13 results about "Cluster tree" patented technology

The invention relates to the technical field of digital data processing, in particular to a circuit breaker fault detection method based on electrical parameter analysis, and the method comprises the steps: obtaining monitoring data, and constructing a parameter matrix; screening suspicious elements based on prediction deviation of elements in the parameter matrix and a mutation point detection result, and determining a first characteristic value according to a starting moment determined by the suspicious elements and a rule degree of data fluctuation in a fluctuation duration of the suspicious elements, and determining a fault interval by combining a consistency analysis result of irregular data fluctuation in parameter matrixes of different types of monitoring data, screening target data from residual terms, obtaining a plurality of clustering trees based on the target data by using an agglomerated hierarchical clustering algorithm, and determining a fault detection result of the circuit breaker based on differences among different clustering trees. According to the method, the residual component is used for replacing the initial data, so that the clustering effect of the clustering number on the monitoring data with different fault sensitivity degrees is improved while the clustering data size is reduced.

The invention discloses a crowd test report sorting method based on big language model multi-modal multi-layer clustering tree analysis. A clustering module, a structuring module and a sorting module are included. And the clustering module comprises feature extraction and clustering, performs semantic understanding on the image and the text by utilizing LLM, constructs a clustering prompt according to a preset prompt template, and inputs the LLM to generate a clustering result. The structuring module analyzes the clustering result output by the LLM into a multi-level clustering tree so as to represent the hierarchical relation and category between the defect reports, and metadata such as access time and accessibility are recorded on nodes. And the sorting module generates a report sorting sequence which covers different categories and is not repeated by adopting a recursive selection strategy based on the clustering tree and the node metadata so as to improve the examination efficiency and the defect discovery capability of the test report.

The invention belongs to the technical field of Internet of Things, particularly relates to an optimal routing selection method based on routing cost, and aims to solve the routing selection problem in a multi-routing cluster tree wireless MESH network. According to the method, the local routing table containing the destination node address, the next hop address, the routing cost and the updating time is maintained at each routing node, so that the function of dynamically updating the routing information is realized. The routing cost is a result obtained by adding the cost factors (such as electric quantity, link quality, available load and the like) of each node on the message transmission path according to a preset weight. The mechanism ensures that various influence factors can be comprehensively considered during route selection, so that the optimal path is selected. No matter the uplink message forwarding or the downlink message forwarding, the method can guide the routing selection through the routing cost updated in real time, thereby effectively improving the success rate of message transmission and reducing the transmission time delay.

The invention discloses an electric power data fingerprint evaluation system based on a Bayesian network, relates to an electric power data fingerprint evaluation device, and aims to solve the problem that an electric power data fingerprint evaluation result is inaccurate due to the fact that an existing evaluation method cannot objectively reflect information of electric power data fingerprints. A fingerprint analysis module is used for performing performance analysis on electric power data and selecting an alternative index set; the screening and evaluating module is used for screening the alternative index set, screening out evaluation indexes influencing the fingerprint security of the power data, performing analytic hierarchy process on the evaluation indexes, and determining the relationship strength among the node indexes; the index quantification module is used for quantitatively giving the prior probability of each node index; the model construction module is used for constructing a structure of a Bayesian network model; the parameter learning module is used for giving Bayesian network parameters; the cluster tree conversion module is used for generating a cluster tree; and the result output module is used for outputting a data fingerprint evaluation result. The beneficial effect is that the accuracy of the evaluation result is improved.

The application discloses a self-adaptive text classification method in the field of scientific information, which comprises the following steps: training a classifier and establishing an original clustering forest; selecting a suitable classifier and establishing an optimal clustering forest, wherein the classifier is a clustering tree algorithm; and testing the latest sample to obtain an integrated final result. The application can improve the classification effect and generalization ability of the text classification in the field of scientific information, thereby improving the accuracy of the text classification model and the accuracy of the text classification.

The invention discloses a tree point cloud branch and leaf separation method and system based on a multi-dimensional super voxel random forest, terminal equipment and a medium, and relates to the technical field of tree point cloud branch and leaf separation, and the method comprises the steps: carrying out the clustering of tree point clouds, and obtaining a tree super voxel set; performing feature extraction on the tree super voxel set to obtain multi-dimensional features of the tree super voxel set; inputting the multi-dimensional features of the tree super-voxel set into a pre-trained random forest classifier model to obtain an initial branch super-voxel set and an initial leaf super-voxel set; and performing category distance constraint processing on the initial branch super-voxel set and the initial leaf super-voxel set to obtain a branch super-voxel set and a leaf super-voxel set. According to the method, point cloud missing and noise influence are reduced through the super voxel unit, branches and leaves are accurately separated in combination with multi-dimensional features and random forest robustness, support is provided for tree parameter extraction, three-dimensional modeling and forestry management, and the method is suitable for a complex tree structure scene.

