40 results about "Computational topology" patented technology

The present invention relates to a hierarchical system and method for task assignment (TA), coordination and communication of multiple Unmanned Aerial Vehicles (DAV's) engaging multiple attack targets and conceives an ad-hoc routing algorithm for synchronization of target lists utilizing a distributed computing topology. Assuming limited communication bandwidth and range, coordination of UAV motion is achieved by implementing a simple behavioral flocking algorithm utilizing a tree topology for target list routing. The TA algorithm is based on a graph-theoretic approach, in which a node locates all the detectable targets, identifies them and computes its distance to each target. The node then produces an attack plan that minimizes the sum of distances of the UAV's in the subtree of a given node to the targets.

Methods and apparatus to optimize workflow execution by the intelligent dispatching of workflow tasks against a grid computing system or infrastructure are provided. For some embodiments, a grid task dispatcher may be configured to dispatch tasks in a manner that takes into account information about an entire workflow, rather than just an individual task. Utilizing information about the tasks (task metadata), such a workflow-scoped task dispatcher may more optimally assign work to compute resources available on the grid, leading to a decrease in workflow execution time and more efficient use of grid computing resources.

The invention discloses a drilling strain network topological structure anomaly detection method based on a multi-channel singular spectrum. The method comprises the following steps: selecting N drilling strain stations as N network nodes; setting a time window length n and a moving step length step, and constructing a multi-dimensional node matrix of which the drilling strain in the time window is n * N; multi-channel singular spectrum analysis is carried out on the multi-dimensional node matrix, and obtaining a feature vector and a corresponding space time principal component after decomposition; selecting a high-frequency component to reconstruct an effective component of each network node; calculating a Pearson correlation coefficient between every two nodes; communicating two networknodes of which the correlation coefficients are greater than 0.8, constructing edges of the strain network, and writing an adjacent matrix of the strain network; calculating topological structure attributes of the strain network; moving the time window, constructing a strain network in the time window, and calculating topological structure attributes; and outputting an evolution diagram of the borehole strain network structure, and judging earthquake-related anomalies. According to the invention, multi-station combined analysis of the borehole strain data is realized.

The present invention provides a method and system for determining information similarity. Based on the time when information is consumed, the information consumed by each user is sorted to obtain the corresponding first information behavior sequence; the first information behavior sequence corresponding to all users is The sequence is integrated to obtain the second information behavior sequence; calculate the first similarity between each pair of information with a co-occurrence relationship in the second information behavior sequence; based on the second information behavior sequence, construct a topological map between information; use The graph convolution algorithm calculates the second similarity between each pair of nodes in the topological graph; based on the first similarity between each pair of information with a co-occurrence relationship, combined with the second similarity between each pair of nodes in the topological graph The second similarity is to determine the final similarity between each pair of information in the second information behavior sequence. Combine the two methods to calculate the final similarity between information, increase the amount of information covered by the method of calculating similarity, and then improve the accuracy of information recommendation.

A comprehensive topology optimization method, which divides network topology optimization into two stages. Firstly, it judges the operating status of the current system. The criticality of each node in the network, and keep the variance of the nodes in the entire topology within a certain range, effectively reducing the impact of the failure of each node and improving the robustness of the network topology; after the network topology of the system is running, the proposed The topology optimization method based on credibility and weighted load flow, quantitatively calculates the credibility of each node and the weighted load flow of each link, and keeps the credibility of the node and the weighted load flow of each link at A certain level is not only conducive to the optimization of the topology, but also helps the topology to quickly discover the failure of a certain node or a certain link.

The invention discloses a method and system for identifying vulnerable nodes in a power distribution system of a data center, which belongs to the technical field of power analysis, and can fully consider the topology structure of the power distribution system and the load and load level of load equipment, and accurately analyze the Identification results of vulnerable nodes. The method includes: based on the topological structure G of the power distribution system and the load parameters of the load equipment, sequentially calculate the importance index of each node in the topological structure G; screen out a plurality of different node combinations according to the important index, and combine each node combination The corresponding nodes in are excised from the topology G, and the recovery topology T corresponding to the number of node combinations is obtained n ; Based on the recovery topology T n The load parameters of the medium load equipment and the load parameters of the load equipment in the topology G are calculated for each restoration topology T n The load loss index; according to each restoration topology T n Corresponding load loss index and self-reliability index, screen out the recovery topology T with the highest vulnerability n , and output the vulnerable node identification results.

The invention discloses a GPU topology partitioning method and device, including: determining the interconnection bandwidth between multiple GPUs according to the physical topology information of multiple GPUs, and generating a GPU topology map including multiple GPUs; GPUs are randomly divided into two partitions; the migration gain of all GPUs in the GPU topology map is calculated, and the GPU with the highest migration gain in the partition including more GPUs is migrated to the partition including fewer GPUs, and the crossover of the current partitioning scheme is calculated. The number of partition connections, and remove the migrated GPU from the GPU topology map; repeat the previous step until all GPUs in the GPU topology map are removed, and select the partition scheme with the smallest number of cross-partition connections as the partition result. The present invention can specifically optimize the topological partition of the GPU from the bottom layer according to different connection relationships between the GPUs, reduce the time-consuming transmission between the GPUs, and improve the calculation speed of artificial intelligence.

The present invention relates to a three-dimensional reconstruction method, device, electronic equipment and computer storage medium, wherein, the method includes: calculating the relative pose similarity transformation between the SFM models in the topological relationship diagram, the topology relationship diagram is when the image set is collected The obtained spatial position prior, each SFM model is obtained by three-dimensional reconstruction based on the SFM algorithm of each image set of a large-scale scene block; the maximum spanning tree is obtained by processing the topological relationship graph, and the maximum spanning tree is a subset of the topological relationship graph; Find the node with the largest degree in the maximum spanning tree as the root node, and use the root node as the reference coordinate system to perform similar transformation on the SFM models on other nodes to unify the coordinate systems of multiple SFM models and obtain the combined target 3D map, and perform a global optimization of the pose graph to obtain the optimized target 3D map. Through this application, the problem of low efficiency in three-dimensional reconstruction of large-scale scenes is solved.

The invention relates to a method for judging topology recognition rate based on line loss, which belongs to the technical field of smart grid, including obtaining the result of automatic topology recognition in the distribution transformer station area, and calculating the topology result according to distribution transformer-outlet cabinet-branch box-table The level division of the box; obtain the power data of the low-voltage side of the distribution transformer, the outgoing line of the outlet cabinet, and the incoming line of the meter box; calculate the line loss rate from the low-voltage side of the distribution transformer to the meter box, and calculate the line loss rate from the outlet cabinet to the meter box. Determine whether the current line loss is within the preset range. If it is within the preset range, judge whether all the branch lines have been judged. If the judgment is completed, calculate the accuracy of topology recognition; if it exceeds the preset range, it is judged as abnormal. Classify and judge the abnormal line loss rate, and record the topology. The invention is directly used for judging whether the topology of the low-voltage station area is accurate, and provides important basis and scientific support for whether the automatic topology recognition of the low-voltage station area is accurate.