20 results about "Maximum weight matching" patented technology

The invention belongs to the technical field of computers, and particularly relates to a data center server resource allocation method and system based on dynamic load balancing, and the method comprises the steps: collecting a physical resource state, instance constraint and network flow through a multi-dimensional probe, constructing a dynamic weighted load scoring model, and introducing a migration penalty term to correct node load evaluation; solving maximum weight matching of the bipartite graph with constraint by combining a Hungary algorithm, and realizing global optimal mapping of the migration instance and a target node; a memory, a CPU and network resources are pre-configured before migration, a delay recovery mechanism is triggered after migration, and the service quality is guaranteed. The system comprises a resource acquisition module, a score generation module, a matching solution module, a pre-configuration module, a self-adaptive period adjustment module and the like. The cluster resource utilization rate is obviously improved by 28%, hotspots are reduced by 76%, the SLA default rate is reduced by 92%, meanwhile, energy is saved by 15%, and collaborative optimization of efficiency and service quality is achieved.

The invention discloses a logistics supply chain digital management method and system. The method comprises the steps of obtaining a to-be-allocated task set and an available vehicle set from a logistics supply chain management system; task features and vehicle features are extracted from the task set T and the vehicle set V respectively; generating a task combination set by adopting a space-time clustering algorithm; constructing an extended vehicle set, and defining a state transition function; calculating matching feasibility and cost of each virtual node and each task combination based on a state transfer function; constructing a bipartite graph with constraints; applying a Kuhn-Miscans algorithm to the bipartite graph G with constraints to solve a maximum weight matching M; and aggregating the matching result M according to entity vehicles, and outputting a vehicle-task combination sequence allocation scheme. In view of the problems of constraint violation and state inconsistency during multi-stage distribution optimization of logistics vehicles in the prior art, the dynamic multi-stage decision problem is converted into a static graph matching problem, and effective transmission of constraints is realized.

The present application relates to the technical field of image processing, more particularly, the present application relates to a kind of based on image processing's fertilizer granularity detection method and system, comprising: identifying cluster region in image, the cluster region is the connected domain of area greater than preset area threshold and convex hull area ratio less than preset non-convexity threshold, extract the edge profile of the cluster region, and based on the multiscale curvature difference response of the edge profile obtains true concave point set;Cutting obstruction potential energy value is constructed inside the cluster region, and the cutting obstruction potential energy value is positively correlated with the distance value of pixel point to cluster region background boundary.The present application combines topological affinity and global maximum weight matching algorithm with non-intersection constraint, effectively solves the problem that path passes through granular entity and segmentation line intersection in complex ring-shaped stacked structure, significantly improves the accuracy of online detection.

The application discloses a differential privacy synthetic data publishing method based on maximum weight matching, which comprises the following steps: firstly, pre-processing user data collected by a reliable third-party server; then, using a graph model method to perform information representation on the processed data attributes, so as to obtain an attribute association graph; then, selecting a suitable low-dimensional marginal set according to a maximum weight matching algorithm; subsequently, adding noise to the low-dimensional marginal set, wherein the noise satisfies a differential privacy definition; then, performing post-processing on the low-dimensional marginal set with noise, so as to obtain a standardized low-dimensional marginal set; performing data synthesis according to the low-dimensional marginal set, so that the synthetic data set is as similar as possible to the original data set in statistical information; and finally, publishing the synthetic data set. By using the technical method, the computational complexity can be reduced while ensuring the utility of the synthetic data set, and the method has better utility for high-dimensional data.

The invention provides a constellation network topology planning method based on robustness optimization. According to the method, on the basis of multi-dimensional physical constraints, link feasibility and a network topology optimization target are considered, and periodic updating and rolling planning of a link state are realized by establishing a time slice discretization model. Furthermore, algebraic connectivity is used as a network robustness index, a continuous weight solution is solved by adopting positive semidefinite programming, and a discrete link scheme meeting node degree constraint is obtained through weight polarization and bipartite graph maximum weight matching, so that connectivity and structural stability of the constellation network are effectively improved.

