13 results about "Primary clustering" patented technology

The invention discloses a regulation and control quota distribution method and system fusing subjective and objective multi-dimensional features. The method comprises the following steps: acquiring electrical load data, historical response behavior data and subjective response intention information of a user; firstly, a Ward system is used for clustering, users are clustered according to active power, then a primary clustering center is used as an initial clustering center of secondary clustering, FCM clustering is carried out, and a typical load curve of the users is described based on a secondary clustering method; load prediction is carried out based on an NARX neural network, a predicted load curve is compared with a typical load curve, and adjustable potential is calculated; constructing a DR feature data set; and the DR feature data set is fused with an entropy weight method and an analytic hierarchy process to obtain a combined weight, a fuzzy relation matrix is constructed, a user comprehensive score is quantified, and a comprehensive response potential score of the user is formed.

The invention provides an existing public building performance quantitative evaluation method and system, and the method comprises the steps: obtaining existing public building performance multi-source evaluation data, carrying out the screening, obtaining a performance evaluation index set, carrying out the primary clustering, obtaining a primary clustering result, carrying out the secondary clustering of the primary clustering result based on a dual-channel clustering algorithm, and obtaining a performance evaluation result. A secondary clustering result is obtained, a performance evaluation index system is constructed according to the secondary clustering result, a building performance evaluation model is constructed and pre-trained, a performance evaluation index evaluation rule is established, and a self-adaptive deployment result of the performance evaluation index system is obtained. And generating a performance evaluation result according to the self-adaptive deployment result and the performance evaluation index evaluation rule, obtaining existing public building data, performing building performance evaluation, obtaining the performance evaluation result, and performing verification and rule optimization on the performance evaluation result. Therefore, the comprehensive value of the existing public building performance is objectively and accurately evaluated.

A database system is operable to generate an IO pipeline that includes a primary cluster key pipeline element serially after a secondary index pipeline element. The IO pipeline is executed based on applying the secondary index pipeline element of the IO pipeline to generate row list output, and applying the primary cluster key pipeline element to determine a first set of row ranges, determine a second set of row ranges of row list output generated by the secondary index pipeline element, and generate, from the first set of row ranges and the second set of row ranges, a result set having a plurality of outputs, each indicating a cluster key and a row range for the cluster key.

The application belongs to the field of hydrological prediction, and particularly relates to a combined machine learning annual runoff prediction method based on secondary decomposition clustering, which comprises the following steps: decomposing an annual runoff sequence into multiple intrinsic mode function (IMF) components and a residual sequence; clustering the intrinsic mode function components to synthesize high-frequency components, medium-frequency components and low-frequency components; performing secondary decomposition on the high-frequency components to generate multiple new intrinsic mode function (IMF') components; performing second clustering reconstruction on the residual sequence generated by the primary decomposition, the medium-frequency components and the low-frequency components generated by the primary clustering, and the IMF' components generated by the secondary decomposition to generate intrinsic mode function (IMF'') components; constructing an annual runoff prediction model based on a snowmelt algorithm optimized long short-term memory network model, and respectively predicting the intrinsic mode function (IMF'') components obtained by the secondary clustering based on the annual runoff prediction model. The application can be used to improve the prediction accuracy of the annual runoff sequence based on the machine learning model through the secondary decomposition and clustering processing of the annual runoff sequence.

The application relates to the technical field of file analysis, and particularly discloses a file analysis method and system applied to an ofd reader, which comprises the following steps: performing primary clustering on a page based on a content distribution array, obtaining the structural features and content features of each page after primary clustering, performing secondary clustering on the page according to the structural features and content features, synchronously determining the analysis score of each page according to the primary clustering process and the secondary clustering process, selecting the page according to the analysis score, executing an analysis process, constructing a simplified analysis process of the same type of page after secondary clustering based on the analyzed content, and performing gradient analysis on each page based on the simplified analysis process. The ofd file is initially identified, the page is clustered according to the initial identification result, and on the basis of executing a traditional analysis process, a pre-simplified analysis operation is performed on the same type of page according to the analysis parameters of the processed page, so that the analysis efficiency is greatly improved.

The invention discloses a millimeter wave radar point cloud clustering method and device, electronic equipment and a storage medium. The method comprises the following steps: carrying out region segmentation on point cloud data to obtain at least two sub-regions; executing a preset clustering algorithm on the point clouds in each sub-region in parallel to obtain a primary clustering result corresponding to each sub-region; wherein the preset clustering algorithm at least comprises a DBSCAN clustering algorithm; determining a target boundary point located in the boundary region after primary clustering in each sub-region; performing clustering processing on all the target boundary points according to a preset clustering algorithm to obtain a secondary clustering result; and combining the primary clustering results and the secondary clustering results to obtain a final clustering result. According to the technical scheme, region segmentation and parallel calculation are carried out on the point cloud data, so that the processing time consumption is remarkably reduced, and the vehicle-mounted real-time requirement can be met; and secondary clustering is carried out by extracting boundary points, so that splitting caused by independent clustering of the sub-regions is eliminated, and the consistency and accuracy of final clustering results are ensured.

