184 results about "DBSCAN" patented technology

The invention relates to the field of gas leakage identification, and discloses a gas leakage identification method and system based on sound localization, a medium and equipment, and the method comprises the steps: collecting a leakage sound wave signal through a microphone array, and extracting acoustic features in a complex background; simulating a diffusion process of gas in a turbulence environment after gas leakage through an established leakage gas convection-diffusion model, and combining gas concentration distribution in the simulated diffusion process with acoustic characteristics to predict sound source localization at a leakage position; modeling a sound source localization search process as POMDP, and iteratively updating a confidence state of a sound source position through a particle filter; spatial distribution features are extracted from the confidence state through DBSCAN clustering to serve as input of the LSTM-DQN network, spatial and temporal features are fused through the constructed LSTM-DQN network, cross-scene generalization is achieved through transfer learning, and sound source coordinates are dynamically optimized and output through the Q network. According to the invention, the positioning accuracy and real-time performance in a complex environment are improved.

The invention belongs to the technical field of integrated circuits, and particularly relates to a time mismatch extraction method based on parallel windowing autocorrelation. According to the method, firstly, clustering processing is carried out on signal points; thirdly, preliminarily screening out legal signals based on the time occupancy rate of the signals, and constructing an initial legal signal library; a similarity score between the signals is further calculated through an IDK algorithm, and adjacent frequency bands are divided through a Pettitt method; and finally, continuation of legal signals is realized by means of a DBSCAN clustering algorithm, and automatic construction of a legal signal library is completed. The method is not only suitable for scenes without prior information, but also can be applied under the condition with partial prior information, and can be dynamically updated according to actual requirements. The method is based on radio frequency spectrum monitoring big data, can construct a legal signal library with high reliability and strong generalization ability, provides effective support for subsequent illegal signal monitoring, and has important application value.

The invention relates to the technical field of echelon utilization of decommissioned lithium ion batteries, in particular to a decommissioned power lithium battery sorting method and system based on DBSCAN and WOA-FCM. The method comprises the steps that static characteristics of decommissioned power lithium batteries are determined; based on the static characteristics, first-stage sorting is carried out on the retired power lithium batteries through a density-based noise application space clustering algorithm, so that abnormal batteries are eliminated, and the clustering number of normal batteries is determined; and for the remaining normal batteries after the abnormal batteries are removed, dynamic characteristics are extracted based on a battery discharge voltage curve, second-stage sorting is carried out through a whale optimization-based fuzzy C-means clustering algorithm, and a battery sorting result is obtained. Therefore, through two-stage collaborative optimization, the sorting accuracy and efficiency are improved, and the high consistency of the recombined battery packs is ensured.

The invention provides a gas field reservoir lithology earthquake prediction method based on CCIDI cleaning and TransMSAE, and the method comprises the steps: constructing a CCIDI cleaning module, and employing a well-divided lithology-divided strategy: employing IQR to remove statistical anomalies, employing DBSCAN to remove boundary noise, employing Isolation Forest to remove high-dimensional residual noise, and forming a high-quality logging data set; the method comprises the following steps: constructing a TransMSAE model: pre-standardizing a Transform encoder to capture long-range dependence, and providing robust time sequence representation; multi-scale convolution is introduced into the MSCCAM to extract a local-global mode, and cross attention is fused to improve complex deposition dynamic perception; aMFEM gating residual self-adaptively enhances minority class features, and minority class recognition is remarkably improved; designing NAFL-ProtoLoss mixed loss: NAFL focus inhibition easy-to-classify samples, ecological niche punishment over-confidence prevention, ProtoLoss zoom-in class inward pushing-out between classes, and optimization of unbalanced classification; finally, training and reasoning are carried out based on the model, and high-precision lithology prediction is achieved. The method provided by the invention has high generalization and accuracy, and has excellent reservoir prediction performance in a complex geological environment.

