adopting depth-first search (DFS), and screening out an optimal joint point pair in combination with a pruning strategy; and S6, performing triangularization processing on the screened optimal articulation point pair to obtain a three-dimensional skeleton point, and outputting a three-dimensional human skeleton structure.

The embodiment of the application relates to the technical field of quantum computing and quantum processor resource management, and provides a neutral atom quantum processor virtualization scheduling method based on a capability block, which comprises the following steps: receiving and analyzing a quantum algorithm task, and extracting a structural feature set; converting the structural feature set into an algorithm capability image based on a feature and capability mapping rule; acquiring a capability block resource pool containing multiple capability blocks and block capability images of the capability blocks; performing capability matching on the algorithm capability image and the block capability images, and judging a covering relationship; determining candidate capability blocks based on the covering relationship and performing feasibility judgment; performing cross-zone combination search on the candidate capability block set, and constructing candidate block combinations; performing cost evaluation on each candidate block combination; selecting a target block combination based on the combination cost, constructing a virtual processor description object, and outputting the virtual processor description object to an execution layer. Therefore, the implementability and scalability of a complex quantum algorithm on a large-scale neutral atom quantum processor are realized.

The present application relates to the technical field of face picture search, and particularly relates to a combined search result optimization method and device, computer equipment and a storage medium. The method constructs a high-confidence search support set, updates the similarity of each picture in the high-confidence search support set to a comprehensive similarity relative to N to-be-retrieved face pictures, and updates the similarity of each picture in the remaining picture set to a comprehensive similarity relative to N to-be-retrieved face pictures and M high-confidence search support set pictures, thereby replacing the original similarity based on a single picture with a comprehensive similarity based on multiple pictures, fully considering the representation difference of face features between different pictures, and being able to ensure that the higher the comprehensive similarity, the greater the probability that the picture and the to-be-retrieved face picture belong to the same person, and finally arranging in descending order according to the updated similarity, which can realize that the pictures of the same person as the to-be-retrieved face picture are as far forward as possible, and the pictures of different persons are as far backward as possible.

The invention belongs to the technical field of vehicle-mounted intelligent control, and discloses an automobile control method and device based on combinatorial search, and the method comprises the steps: obtaining user voice data, carrying out the intention analysis, and obtaining an optimization target and a constraint condition; generating a formalized task description according to the optimization target and the constraint condition; obtaining a user portrait, vehicle environment data and vehicle state data, and constructing a state space vector; performing combined search in a preset atomic action library according to the formalized task description and the state space vector by utilizing a heuristic graph search algorithm to obtain a plurality of alternative control strategies; calculating a comprehensive score of each alternative control strategy by adopting a multi-target evaluation algorithm; and determining and executing a target control strategy according to the comprehensive score or the user selection instruction. According to the method and the device, feasible and cross-domain alternative control strategies can be generated in the preset atomic action library which can be executed by the vehicle and the multi-dimensional environment state space.

The invention discloses power load decomposition based on an interpretable decision tree and a hidden Markov model, relates to the technical field of power energy consumption monitoring, and solves the problems that NILM research focuses on improvement of algorithm performance and decomposition precision and the model interpretability is insufficient increasingly, and a deep learning method is relatively high in precision and relatively low in reliability. And the black box characteristic causes the lack of explicit mapping between input and output, so that the result credibility is difficult to verify. The invention discloses power load decomposition capable of explaining a decision tree and a hidden Markov model. The power load decomposition comprises the following steps: S1, constructing an equipment feature library; s2, training and applying a decision tree pruning model; s3, constructing a hidden Markov model; s4, carrying out optimal state combination search and power decomposition; and S5, outputting and visualizing a result. According to the invention, by constructing an intelligent double-level analysis framework, accurate decomposition and state identification of the hybrid load are realized, and technical support is provided for energy consumption management in the electric power field.

The present application relates to the field of data analysis, especially to a heat stroke pathological data integration information collection processing method based on deep learning, comprising: determining the data state of user analysis data according to the data richness and image transformation completion degree of the user analysis data; when the keyword selection mode is the association group selection generation, determining the number of selected keywords according to the group influence coefficient, and selecting the keywords according to the order from large to small of the sensitive coefficient; when the keyword selection mode is the image transformation supplement generation, determining the keyword generation mode according to the feature representative degree of the divided paragraph, which is to generate keywords according to the median point or outlier point; under the initial search completion condition, optimization analysis is carried out, and the search optimization mode is determined to be combination search or sequential search according to the repeated keyword analysis state, which improves the pathological data integration search efficiency.