58 results about "Sequence clustering" patented technology

The application discloses a student intelligent psychological state evaluation system, and relates to the technical field of data communication. First, the quality of student published data is checked and automatically corrected to ensure the reliability of subsequent processing input. Then, data sequence clusters are divided through clustering analysis, and resource scheduling strategies are dynamically adjusted based on the similarity confidence of each cluster to realize the optimization of analysis frequency and sampling rate of different clusters. Meanwhile, the resource scheduling strategy in the cluster is dynamically adjusted, and the optimal task allocation plan is intelligently calculated and issued according to the scheduling exposure index and resource proportion index of each sequence cluster to realize the fine matching of computing power and manpower. Finally, the real-time monitoring of key performance indicators improves the real-time performance and accuracy of student psychological state evaluation, and significantly optimizes the resource utilization efficiency and stability of the system in a high-concurrency scenario.

The invention discloses a construction method and application of a pathogenic microorganism genome database. The method comprises the steps of genome acquisition, genome screening, target region grabbing, strain verification, target region sequence clustering, redundancy elimination and database merging. The construction method of the pathogenic microorganism genome database is suitable for targeted high-throughput sequencing, and the corresponding amplification relationship between the primer and the corresponding microorganism genome sequence is obtained through the operations of screening, filtering, simulation amplification and the like on pathogenic microorganism genome data; the risk of wrong identification of pathogenic species can be reduced by constraining the corresponding relationship.

The invention provides a time sequence clustering method and device, and the method comprises the steps: obtaining subsequences of a to-be-clustered time sequence, taking m subsequences as m clustering centers, and combining the m clustering centers into a clustering center set of the to-be-clustered time sequence; a target function is constructed with the purpose of maximizing the sum of the shape similarities of all the subsequences and the corresponding clustering centers, and one subsequence corresponds to one clustering center; and performing approximate solution on the target function based on a greedy strategy, and accelerating the solution process through a preset effective boundary to determine an approximate optimal subsequence combination in the clustering center set. According to the invention, the time complexity is significantly reduced by selecting the subsequences as the clustering center; a greedy strategy is introduced to enable a suboptimal solution to quickly approach a globally optimal solution, the solving process is accelerated by setting an effective boundary, and the clustering efficiency of a long-time sequence is effectively improved.

The invention provides a heuristic biological sequence clustering method based on semi-global comparison. The method comprises the following steps: firstly, reading all biological sequences, removing repetitive sequences, and carrying out descending sorting on the biological sequences according to sequence lengths; a first sequence is used as a representative sequence of a first category, then a next sequence is read, the similarity between the current sequence and the representative sequence is calculated by adopting a semi-global sequence comparison algorithm, each sequence can be compared to the most similar area in the representative sequence through semi-global sequence comparison, the similarity between the sequences is found to the maximum extent, and the similarity between the sequences and the representative sequence is calculated. Obtaining a relatively high comparison similarity value between the sequences; if the similarity meets a clustering threshold value, adding the similarity into a category with the same representative sequence, otherwise, taking the similarity as a new representative sequence, and generating a new category; repeating the steps until all the sequences are processed; the number of the final representative sequences is the number of clustering categories, and the categories which are the same as the representative sequences are member sequences of each category.

The invention provides a viral genome splicing method and system combining a reference sequence and de novo splicing, and the method comprises the steps: carrying out the quality control of the sequencing data of a viral genome sequence, and removing a host genome sequence in the sequencing data based on a k-mers strategy; splicing the sequencing data from the beginning to generate a preliminary assembling sequence of the viral genome; performing virus classification annotation on the virus genome preliminary assembly sequence based on a blastn tool, and selecting a highly similar virus genome sequence from a virus library as a reference sequence according to the virus classification annotation; generating a data set according to the reference sequence and the virus genome preliminary assembly sequence, dividing a plurality of sequence clusters according to the similarity between the sequences in the data set, and generating a consensus sequence; comparing the reads after the host is removed with a consensus sequence, and adjusting the consensus sequence according to a comparison result to obtain a final virus genome sequence. According to the method, the accuracy of virus genome splicing is improved, and the universality is high.