The invention relates to a manual driving vehicle overspeed behavior identification and intervention method based on a detection carrier, and the method comprises the steps: building a detection carrier system, collecting adjacent HV point cloud data through a laser radar, carrying out the median average filtering, K-Means and DBSCAN combined clustering preprocessing, and achieving the multi-target dynamic tracking through a Point Pill + DeepSORT model; and calculating the real average speed of the HV by using a 1-3s self-adaptive speed measurement window, and comparing the road speed limit to complete the overspeed identification. And when overspeed is judged, dynamically triggering a corresponding high-speed camera to capture, extracting license plate information, fusing speed and GPS data to generate illegal evidence. The method comprises the following steps: quantifying the influence of a detection carrier on an HV individual overspeed behavior, training an HV swarm agent based on VAE-WGAN data enhancement and a double-branch deep Q network, simulating multiple traffic scenes in SUMO, and determining an optimal deployment scheme of the detection carrier by using a dynamic Gaussian mixture Bayesian network and a double-layer double-target optimization model. The method breaks through the limitation of traditional speed measurement, improves the recognition precision and deterrence effect, and balances the traffic safety and efficiency.

The invention discloses an analysis method and system based on AI video recognition behavior monitoring, and relates to the technical field of video recognition, and the method comprises the steps: collecting video data, calculating the overall pixel gradient change of adjacent frames, marking preliminary candidate frames, calculating the edge saliency intensity of each frame, and analyzing a salient region point set; clustering and calculating the geometric center point of each cluster as an anchor point value, and forming an anchor point set through the center points of each cluster; according to the method, concentrated extraction of salient points is realized through a density-based clustering algorithm DBSCAN, so that an extraction result has logic uniqueness and local integrity of a salient target, a local feature map is extracted by using a shallow convolutional network, target texture and edge information is effectively captured, bounding box optimization is realized by rotating an IoU loss function, and the robustness of the target is improved. And the fitting precision of the boundary of the salient region is improved.

The invention discloses a robot sensing and autonomous navigation method and system for a three-dimensional topological map, and the method comprises the steps: obtaining the three-dimensional point cloud data of a surrounding environment, the acceleration and angular velocity data of a robot, and predicting the pose of the robot; identifying a free space and an obstacle area, determining an occupation state of each point cloud voxel, and generating a three-dimensional occupation map; dividing a free space into a plurality of convex clusters through neighborhood search based on a DBSCAN algorithm, and constructing a three-dimensional topological graph; identifying a vertical channel, and setting corresponding edges to be connected with portal nodes to form a global three-dimensional topological graph; and according to the global three-dimensional topological graph, determining a global path and a local obstacle avoidance path, and combining the global path and the local obstacle avoidance path to obtain a global optimization path. The technical bottlenecks that a two-dimensional map lacks a vertical dimension and a traditional voxel method is insufficient in real-time performance are broken through on the whole, and an efficient, safe and coherent solution is provided for autonomous navigation of a robot in a complex three-dimensional environment.

The invention relates to an ultrasonic detection and damage evaluation method and system for a curved-surface composite component, and the method comprises the steps: collecting three-dimensional point cloud data of a surface damage region of the curved-surface composite component, and constructing a detection path point sequence; the industrial robot and the ultrasonic phased array are controlled to move and scan synchronously according to time, and ultrasonic A scanning signals and robot tail end pose information are collected; three-dimensional space coordinates of all reflection points in the curved surface composite material component are calculated in combination with pose derivation, and an initial three-dimensional damage point cloud is constructed; setting a segmentation threshold to construct an initial segmentation damage point cloud; a traditional DBSCAN clustering algorithm is improved for clustering segmentation, and point clouds of all independent damage areas are obtained; and voxelizing the point cloud of the independent damage area into a three-dimensional voxel model to realize quantitative assessment of the damage. According to the method, automatic and high-precision three-dimensional damage detection and evaluation of the complex curved surface component are realized, and the problems that a traditional ultrasonic detection method depends on manual operation and is poor in adaptability, the detection result is not visual and the like are effectively solved.