The application relates to the field of artificial intelligence and discloses a content generation method and device based on retrieval and electronic equipment. The method comprises the following steps: determining a first candidate text block from a plurality of preset text blocks based on a user input problem to be solved; determining the first contribution degree of each first candidate text block based on the first candidate text block and the problem to be solved; obtaining the clustering label information corresponding to the first candidate text block by querying a text block clustering tree based on the first contribution degree and the first candidate text block; determining a second candidate text block from the plurality of preset text blocks based on the first contribution degree and the clustering label information; determining target content based on the first candidate text block and / or the second candidate text block, and instructing a preset model to generate an answer. The method introduces a text block clustering tree structure, utilizes a retrieval mechanism based on contribution degree, realizes dynamic focusing and semantic diffusion in the retrieval process, and improves retrieval precision and content generation quality under complex queries.

This application relates to the technical field of digital data processing, specifically to a circuit breaker fault detection method based on electrical parameter analysis. The method includes: obtaining monitoring data and constructing a parameter matrix; screening suspicious elements based on the prediction deviation of elements in the parameter matrix and the mutation point detection result, determining a first eigenvalue according to the starting moment determined by the suspicious elements and the degree of regularity of data fluctuations within the fluctuation duration of the suspicious elements, and then combining the consistency analysis result of irregular data fluctuations in the parameter matrices of different types of monitoring data to determine a fault interval. Furthermore, target data is screened from the residual term, and a plurality of clustering trees are obtained based on the target data using the agglomerative hierarchical clustering algorithm. The fault detection result of the circuit breaker is determined based on the differences between different clustering trees. In this application, residual components are used to replace the initial data, reducing the amount of clustering data while improving the clustering effect of the clustering number on monitoring data that is sensitive to different faults.

The invention provides a satellite communication system user grouping method based on a clustering tree. The method comprises the following steps: setting a clustering detection condition; setting clustering features of the clusters, wherein the clustering features of the clusters are used as the clustering features of the user groups and are used for obtaining the clustering features of the nodes; initializing a user group and a clustering tree; for a current user to be grouped, searching a cluster with the strongest correlation with the feature attributes of the user, and detecting whether the user can join a user group corresponding to the cluster according to a cluster detection condition; correspondingly updating the user group and the clustering tree according to whether the user joins the user group or not; returning to the step of searching the cluster with the strongest correlation; and obtaining a final grouping result until all users are grouped. The method provided by the invention is designed for a satellite network, and can combine the users which meet the clustering detection condition and have similar attributes into the user group through the set clustering features of the clustering tree and the clustering detection condition under the condition that the clustering number does not need to be preset in advance.

This application belongs to the technical field of spectral analysis and discloses a method, device, electronic device, and storage medium for identifying spectral transition times. The method includes: periodically acquiring real-time spectral data of the sample to be tested; employing an agglomerative hierarchical clustering method, combined with Mahalanobis distance and minimum variance method, merging the real-time spectral data into corresponding cluster tree structure data according to the time series; determining the transition point data that meets preset clustering distance abrupt change conditions and preset frequency domain feature jump conditions based on the merging distance in the cluster tree structure data using a sliding window algorithm and a Fourier transform algorithm; verifying the transition point data according to the isolated forest algorithm to determine the time point corresponding to the transition point data of non-abnormal data as the transition time point; the above method improves the identification efficiency of spectral transition times.

The application relates to the technical field of automatic driving, and specifically provides a clustering method, a target contour and type acquisition method, equipment, a medium and a vehicle, and aims to solve the problem of improving point cloud clustering accuracy. The method provided by the application comprises the following steps: voxelizing a three-dimensional point cloud to form a plurality of voxels; establishing an index relationship between each voxel and a first neighborhood voxel with the smallest spatial position in the neighborhood of each voxel; establishing a plurality of clustering trees according to the index relationship of each voxel; and respectively merging a second neighborhood voxel in each voxel and the neighborhood of each voxel into the same clustering tree, wherein the second neighborhood voxel is a neighborhood voxel similar to the multi-source information between the voxels, and the multi-source information comprises result information corresponding to the voxels in the results of each target perception task after the three-dimensional point cloud is subjected to a plurality of target perception tasks. Through the above method, the efficiency and accuracy of point cloud clustering can be improved simultaneously, and the accuracy of target classification according to the clustering result is improved.

The present invention relates to a sensor node load balancing method, device and medium based on clustering tree, comprising: dividing wireless sensor nodes in a network into a multi-layer structure based on a clustering tree; constructing a clustering tree in the network, wherein each node added to the clustering tree will immediately search for its child nodes; using a balancing algorithm to perform multiple rounds of load balancing on the constructed clustering tree, and finally balancing the node loads in the network by running the load balancing algorithm multiple times on each node. The present invention can effectively balance node energy consumption, extend the stability period and life cycle of the network, and improve the throughput of the network.

The present application relates to a method, device, computer equipment and storage medium for classifying web addresses. The method comprises: obtaining a web address to be classified, performing field segmentation on the web address to be classified to obtain a target field corresponding to a first node relationship; obtaining a target clustering structure tree; matching each target field with each node in the target clustering structure tree layer by layer according to the first node relationship; when each target field matches each node in a target path, determining the web address category of the web address to be classified according to the target path; the target path is the path from the root node to the leaf node in the target clustering structure tree. The above scheme can directly implement parallel computing based on the target clustering structure tree to classify the web addresses to be classified, quickly output the web address classification results of the web addresses to be classified, and effectively ensure the efficiency of web address classification.