The invention discloses a reservoir bank slope short-term deformation trend IKHM-EFS-TimeMixer prediction method considering a prediction interval hydro meteorological change. The method comprises the following steps: 1) collecting a reservoir bank slope historical deformation sequence and carrying out clustering distinguishing; 2) an IKHM algorithm is initialized, and IKHM clustering is carried out on the reservoir bank slope variable factor sequences with similar characteristics; 3) constructing a clustering feature analysis matrix of the variable factor sequence; and 4) carrying out EWT decomposition on the displacement sequences in the clustering feature analysis matrix, removing noise sequences determined in each decomposed sequence, and reconstructing each group of displacement sequences. And 5) based on the updated clustering feature analysis matrix and considering the hydro meteorological change of the prediction interval, applying a TimeMixer model to predict the deformation trend of the reservoir bank slope in any step length in the short-term future. According to the method, the clustering result with obvious global characteristics is obtained through optimization clustering, the superposed noise item and the feature item after deformation sequence clustering are distinguished, and the hydro-meteorological change of a certain prediction interval is considered, so that the method has better performance in the aspect of predicting the deformation trend of the bank slope in a short period.

A complex bait data set construction method based on structure and sequence collaborative redundancy elimination belongs to the field of bioinformatics, and comprises the following steps: screening an initial protein complex structure set, removing entries containing nucleic acids, small molecules or non-protein chains, and selecting binary complexes meeting integrity and resolution requirements; secondly, structure clustering and sequence clustering are carried out based on three-dimensional structure similarity and sequence homology, combined comparison is carried out on the two results, and redundant compound entries which are highly similar in structure and sequence are removed; then, taking each cluster representative compound as a target, generating a plurality of groups of bait structures by using a molecular docking or prediction modeling method, and calculating a quality index; and finally, performing stratified sampling and proportion balance based on the score interval of the quality index, and constructing a high-quality protein complex bait data set with structure and sequence collaborative redundancy elimination and balanced quality distribution. The data set generated by the method has the advantages of low redundancy, high diversity and quality distribution controllability.

The invention relates to the field of agricultural data analysis, in particular to a plough layer capacity expansion regulation and control method fused with an organic material carbon nitrogen conversion model, and the method comprises the steps: obtaining carbon nitrogen mineralization rate time sequence original data; performing coupling evaluation on mineralization energy gravity center moment difference and morphological overlapping characteristics to obtain a phase hysteresis correction factor; performing normalized attenuation mapping on the sequence local oscillation energy functional and the total sample average oscillation energy to obtain a transient excitation smoothing factor; performing optimization factor multiplicative correction on the Euclidean distance to obtain a corrected distance and performing clustering iterative calculation; and performing carbon nitrogen conversion model parameter matching and regulation strategy reverse derivation on a clustering result to obtain a plough layer capacity expansion regulation scheme so as to solve the problems of regulation type region division distortion and model matching precision reduction caused by inoculation hysteresis translation and excitation effect high-frequency oscillation in the existing Euclidean distance-based mineralization rate time sequence clustering.

The invention discloses a clustering method, system and equipment based on a sequential sequence and a storage medium, and the method comprises the steps: carrying out the reasonable classification and analysis of tree nodes through modular operation by employing a sequential sequence clustering method based on multi-view learning according to an energy consumption curve; different energy consumption curves are clustered, the evaluation content of each cluster is automatically output, and possible reasons are pushed, so that the operation and maintenance personnel can conveniently check, and the energy consumption check efficiency and accuracy are improved; and therefore, corresponding more reasonable and appropriate energy-saving measures can be formulated. In view of prominent energy consumption problem and high energy cost ratio caused by medical service demand increase and hospital scale expansion, the invention aims to reasonably classify tree nodes according to an energy consumption curve so as to formulate energy-saving measures.

Computerized methodologies are disclosed that are directed to detecting anomalies within a time-series data set. An aspect of the anomaly detection process includes determining one or more seasonality patterns that correspond to a specific time-series data set by evaluating a set of candidate seasonality patterns (e.g., hourly, daily, weekly, day-start off-sets, etc.). The evaluation of a candidate seasonality pattern may include dividing the time-series data set into a collection of subsequences based on the particular candidate seasonality pattern. Further, the collection of subsequences may be divided into clusters and a silhouette score may be computed to measure the clustering quality of the candidate seasonality pattern. In some instances, the candidate seasonality pattern having the highest silhouette score is selected and utilized in anomaly detection process. In other instances, a plurality of seasonality patterns may be combined forming a time policy, where the time policy is utilized in anomaly detection process.