The invention discloses a sea route network construction method based on multistage grid clustering, and the method comprises the steps: setting a navigation region, collecting navigation data, carrying out the noise reduction, preprocessing and segmenting, and obtaining processed data; a space grid structure and a KD-tree structure are introduced on the basis of the DBSCAN algorithm to form a GBSC algorithm; a dynamic threshold mechanism is introduced on the basis of a Douglas-Peucker algorithm, and an adaptive DP algorithm is obtained; screening, clustering and correcting the trajectory segments and the trajectory points by using a GBSC algorithm and an adaptive DP algorithm to obtain a feature point set; a feature point closest to the track point is found in the feature point set through a KD-tree structure, a feature point sequence is obtained, repeated nodes are compressed to obtain a node sequence, adjacent nodes are connected, and an edge set is obtained to construct an adjacent matrix and an edge weight matrix to form an air route network; the method can truly reflect the actual navigation structure of the ship in the sea area, has a high coincidence degree with the original AIS trajectory, and significantly improves the accuracy and reliability of air route extraction.

The invention discloses an NRLC enhanced two-stage DBSCAN clustering rock discontinuous surface identification method, and relates to the field of rock mass structural surface intelligent identification, and the method comprises the steps: S1, identifying concave, convex and boundary feature points in a rock point cloud; s2, calculating a rock point cloud normal vector based on local fitting; s3, based on the rock point cloud normal vector, identifying a group discontinuous surface through DBSCAN clustering; s4, DBSCAN clustering is carried out based on the group discontinuous surfaces to obtain all discontinuous surfaces in the point clouds, automatic recognition and accurate representation of the rock mass discontinuous surfaces are achieved through normal vector clustering analysis, a set of efficient and accurate rock mass structural surface intelligent recognition technology system is established, and reliable three-dimensional data support is provided for rock mass stability evaluation.

The invention discloses an artificial intelligence-driven urban potential updating area identification and division method, which comprises the following steps of: firstly, constructing an index system according to three dimensions of building space, industry form composition and crowd vitality by using the collected building number, land function data, POI (Point of Interest) industry form data, LBS (Location Based Service) crowd data, high-definition remote sensing image data and road network data; the method comprises the following steps: carrying out hierarchical classification identification on an update object through a natural discontinuity point method and a K-Means clustering algorithm, carrying out intelligent division and optimization on a city update unit through a DBSCAN clustering algorithm and a reinforcement learning model of a graph network structure database, and finally carrying out division and boundary optimization on a city update area. According to the invention, automatic identification and division of potential updating objects, updating units and updating areas of the city can be realized, and digital and intelligent application conversion of city planning work is promoted.

The invention provides a comprehensive analysis method for compaction quality of a lime treated soft soil roadbed based on intelligent compaction. The method comprises the following steps: S1, acquiring vibration modulus data corresponding to a plurality of detection units respectively; s2, calculating a compaction degree passing rate and a stability passing rate based on the vibration modulus data corresponding to the plurality of detection units, obtaining a uniformity passing rate by adopting a multi-scale variation structure analysis method, and calculating an abnormal aggregation index through a DBSCAN clustering algorithm; s3, determining the weight of each index, wherein the weight comprises the weight of the compaction degree passing rate, the weight of the uniformity passing rate and the weight of the stability passing rate; s4, calculating a compaction quality value based on each index and the corresponding weight; s5, the roadbed compaction quality grade is judged based on the compaction quality value, and roadbed repairing is conducted based on the roadbed compaction quality grade and the abnormal aggregation index. Compared with the prior art, the method has the advantages of improving the spatial resolution and defect identification precision of uniformity detection of the roadbed compactness and the like.