The invention provides an airport boarding bridge guarantee service personnel automatic dispatching method comprising the following steps: S1, obtaining flight information in a future preset time period, and generating a boarding bridge guarantee task set K corresponding to flights; s2, obtaining available employee information at the current moment, and generating an available employee set S; s3, calculating the comprehensive matching degree between the employees and the boarding bridge guarantee tasks according to the boarding bridge guarantee task set K and the available employee set S, and constructing a weighted bipartite graph of the employees and the boarding bridge guarantee tasks according to the comprehensive matching degree; and S4, solving the weighted bipartite graph by applying a bipartite graph maximum weight matching algorithm to generate an employee dispatching result, and carrying out emergency processing on tasks which are not successfully matched in the scheme. According to the invention, automation of boarding bridge guarantee task and personnel matching is realized, and improvement of airport ground service guarantee efficiency is facilitated.

This invention discloses a deep learning-based method and system for pairing gravitational waves using gravitational lensing. The method includes: acquiring time-series strain data from a gravitational wave detector; performing frequency domain whitening on the data and converting it back to the time domain, and extracting fixed-length whitened strain segments; inputting the data into a twin encoder; selecting several other events with the highest similarity for each gravitational wave event based on the cosine similarity between feature vectors to form a candidate pairing set; calculating the waveform cross-correlation using a generalized cross-correlation-phase transform algorithm to obtain a refined similarity score and an estimate of the relative time delay between event pairs; constructing a weighted graph, and obtaining the final pairing result by solving for the maximum weight matching of the graph. In this invention, the twin encoder is used to reduce the dimensionality of high-dimensional time-domain data, and a coarse-screening strategy avoids full pairwise comparisons of the entire catalog, greatly improving the search speed on large-scale datasets.

The application belongs to the technical field of data mining, and specifically discloses a cross-field technology fusion opportunity finding method based on science and technology fusion interaction, which comprises the following steps: obtaining scientific paper and technology patent data, identifying cross-field scientific fusion and technology fusion based on semantic anchor points and a similarity threshold; using a trained deep text clustering model to respectively perform deep clustering on scientific fusion and technology fusion, and obtaining corresponding clustering sets; obtaining text representation corresponding to each cluster, filtering through a threshold, and establishing association between two types of clusters through bipartite graph maximum weight matching to identify shared topics and exclusive topics; using a dynamic time warping method to align the time series of scientific and technological trends, judging topics with scientific convergence preceding technological convergence, and identifying technology fusion opportunities with growth potential based on the development trend of scientific papers. The application can deeply mine cross-field science and technology semantics, improve clustering accuracy, quantify science and technology association, and dynamically find forward-looking technology fusion opportunities.

The application provides a wharf vehicle operation near connection method and device, which is applied to the field of information technology, maps intersections, lanes, vehicles, tasks and matching relations into Actor instances based on an Actor distributed model, responds to vehicle and task messages in real time, predicts the time for each vehicle to reach a target loading point by searching for the fastest driving path in combination with all state vehicles in the task allocation range, establishes the association between operation vehicles and tasks under the condition of meeting the task time window, constructs one or more groups of weighted bipartite graphs with controllable scales by taking the inverse number of the predicted empty driving time as the weight, solves the maximum weight matching by using the Hungarian algorithm in parallel, eliminates the multi-task matching conflicts caused by graph segmentation by the vehicle Actor, and cyclically executes the above process when the vehicle or task state changes, so that the global empty driving total time is minimized, the operation energy consumption and operation cost are significantly reduced, and the overall efficiency and resource utilization of the wharf loading and unloading operation are improved.

The invention relates to a similarity calculation method of a affair graph and a related product. The method comprises the following steps: acquiring a plurality of affair maps; based on different types of information similarity models, respectively analyzing the similarity of the target information between the affair maps to obtain different types of information similarity features; on the basis of domain weights, the information similar features are fused, associated features are obtained, and the domain weights are determined according to domain knowledge; fusing the weight of each matching edge under the maximum weight matching of the bipartite graph to obtain a graph similarity; the bipartite graph is determined according to the associated features. According to the scheme provided by the invention, the domain weight is determined through the domain knowledge, and the fusion process of the information similar features is guided to improve the influence of key elements in the information similar features on the associated features, so that the related parameters of the bipartite graph more accurately reflect the similarity degree between the two affair maps, and accurate calculation of the map similarity is facilitated.