This invention discloses a sequence recommendation method based on multi-scale encoding and capsule intent purification. The method acquires and preprocesses user dynamic interaction sequences, and extracts logically related sequence pairs based on the principle of prediction target consistency. Using a multi-scale segment mapping method, the preprocessed user dynamic interaction sequences are deconstructed into behavioral segments of different granularities, and features are extracted and weighted to form a fused sequence representation. The generated fused sequence representation is then subjected to primary clustering, and the dynamic routing mechanism of the capsule network is used to perform nonlinear mapping and semantic enhancement on the cluster centers, generating a purified high-order intent capsule set. A joint loss function is constructed to uniformly optimize the main recommendation task and the multi-dimensional self-supervised auxiliary task. The parameters of the multi-scale encoder, capsule network, and item embedding matrix are synchronously updated using the backpropagation algorithm until the recommendation model converges, resulting in a trained recommendation model. Recommended sequences are then obtained based on the trained recommendation model.

The invention relates to the technical field of file analysis, and particularly discloses a file analysis method and system applied to an ofd reader, and the method comprises the steps: carrying out the primary clustering of pages based on a content distribution array, and obtaining the structural features and content features of each type of pages after the primary clustering; performing secondary clustering on the pages according to the structural features and the content features, and synchronously determining an analysis score of each page according to a primary clustering process and a secondary clustering process; selecting pages according to the analysis score, executing an analysis process, constructing a simplified analysis process of similar pages after secondary clustering based on the analyzed content, and performing gradient analysis on each page based on the simplified analysis process; according to the method, the ofd file is subjected to initial identification, the pages are clustered according to the initial identification result, and on the basis of executing the traditional analysis process, the similar pages are subjected to preposed simplified analysis operation according to the analysis parameters of the processed pages, so that the analysis efficiency is greatly improved.

The invention discloses a photovoltaic cluster power prediction method based on step-by-step spatial feature clustering and an improved graph attention network, and relates to the photovoltaic field, and the method comprises the following steps: constructing a distributed photovoltaic cluster data set, the method comprises the following steps: performing primary clustering division by taking the physical characteristics of a photovoltaic module as characteristics to be input into an affinity propagation AP algorithm, then performing secondary clustering division by taking a solar altitude angle sequence as characteristics to be input into the AP algorithm, and finally dividing a photovoltaic cluster into a plurality of sub-clusters; for each photovoltaic sub-cluster, sorting and merging historical power and historical meteorological time sequence data, and inputting the historical power and historical meteorological time sequence data into a GAT-Encoder-Decoder deep learning model for training; reasoning and outputting a day-ahead power prediction result of each power station in the sub-cluster; and accumulating the prediction results of all the power stations to obtain a power prediction result of the whole photovoltaic cluster. According to the invention, clustering calculation is carried out through step-by-step spatial features so as to obtain a sub-cluster division result which can better reflect the spatial feature state of the photovoltaic power station, and the correlation of the output of the photovoltaic power station in the sub-cluster is improved.

An object-tracking apparatus using a LiDAR sensor of an embodiment includes: a LiDAR sensor configured to acquire a point cloud related to an object; a first clustering unit configured to primarily cluster the point cloud; and a second clustering unit configured to secondarily cluster the result of primary clustering in response to the time taken to perform primary clustering.

The invention is suitable for the technical field of data processing, and provides a new energy power station grid-connected performance comprehensive evaluation method based on multi-level coordination control, and the method comprises the steps: obtaining grid-connected operation data, equipment state data and historical operation data of a new energy power station; generating primary clustering data, and performing multi-dimensional fusion based on dynamic weight distribution to generate target evaluation data; performing feature extraction to generate a spatio-temporal feature vector representing a current operation state and a spatio-temporal feature vector representing a historical operation mode; determining the current operation mode characteristics of the equipment based on the association mapping of the two; inputting the historical operation data conforming to the current operation mode characteristics of the equipment into the performance evaluation model for iterative optimization to obtain an updated performance evaluation model; and performing comprehensive evaluation through the updated performance evaluation model, and outputting a grid-connected performance grade result. The equipment operation mode is accurately analyzed, the evaluation accuracy is improved, and a scientific data basis is provided for power station operation management.

The application discloses a partial discharge diagnosis method based on regional clustering, which firstly extracts pulse repetitive discharge spectrum data and corresponding labels in a partial discharge training set and stores the data as a mapping matrix. Through a primary clustering function, corresponding selected region coordinate matrices are obtained. The selected region coordinate matrices are classified according to the labels, and the result is stored in a clustering data dictionary. The labels and the selected region coordinate matrices in the clustering data dictionary are iterated, and are sent into a secondary clustering function to obtain final diagnosis region coordinate matrices. Whether the collected PRPD spectrum data to be diagnosed is faulty is judged in combination with a discharge frequency threshold, a corresponding fault occurrence probability is calculated, and finally a partial discharge fault diagnosis result containing the fault occurrence probability is generated.

The invention belongs to the technical field of network security, and relates to a network risk information early warning method and system based on cascade aggregation and reinforcement learning. The method comprises the following steps: identifying valid data in network information, and judging the emotional tendency of the valid data; performing primary clustering on the network information with negative tendency, clustering the network information with similar content expressions into the same category, and extracting center point data of different category clusters; performing secondary clustering on the center point data of different clusters to obtain a network information set; for the network information set after secondary clustering, adopting an early warning classification model based on a generative large language model to judge whether early warning is needed or not from the dimensions of two aspects of propagation features and early warning matching features, and pushing early warning content to a user; and after the early warning content is generated, updating an early warning classification model and a secondary clustering threshold by using a reinforcement learning mechanism. According to the invention, network risk information early warning requirements of users in different scenes can be met, and the efficiency of network information analysis can be improved.