The invention discloses a DBSCAN unmanned aerial vehicle path planning method based on depth estimation and optical flow analysis, and relates to the technical field of unmanned aerial vehicle autonomous navigation, and the method comprises the steps: extracting an image depth feature through a VGG16 model, carrying out the enhancement and segmentation of an image through Gamma transformation and a region growing algorithm, optimizing a seed group through optical flow analysis and a conditional random field algorithm, and carrying out the optimization of the depth feature of the image. A DBSCAN algorithm is used for carrying out path clustering, an A * algorithm is used for carrying out obstacle avoidance path planning, and a double-loop PID control algorithm is combined for execution. According to the invention, through fusion of multi-modal data and path clustering optimization, navigation precision and robustness in a complex environment are improved, seed group dynamic updating is realized based on seed probability calculation and optical flow analysis of a conditional random field, sensing precision, environmental adaptability and real-time reaction capability are improved, and through combination of double-loop PID control and an A * algorithm, the real-time response capability of the system is improved. And the real-time obstacle avoidance capability is optimized.

This invention discloses an improved DBSCAN algorithm and a many-to-many matching algorithm for 6G vehicle-to-everything (V2X) networks, belonging to the field of V2X technology. It designs an improved density-based spatial application noise clustering algorithm to cluster more dispersed vehicles, employs a many-to-many matching algorithm to pair vehicles with auxiliary MEC servers, and designs a multi-agent-based task offloading mechanism to reduce latency and improve resource utilization efficiency. The improved DBSCAN algorithm proposed in this invention overcomes the sensitivity of parameter selection, does not require particularly precise parameter settings, performs well on datasets with uneven density, overcomes the influence of vehicle density on clustering results, and the proposed many-to-many matching mechanism improves the utilization of remaining vehicle computing resources and greatly alleviates the pressure on edge servers, while ensuring that high-priority tasks are computed first.

The embodiment of the invention provides a call quality detection method and device, electronic equipment and a storage medium, and is applied to the technical field of communication. The method provided by the embodiment of the invention is applied. The method comprises the following steps: acquiring a single call MOS value, a single call packet loss rate, single call average jitter, single call average delay, single call maximum jitter, a single call packet loss rate peak value, a single call codec type, a single call network type and a single call duration of each call of a user in a preset historical duration; and calculating the first data and constructing a user feature vector to obtain the user feature vector representing the call quality of the user. And respectively analyzing the feature vectors of the users through a K-means clustering algorithm and a DBSCAN clustering algorithm, identifying abnormal call users, and clustering the users, so that the users with the same or similar call quality are clustered in the same cluster. Therefore, the long-term call quality of the user is judged.

The application discloses a kind of based on user behavior portrait's layered content propaganda push effect quantitative evaluation method and system, including the construction fusion basic attribute, behavior, demand and effect data multidimensional user portrait system, dynamic update user rule of law literacy label;Design double-layer push algorithm, first round is matched group scene content through rule engine, second round is generated personalized content set through weighted cosine similarity and knowledge weak point incentive factor;Build three-level evaluation system, combine entropy weight method and subjective adjustment coefficient to calculate quantitative score of effect of popularization of law;The evaluation result is inputted push algorithm and portrait update in reverse, realize the closed-loop optimization of push strategy;And based on TFIDF and DBSCAN algorithm, identify regional high-frequency legal risk points, generate early warning report and supporting popularization of law content;The application solves the technical problems, such as single portrait dimension, low push matching degree, effect is difficult to quantify, regional governance support is insufficient, and is suitable for multi-terminal popularization of law scene.

The application provides an automatic driving rear-end collision and sudden state prediction method based on space-time occupation characteristics, and relates to the field of automatic driving.The application proposes a prediction network based on a space-time graph Transformer, solves the problem of lack of space-time occupation characteristics in automatic driving vehicle dangerous scene analysis.Through integration of multi-dimensional characteristics from a nuScenes data set, a rear-end near-collision scene knowledge graph RNSKG is constructed, and DBSCAN unsupervised clustering is used to classify dangerous states and emergency states.A space-time graph Transformer model coupled with a time sequence and a space attention mechanism is significantly better than a traditional model in predicting dangerous states and emergency states, verifying the effectiveness in capturing complex space-time interactions.