The invention belongs to the technical field of time sequence anomaly detection, and particularly relates to a streaming time sequence sub-sequence anomaly detection method fusing global and local features, which comprises the following steps: clustering all sub-sequences in an initial batch, constructing an inverted index, and performing approximate neighbor search on each sub-sequence; the used calculation formula is used for measuring deviation between the subsequences and neighbors of the subsequences and between the subsequences and cluster centers of clusters containing the neighbors of the subsequences and dispersity of neighbor distribution, and global and local feature fusion is achieved; taking each current cluster as a historical cluster, calculating the maximum distance between each sub-sequence in each historical cluster and the cluster center of the cluster, and taking the maximum distance as a density parameter of the historical cluster; and clustering all the subsequences in the latest batch, calculating the Euclidean distance between the cluster center of each new cluster and the cluster center of each historical cluster, carrying out cluster combination based on the comparison of the Euclidean distance and the density parameter of each historical cluster, and repeatedly executing neighbor search and abnormal score calculation. According to the invention, accurate detection of large-scale streaming data can be realized.

The invention discloses a medicine constant temperature system control optimization method, and mainly relates to the technical field of medicine constant temperature control, and the method comprises the steps: obtaining a pre-arranged temperature sensor array of a target medicine constant temperature area; obtaining a temperature data sub-sequence cluster; obtaining a key temperature value sequence set corresponding to the temperature data sequence set; constructing a key temperature value topology network sequence; performing iterative correlation analysis on the key temperature value topology network sequence according to a time sequence, and determining iterative correlation temperature features; and performing control optimization on the medicine constant temperature system based on the iterative correlation temperature characteristics to obtain control optimization parameters. According to the invention, the technical problems of low control accuracy and slow system response caused by lack of in-depth analysis of temperature in medicine constant temperature system control optimization in the prior art are solved, and the technical effects of effectively reducing the processing complexity of temperature data and improving the accuracy of the temperature control system are achieved.

Computerized methodologies are disclosed that are directed to detecting anomalies within a time-series data set. An aspect of the anomaly detection process includes determining one or more seasonality patterns that correspond to a specific time-series data set by evaluating a set of candidate seasonality patterns (e.g., hourly, daily, weekly, day-start off-sets, etc.). The evaluation of a candidate seasonality pattern may include dividing the time-series data set into a collection of subsequences based on the particular candidate seasonality pattern. Further, the collection of subsequences may be divided into clusters and a silhouette score may be computed to measure the clustering quality of the candidate seasonality pattern. In some instances, the candidate seasonality pattern having the highest silhouette score is selected and utilized in anomaly detection process. In other instances, a plurality of seasonality patterns may be combined forming a time policy, where the time policy is utilized in anomaly detection process.

The invention relates to the technical field of clinical medical treatment, and discloses a disease deterioration risk identification method and system based on depth time sequence clustering enhancement, and the method comprises the steps: 1, obtaining multi-modal clinical sequence data of a patient, including physiological indexes of a time sequence, a laboratory detection result, historical diseases and medication data; 2, performing feature extraction and classification on the patient sequences by adopting a knowledge enhanced sequence clustering method, and grouping the patient sequences according to future outcome distribution of the patient sequences; 3, enabling the model to quickly adapt to a prediction task of a new patient subgroup through a meta-training process; and step 4, based on the trained meta-model, carrying out rapid adaptation on the new patient subtype, and predicting the possibility that the new patient subtype has a deterioration event in a certain time window in the future. The method and the system can effectively learn the disease change mode of the patient under the condition of limited clinical data, improve the prediction accuracy of the new patient subgroup, and are especially suitable for clinical prediction scenes under the condition of small samples.

The invention discloses an elder accompanying intelligent dialogue method and system based on a user portrait, and relates to the technical field of intelligent interaction, and the method comprises the steps: calculating user features based on multi-modal data, splicing the user features into a user portrait vector, generating a frame sequence from voice, calculating an interaction feature vector, constructing an interaction sequence, generating a prefix tree, and extracting all paths. Calculating a support degree, and generating a candidate mode set; extracting an intention label corresponding to each mode in the candidate mode set, generating a frequent mode set, calculating similarity, generating an associated intention label and confidence, extracting frequency features in user portrait vectors, splicing the frequency features into interest feature sub-vectors, calculating joint utility, and selecting a maximum value as an optimal intention. According to the method, through combination of the prefix tree and frequent pattern mining and interaction sequence clustering, the capability of capturing long-term interaction behavior rules of old people is enhanced, and through combination of interest utility and matching utility, the accuracy of intention inference is improved.