The invention discloses a gravitational lens gravitational wave pairing method and system based on deep learning, and the method comprises the steps: obtaining time sequence strain data of a gravitational wave detector, carrying out the frequency domain whitening processing of the data, converting the data back to a time domain, and carrying out the interception to obtain a whitening strain fragment with a fixed length; inputting into a twin encoder; selecting a plurality of other events with the highest similarity for each gravitational wave event based on the cosine similarity between the feature vectors to form a candidate pairing set; a generalized cross-correlation-phase transformation algorithm is adopted to calculate waveform cross-correlation, and refined similarity scores and relative time delay estimation between event pairs are obtained; and constructing a weighted graph, and obtaining a final pairing result by solving the maximum weight matching of the graph. According to the method, dimension reduction is carried out on the high-dimensional time domain data by using the twin encoder, and full-amount pairwise comparison on directories is avoided through a coarse screening strategy, so that the search speed on a large-scale data set is greatly improved.

The application relates to the technical field of semantic analysis, in particular to a bid evaluation clause extraction system based on semantic analysis, which comprises a bid text analysis module, wherein according to a bid text segment in an extracted bid, concept node information entropy weight is generated, the concept node information entropy weight is calculated with an extracted node associated semantic vector, and a binary semantic feature fingerprint is generated. In the application, the maximum weight matching node set is subjected to position alignment splicing, numerical parameter and text string reconstruction, and an index label is established in combination with entity name content and time node value, finally forming a bid evaluation rule extraction dictionary, so that the output result is changed from pure semantic interpretation into a structured rule unit which is searchable, archivable, callable and reviewable, and the bid evaluation rule extraction scene presents stronger clause restoration capability, rule aggregation capability, heterogeneous expression compatibility and subsequent reuse capability.

The application relates to the technical field of metallurgical automation, and discloses a method, equipment, medium and program product for cross-process acting of cold and hot coil excess materials. The method for cross-process acting of cold and hot coil excess materials provided by the application is used for an electronic device, the method acquires multiple future contracts and multiple cold and hot coil excess materials of the same steel grade, takes the future contracts and the cold and hot coil excess materials as nodes in a weighted bipartite graph, forms at least one matching pair composed of one future contract and one cold and hot coil excess material to construct an acting matching relationship graph, determines the matching weight of each matching pair in the at least one matching pair, and then solves the acting matching relationship graph through a bipartite graph maximum weight matching algorithm to obtain an acting scheme, so that the automation and large-scale of the cross-process acting of the cold and hot coil excess materials are realized, the acting efficiency of the future contracts and the cold and hot coil excess materials is improved, and globally optimal excess material acting can be realized.

The application relates to the technical field of data security management, in particular to an archive data access control method and system based on multi-level encryption, which comprises the following steps: constructing a signal strength benchmark and an environmental background noise vibration benchmark by using continuous RSSI and acceleration data; secondly, extracting a footstep vibration and a signal drop event, analyzing the correlation characteristics of the two by using a bipartite graph maximum weight matching algorithm, and calculating a walking signal consistency score; finally, dynamically updating trust points according to the score, generating a decryption key only when the trust points meet the standard, and destroying the key otherwise. The application aims to control the real-time safe access of encrypted archive data by accurately checking the consistency of the walking behavior of the escort.

The invention provides a constrained multi-objective collaborative optimization method and system, and belongs to the technical field of port logistics, and the method comprises the steps: generating an initial path solution through a LACAM algorithm; judging whether termination conditions such as time, times or improved stagnation are met or not, and if yes, outputting; if not, a compromise graph is constructed, nodes of the compromise graph are AGVs, edge weights are the sum of historical compromise numbers, and maximum weight matching is solved to determine a modification set M; fixing a non-modified set path as an obstacle, and re-planning an AGV path in M; and updating the solution by using the new path and returning to iteration. The feasibility and completeness of the initial solution of the LCAM are ensured; the compromise graph quantifies the AGV abnormal influence, and the high-optimization potential AGV pair is accurately positioned through maximum weight matching; the calculation complexity is reduced by local re-planning, the compromise number is continuously reduced by iterative optimization, and the solution quality is improved; the method gives consideration to real-time performance and global optimization, and is suitable for a dynamic AGV scheduling scene.