The application discloses a kind of based on data-driven wasserstein fuzzy set active distribution network distribution robust day-ahead scheduling method, realize the construction compact DRO fuzzy set, and effectively reduce the conservativeness of day-ahead scheduling result;The method constructs the wasserstein condition generated adversarial network model (CWGAN-GP) based on gradient penalty norm, for wind, light output day-ahead scene generation, and proposes the abnormal sample identification method of improved DBSCAN clustering combined with CNN-BiGRU automatic encoder, to improve the credibility of generated scene set;Adopt the compact boundary of data support set determined based on non-parametric kernel density estimation (NKDE) confidence interval, and combined with wasserstein metric to construct DRO fuzzy set.The application compared with the existing DRO day-ahead scheduling method, realizes the deep combination of data-driven method and DRO model, effectively reduces the conservativeness of fuzzy set, and improves the economy of day-ahead scheduling scheme and the adaptability of coping with new energy output uncertainty under the premise of guaranteeing decision robustness.

The application discloses a meter box line loss compensation method and system based on a neural network. The method first collects voltage, current, power supply end electric energy and power supply end electric energy, and calculates the abnormality degree and line loss rate of the meter box accordingly; then a data point set is constructed, DBSCAN clustering based on abnormality density weighting is performed, and multiple clustering clusters are formed; based on the abnormality distribution proportion vector similarity in each clustering cluster, the abnormality degree level is adaptively divided, and the variance of the line loss rate in each level is calculated as the line loss sensitivity; a long short-term memory neural network model is constructed, and an adaptive sample weight based on the line loss sensitivity is introduced into the loss function for training; finally, line loss compensation is performed according to the model output. The application retains key sparse samples through weighted clustering, adaptively divides abnormal levels and quantifies compensation difficulty, and combines sensitivity weighting training, so that the precision and robustness of line loss compensation in a complex scene are significantly improved.

The application discloses a kind of road width automatic calculation and quality checking method of road network data, it is related to road network data processing technical field, to solve the problem of low efficiency of multi-source data fusion in existing method;The application is through collecting multi-source road network data and pre-standardization processing, unified coordinate system mapping and metadata annotation generate labeled original dataset;Perform noise adaptive filtering and occlusion detection, utilize U-Net semantic segmentation to identify interference area, and apply GAN local completion output purification dataset;Construct the multi-modal network calculation dynamic fusion weight of attention mechanism, and parallel integration forms fusion feature vector;Through Canny optimization edge extraction and DBSCAN clustering calculation width parameter, embedding topological structure;Use multi-scale consistency index to check result, if threshold backtracking optimization;Output result and iteratively update parameter, form closed loop mechanism.The method improves fusion efficiency and calculation accuracy, applicable to traffic navigation and urban planning, supports real-time dynamic application.

A kind of distributed power supply power prediction method, system and equipment based on LSTM, method is first obtained the meteorological data of distributed power supply target area, calculate theoretical electric power data, then based on DBSCAN algorithm clustering analysis is carried out, and the dimension of data is expanded, obtain normal value and abnormal value data set, then the loss function of LSTM prediction network is optimized, and the future normal value and abnormal value prediction result is obtained, finally set parameter adjustment condition, repeat the above steps until completion prediction;The present application is aimed at the problem that the type of historical electric power data is various, and the abnormal value is much, the distance definition of DBSCAN clustering algorithm is improved, the distance between data point and theoretical output power is combined, the stability of normal value extraction is improved, and LSTM prediction network is introduced, the data limitation of only predicting normal value is avoided, the loss function is improved by combining physical constraint and theoretical power, and adaptive adjustment parameter is optimized, so that the algorithm is more efficient.