The invention belongs to the technical field of data management and intelligent analysis, and particularly relates to a data loading and conversion processing method under a lake-warehouse fusion architecture, which comprises the following steps of: after an externally uploaded file is received, performing compatibility verification and content verification on the externally uploaded file in sequence; identifying that the tables and the partitions which are influenced by data loading and need to be synchronously updated are added, executing atomic data loading operation, synchronously updating the tables and the partitions in the table and the partitions, and generating corresponding version numbers for each table and each partition at the same time; for a data block set under the lake-warehouse fusion architecture, performing hierarchical adjustment according to the access popularity of the data block set; income-cost evaluation is carried out on each data block, and the data blocks needing to be subjected to Z-sequence clustering rearrangement are recognized and added into a requeuing column; and performing Z-sequence clustering rearrangement on each data block in the requeuing column.

This invention relates to the field of agricultural data analysis, and particularly to a method for expanding and regulating the topsoil volume by integrating an organic material carbon and nitrogen conversion model. The method includes: acquiring raw time-series data of carbon and nitrogen mineralization rates; obtaining a phase hysteresis correction factor by coupling and evaluating the temporal differences and morphological overlap characteristics of the mineralization energy centroid; obtaining a transient excitation smoothing factor by normalizing and attenuating the local oscillation energy functional of the sequence with the average oscillation energy of the entire sample; obtaining a corrected distance by performing multiplicative optimization of the Euclidean distance and performing iterative clustering calculations; and obtaining a topsoil volume expansion and regulation scheme by matching carbon and nitrogen conversion model parameters and inversely deriving the regulation strategy from the clustering results. This addresses the problems of distortion in the division of regulation type regions and decreased model matching accuracy caused by inoculation hysteresis translation and high-frequency oscillations of excitation effects in existing Euclidean distance-based mineralization rate time-series clustering.

The invention discloses a single metagenome contig sequence clustering method and system fused with semantic features. The method comprises the following steps: preprocessing metagenome original data to obtain a DNA contig sequence, extracting 4-mer features from a contig, and performing dimensionality reduction on the 4-mer features to obtain composition features of the DNA contig sequence; selecting a training sequence from the DNA contig sequence, further pre-training the DNABERTS model by using the training sequence, and performing semantic feature extraction to obtain semantic features; inputting the composition features and the semantic features into a pre-constructed VAE variational auto-encoder for feature fusion reconstruction, and calculating a reconstruction error; and when the reconstruction error is stable or reaches a predetermined training round, obtaining potential spatial characteristics of the VAE variational auto-encoder, and according to the potential spatial characteristics, completing clustering of the DNA contig sequence through a k-means clustering algorithm. According to the method, the problem that the clustering precision is low when coverage rate features are lacked is solved, and the clustering effect of the single-sample metagenome is improved.

The invention discloses a form-based atmospheric pollution time series data feature fragment mining method and system. According to the method, clustering analysis is carried out by extracting time sequence sub-fragments (u-fragments), so that typical modes in the air pollution process are recognized. Specifically, the method proposes a discrete feature representation method to capture representative local morphological features in atmospheric pollution time series data, and at the same time, uses the similarity between extended Euclidean distance degree quantum sequences, and identifies sub-fragments with pollution identification degree through sub-sequence clustering. Based on the method, the invention further constructs a form-based atmospheric pollution time series data feature fragment mining system. The system can mine pollution fragments with identification degree from atmospheric pollution time sequence data, overcomes the limitation that a traditional statistics and frequency domain analysis method is weak in local feature capturing capability, and can be applied to the fields of urban pollution mode recognition, pollution traceability auxiliary decision making, environmental governance optimization and the like.

The application discloses a nucleic acid sequence clustering method and device, a computer readable storage medium and a terminal. The terminal constructs a tree structure with multiple branches to search a specified interval of a nucleic acid sequence, thereby avoiding a large amount of time consumed by traditional calculation of an editing distance. In addition, the application adopts a node drift algorithm to resist interference caused by errors in the nucleic acid sequence. Compared with existing nucleic acid clustering algorithms, the method provided by the application can cluster a large number of unidentified nucleic acid sequences, and has the functions of automatically correcting and comparing the clustered nucleic acid sequences. The method can directly output the corrected nucleic acid original sequence, thereby greatly reducing the processing time after sequencing reading.