Embodiments of the present application provide a computing power scheduling method for large model training, an electronic device, a storage medium and a program product. The method comprises: constructing a global resource view based on real-time state information of heterogeneous computing power resources; obtaining task characteristics of a training task of a large model to be scheduled, and predicting a resource demand change trend of the training task in a preset future time window according to performance data of the training task in a preset historical time window; based on the global resource view and the resource demand change trend, constructing a constrained bipartite graph maximum weight matching problem between a task unit and a resource unit, determining a matching relationship between the task unit and the resource unit; and performing task placement according to the matching relationship. The accuracy of the scheduling decision and the utilization rate of the cluster resources are improved, and the problems of the traditional scheduler, such as lack of awareness of task characteristics, static resource view, and poor adaptation to heterogeneous environments, are solved.

This invention discloses a dynamic order segmentation and matching method for V2V energy transactions, belonging to the fields of electric vehicle energy management and artificial intelligence technology. It includes: real-time collection of order information and supply vehicle status information; segmenting orders into urgent and non-urgent orders based on their urgency; in the first stage, constructing a weighted bipartite graph for urgent orders and prioritizing matching through maximum weight matching; in the second stage, for non-urgent orders, modeling the segmentation ratio decision as a Markov decision process, using Deep Sets encoding to process the dynamically varying number of orders and vehicles, outputting the segmentation ratio and generating sub-orders; subsequently, performing a second-stage weighted matching between the segmented order set and the remaining supply vehicles; and executing the transaction after confirmation. This invention solves the problems of low resource utilization, difficulty in matching low-battery supply vehicles, and resource fragmentation in traditional V2V transactions. By activating idle battery resources and dynamic scheduling, it significantly improves the peak-period transaction volume and the overall energy fit of the system.

The invention relates to the technical field of image processing, and more specifically, the invention relates to a fertilizer granularity detection method and system based on image processing, and the method comprises the steps: recognizing a cluster region in an image, the cluster region being a connected domain with an area greater than a preset area threshold and a convex hull area ratio less than a preset non-convexity threshold, extracting an edge contour of the clustered region, and obtaining a true concave point set based on the multi-scale curvature difference response of the edge contour; and constructing a cutting hindering potential energy value in the cluster-shaped region, wherein the cutting hindering potential energy value is positively correlated with a distance value from the pixel point to the background boundary of the cluster-shaped region. According to the method, the topology affinity and the global maximum weight matching algorithm with the non-cross constraint are combined, the problem that the path penetrates through the particle entity and the segmentation line cross in the complex annular stacked structure is effectively solved, and the accuracy of online detection is remarkably improved.

The application provides a kind of constraint multi-objective collaborative optimization method and system, belong to port logistics technical field, the initial path solution of the application is generated with LaCAM algorithm;Then judge whether to meet time, times or improvement stagnation termination condition, meet then output;Not satisfied then build compromise graph, the node of compromise graph is AGV, and the edge weight is the total sum of historical compromise number, determine the modification set M by solving the maximum weight matching;Fixed non-modification set path is an obstacle, and the AGV path in M is re-planned;Update the solution with new path and return iteration.LaCAM initial solution guarantees feasibility and completeness;Compromise graph quantifies AGV abnormal influence, maximum weight matching accurately locates high optimization potential AGV pair;Local re-planning reduces computational complexity, iterative optimization continuously reduces compromise number, improves solution quality;Real-time and global optimization are considered, suitable for dynamic AGV scheduling scene.

The application relates to a dynamic obstacle filtering method based on clustering tracking, which is characterized in that: the original point cloud is subjected to object-level segmentation through an adaptive Euclidean clustering algorithm of a multi-line laser radar; subsequently, a similarity function is constructed by calculating the distance between clusters and the volume ratio of the boundary box, so as to form a bipartite graph describing the matching degree between frames; the Kuhn-Munkres algorithm is adopted to find the maximum weight matching, so as to realize the association of clusters; the motion trajectory of the cluster center is tracked, the short-term motion consistency is utilized to judge the dynamic obstacle, and the point cloud information of the dynamic obstacle is filtered in the mapping process. The method effectively solves the deficiency of the traditional method in real-time performance and precision, and significantly improves the navigation and positioning performance of the mobile robot in a complex environment. Compared with the prior art, the application has the advantages of high filtering efficiency, high precision and good continuity and real-time performance for dynamic obstacles.