The invention discloses a meat safety risk early warning model based on AI fast detection data and a service method, and relates to the field of food risk control. By extracting four types of risk features of time, space, category and circulation, an LSTM + DBSCAN + XGBoost fusion model is constructed, and the early warning advance period is longer than or equal to 24 hours; a regional thermodynamic diagram is output for the government, a batch prevention and control scheme is pushed for an enterprise, the supervision efficiency is improved by 90%, and the annual average loss of the enterprise is greater than or equal to 80,000 yuan. The method solves the problems of passive response, low data value and single service, realizes active prevention and control of meat safety, and is suitable for government supervision and enterprise quality control scenes.

The invention belongs to the technical field of calculus risk assessment, and provides a calculus risk prediction method based on machine learning and electronic equipment, and the method comprises the steps: to-be-detected original data collection and calculus risk prediction; the construction process of the stone risk prediction model comprises the steps of unmarked data collection, image feature extraction, index feature extraction, text semantic extraction, feature fusion, feature clustering analysis, iterative training, fine adjustment small sample data collection, meta-learning fine adjustment and cluster marking. According to the method, through a semi-supervised learning strategy of unmarked data clustering and small sample adjustment, the demand of marked data is greatly reduced; according to the method, clustering and unsupervised iteration are fused through DBSCAN and DPC double algorithms, so that pre-training under unmarked data is realized, and the reliability of unmarked training is improved; through combination of multi-modal data fusion, clustering iteration and meta-learning fine tuning, the robustness of the model is improved.

A transformer partial discharge positioning detection method based on ultrahigh frequency signals is characterized by comprising the following steps: S1, estimating time delay of signals obtained by an ultrahigh frequency (UHF) sensor array based on a generalized cross-correlation method; and S2, arranging and combining signals of the UHF sensor array, and solving coordinates of a positioning point based on an improved multi-moth optimization (MFO) algorithm according to the time delay obtained in the step S1. And S3, clustering the coordinates of the plurality of positioning points obtained in the step S2 based on a density clustering algorithm (DBSCAN), and selecting the geometric center position of the category with the maximum sample number as the final local discharge source coordinate.

The invention belongs to the technical field of geophysical exploration, and discloses a mine transient electromagnetic anomaly boundary identification method based on DBSCAN clustering. The method comprises the following steps: step 1, establishing a three-dimensional geophysical model; 2, calculating apparent resistivity data by using a small loop transient electromagnetic method; 3, drawing a k-distance map, and obtaining an optimal neighborhood radius according to the k-distance map; step 4, performing DBSCAN clustering on the apparent resistivity data to obtain a clustering result; 5, replacing the intra-cluster apparent resistivity value with a cluster center value to obtain replaced apparent resistivity data; step 6, imaging the replaced apparent resistivity data and identifying an abnormal body; 7, comparing the anomalous body boundary with the anomalous body boundary in the model, and if the difference is too large, adjusting the optimal neighborhood radius and returning to the step 4; and otherwise, taking the current abnormal body as a final identified abnormal body. The method can effectively identify irregular anomalous body boundaries of non-uniform complex stratums, and can adapt to real-time identification requirements of underground anomalous bodies.

The invention provides a Lamb wave benchmark-free damage imaging method based on adaptive clustering and semantic weighting, and belongs to the technical field of air coupling ultrasonic detection.The Lamb wave benchmark-free damage imaging method comprises the steps that a full-path response signal is collected through orthogonal scanning, a wavelet low-frequency approximation coefficient and a normalized symmetric difference factor are extracted to construct a joint feature vector; and performing unsupervised division of'health-damage 'states on a scanning path by adopting DBSCAN clustering. And semantic tags obtained by clustering are further used as adaptive weights to be embedded into the improved RAPID probability imaging model, and dynamic construction of artifact path suppression and soft reference is realized. In a layering defect detection experiment of the carbon fiber reinforced composite material, compared with a traditional RAPID method, the average size measurement error is reduced by 50.5% under the condition that an independent health reference is completely not needed, the measurement precision of 40 * 20 * 0.05 mm defects in the X / Y direction is improved by 81.7%, the measurement precision of 40 * 20 * 0.05 mm defects in the Y / X direction is improved by 65.5%, and meanwhile 92.3% of boundary artifacts are effectively restrained.