Described herein are methods synchronizing sequencing primers within a sequencing cluster and methods of generating long-range sequencing reads. The methods can include hybridizing primers to polynucleotide copies within a sequencing cluster; extending the primers through a first region of the polynucleotide copies using labeled nucleotides according to a sequencing flow order; extending the primers through a second region of the polynucleotide copies using one or more re-phasing flow steps that each include at least two different types of nucleotide bases; and extending the primers through a third region of the polynucleotide copies using labeled nucleotides according to the sequencing cycle. The rephasing flow steps may be initiated after a predetermined number of sequencing flow steps, after a measured sequencing signal strength falls below a predetermined sequencing signal strength threshold, or a measured sequencing signal-to-noise ratio falls below a sequencing signal-to-noise ratio threshold.

The invention provides a multi-system data communication and service fusion method and system based on a front-end instruction. According to the method, operation events such as clicking, inputting, skipping and data reading of a user in a plurality of service systems are collected at the front end of a browser, a cross-system operation track is constructed, key action extraction, sequence clustering and structural feature recognition are carried out on the operation track, and a reusable service mode template is automatically generated. Based on natural language service requirements of a user, system execution intention recognition and task element extraction, task requirements are mapped to a service template, structured task description is generated, a front-end instruction sequence and a cross-system dependency relationship are constructed according to the structured task description, and an executable instruction diagram is formed. The instruction graph is dynamically executed in the browser. According to the method, the original system interface does not need to be transformed, automatic communication and data collaboration of multi-system business processes can be realized, the manual operation cost is remarkably reduced, and the business processing efficiency and the data consistency are improved.

The invention discloses a three-phase load prediction method based on time sequence clustering, and the method comprises the steps: firstly collecting the data of an intelligent electric meter of a user in a target transformer area, and carrying out the preprocessing of the collected current load data; clustering the preprocessed load data based on a time sequence clustering algorithm to obtain K clusters of data of different load types; and clustering is completed by iteratively updating the clustering centroid. Constructing a space-time diagram neural network according to the clustering result and the three-phase phase information of the user; the space-time diagram neural network is used for carrying out space-time feature extraction and fusion on the clustered load data and outputting a three-phase load prediction result; and finally, training the constructed space-time diagram neural network by using the training data, and predicting the future three-phase load by using the trained model. According to the method, data with the same load type are classified into one class through a time sequence clustering algorithm, and a space-time diagram neural network is built according to three-phase phase information of a user to carry out three-phase load prediction.

The invention provides an abnormal behavior detection method and system for a communication information flow, and relates to the technical field of network communication security, firstly, an original communication information flow set containing massive communication session recording units is captured in a network communication link, and each recording unit carries information such as source and destination internet protocol address tags; performing communication behavior pattern analysis on the original communication information flow set, and constructing a cross-session associated interaction behavior sequence cluster; secondly, extracting features of an interactive behavior time sequence chain, and processing the features through a semantic embedding layer and a time sequence convolution encoder to generate a joint embedded feature vector; and inputting the vector into an abnormal behavior prototype network for prototype matching to obtain an initial abnormal behavior mark. And finally, tracing the abnormal behavior propagation path according to the initial mark to obtain an abnormal behavior propagation path sequence set. According to the invention, abnormal behaviors in network communication can be effectively detected, and the detection accuracy and reliability are improved.

The application discloses a macrovirus group data analysis method, and belongs to the technical field of macrovirology. The method comprises the following steps: sequence quality control, sequence assembly, sequence clustering, virus sequence identification, virus sequence checking, virus abundance calculation, species annotation, virus lifestyle judgment, virus host prediction and virus auxiliary metabolism gene analysis. The application uses Trimmomatic software, BWA-MEN, Megahit software, CD-HIT and other tools to execute the analysis process of macrovirus group data. Practice proves that the application can accurately identify and annotate virus species, comprehensively and systematically deeply analyze and mine macrovirus group data, the steps are simple and clear, the analysis time is short, and the effect of macrovirus identification research is greatly improved.