The invention discloses an SAR (Synthetic Aperture Radar) image geometric correction method and system for optimizing a radial basis function neural network based on a genetic algorithm. The method comprises the following steps: normalizing three-dimensional geographic coordinates of ground feature points, and inputting the normalized three-dimensional geographic coordinates to a full connection layer for feature extraction; constructing a ReRBF network, wherein the ReRBF network is formed by stacking a plurality of layers of RBF sub-modules through a residual connection structure; and the output of the previous module is used as a residual error, is spliced with the original input or the intermediate feature, and is jointly used as the input of the next module to form a progressive feature enhancement and residual error learning mechanism, so that the network can refine the coordinate mapping relationship layer by layer. And a DBSCAN algorithm is adopted to perform automatic clustering on training data, so that the generalization ability of the model in multiple scenes is improved. And optimizing the ReRBF network by using a genetic algorithm, and adaptively determining the optimal width parameter of the radial basis function network. According to the method, the geometric positioning precision and the processing efficiency of the SAR image under the complex terrain condition are improved.

The invention provides a traffic risk assessment method, system and device, and a storage medium. The method comprises the following steps: performing DBSCAN clustering on each first alarm point on a first road section according to a pre-configured coordinate generation rule to generate a plurality of first clustering categories; and executing the following operations on each first clustering category: establishing a first scoring system according to the alarm data in the first clustering category and generating a first scoring result. The method can accurately identify accident black spots, dynamically adapts to a non-uniform road network structure, is high in real-time performance, and supports management decisions.

The invention provides a PM2.5 prediction method based on meteorological factor space partitioning, which performs scientific partitioning on monitoring stations based on spatial distribution characteristics of meteorological factors through an ST-DBSCAN clustering algorithm, divides stations with similar meteorological conditions into the same partition, fully considers spatial heterogeneity of the meteorological factors, and improves the accuracy of PM2.5 prediction. Noise interference caused by mixed modeling of all stations is avoided; the spatial constraint value is calculated in each partition so as to quantify the spatial influence of the surrounding stations on the target station, the perception ability of the prediction model on the spatial correlation is enhanced, and the prediction accuracy is improved; and meanwhile, the LSTM neural network model is adopted to fuse the historical PM2.5 concentration and the spatial constraint value for prediction, so that the time dynamic characteristics and the spatial correlation characteristics of the PM2.5 concentration change can be effectively captured, and the space-time modeling capability is higher.

The invention discloses a tree skeleton simplification and completion method based on DBSCAN. The method comprises the following steps: step 1, simplifying a tree skeleton based on a Laplace contraction algorithm; and step 2, complementing the tree skeleton based on a mean value adjacent point interpolation algorithm. The method comprises the following steps: firstly, extracting an initial skeleton with dense point clouds of a tree by using a semantic-based Laplace contraction algorithm, then clustering dense skeleton points in the initial skeleton by using a DBSCAN algorithm, and extracting a skeleton of the dense skeleton points by using a Laplace contraction algorithm, thereby simplifying the dense skeleton points, and improving the clustering precision of the tree. And finally, carrying out interpolation on the missing skeleton points by using a mean value-based adjacent point interpolation algorithm provided by the invention, and carrying out skeleton point smoothing on the interpolated skeleton based on a Gaussian smoothing function. Based on the accurate tree skeleton obtained by the method, help and technical support are provided for forestry resource checking.