The invention discloses an academic analysis method and system based on semantic question and answer sequence clustering, and relates to the technical field of online learning analysis and education, and the method comprises the steps: constructing a multi-dimensional semantic question and answer sequence based on the multi-dimensional data of students; based on the multi-dimensional semantic question and answer sequence, using a Word2Vec technology to perform feature representation learning, and generating a semantic feature vector; clustering by adopting a K-means clustering algorithm based on the generated semantic feature vector to obtain a clustering result so as to identify academic performance features of the student group; and based on the clustering result, analyzing different student groups by adopting a statistical method to obtain an analysis result. According to the method, the limitations of an existing method in the aspects of insufficient heterogeneous data processing, single analysis dimension, incomplete student feature understanding and the like are overcome, and targeted teaching strategies and improvement suggestions can be provided for educators.

Aiming at the defects in a DATC-CT method, the invention provides an active time sequence clustering model (DATC-MC) based on multi-level comparative learning, the model fuses a multi-level comparative learning mechanism comprising instance-level comparison and class-cluster-level comparison on the basis of the DATC-CT method, the sample feature representation quality is optimized, and the distinguishing capability of boundaries between class clusters is enhanced. In the consolidation stage of active sampling, the DATC-MC calculates the density of each sample in a representation space by using a KNN algorithm, preferentially selects the sample with the maximum sample density, and stores the sample with the maximum sample density in an annotation class cluster set ACS or a corresponding auxiliary annotation set AAS, thereby generating high-quality ML and CL constraints. And the DATC-MC model uses comparison loss and constraint loss to jointly optimize coding network parameters and a class cluster center. A contrast experiment result on a public data set shows that the clustering performance of the proposed model is superior to that of an existing active time sequence clustering method.

The invention discloses a bank outlet people flow prediction method and a related device, and relates to the technical field of artificial intelligence. Real-time people flow data and associated data are fused to obtain a real-time time sequence of high-dimensional features; and an improved DTW algorithm is adopted to determine a sub-time sequence classification set with the highest similarity degree with the real-time time sequence from the historical time sequence classification set, and then the real-time time sequence is input into the people flow prediction model corresponding to the sub-time sequence classification set to obtain a people flow prediction result. The improved DTW algorithm is applied to the selection of the sub-historical time sequence classification set, the training and use of the time sequence clustering model, and the updating process of the people flow prediction model, and the algorithm considers the distance of the position between the time sequences and the similarity of the direction. And the calculation precision of the similarity between the time sequences is improved, so that the prediction precision of human traffic prediction of the whole bank outlet can be improved.

The invention relates to the technical field of metagenome sequence clustering, in particular to a method, a system and equipment for clustering metagenome sequences based on multi-modal feature fusion and a storage medium. According to the clustering method, a Dirichlet process Gaussian mixture model and a sparse affinity graph model are combined, and in the clustering process, the clustering time is shortened, and the clustering efficiency is improved. The system can dynamically adjust a processing strategy (for example, whether to transfer into a sparse affinity graph model) according to the clustering number, ensures that a final clustering result meets the requirement (K = 1), adapts to metagenome data sets of different scales, realizes efficient and accurate clustering of metagenome data, can fully explore information contained in metagenome sequencing data, and improves the clustering efficiency. Effective clustering is carried out by utilizing biological characteristics contained in the sequencing data to the greatest extent, and the calculation complexity is reduced and the clustering process is accelerated through a hierarchical clustering method.

The invention relates to a self-adaptive two-stage clustering method for maximizing the distance between load sequence clusters and a storage medium, and solves the problems of poor clustering effect and the like in the prior art, and adopts the technical scheme that the method comprises the following steps of: extracting load data, and establishing a load data set; preprocessing the load data set; obtaining a plurality of first regions of interest (ROI); carrying out weighted calculation on the average cluster spacing through DB Index to maximize the average cluster spacing and complete first-stage adaptive clustering; selecting clusters with relatively large average cluster spacing obtained by the first-stage clustering to obtain a plurality of second regions of interest (ROI); and carrying out weighted calculation on the average cluster spacing through DB-Index to maximize the average cluster spacing and complete second-stage adaptive clustering. The method has the technical effects that the complexity in the power load data is effectively dealt with by maximizing the inter-cluster distance of the load data, a typical load data mode can be captured, and a low-density mode with a special value can be identified.