42 results about "Vector optimization" patented technology

The invention provides a marketing strategy generation method and device, equipment, a storage medium and a program product, and relates to the technical field of data processing, and the method comprises the steps: carrying out the vector space alignment of a user behavior feature and a knowledge base vector, so as to determine the semantic association degree of the user behavior feature and the knowledge base vector; obtaining entity vectors related to the user behavior characteristics from a knowledge base on the basis of the semantic association degree; predicting the consumption intention of the user based on the spliced data of the user behavior characteristics and the entity vectors; generating a marketing strategy based on the user consumption intention; wherein the user behavior features are obtained by fusing semantic features and structured features in the user multi-source heterogeneous data, and the semantic features are obtained by processing the user multi-source heterogeneous data based on a large model. According to the method, the generalization ability of a large model and the semantic association ability of knowledge base vector optimization are fused, and a real-time dynamic intelligent data mining framework is constructed, so that a high-precision personalized marketing strategy is generated.

The invention provides a cylinder processing cutter path adaptive optimization generation method, which belongs to the technical field of cylinder processing, and adopts fractal dimension analysis to identify surface geometric complexity and determine a minimum repetitive unit, and applies a four-color theorem to carry out non-intersection area division on the surface, so as to improve the processing precision of the cylinder processing cutter path. Multi-axis linkage machining parameters are initialized, global path planning and cutter axis vector optimization and collision detection are performed based on an improved Di jkstra algorithm, path adaptive adjustment is performed according to local curvature change of a minimum fractal matrix, a game model is established, and coevolution of a double-layer game model is realized by adopting a cross optimization rule. And finally, an optimized tool path NC code containing the position coordinates, the attitude angle and the feeding speed is generated, and the technical problem that the machining efficiency and the surface quality are difficult to consider in tool path planning under the complex geometrical characteristics of the cylinder surface is solved.

The invention discloses a three-level ANPC converter model prediction control method and device based on Newton iteration method vector optimization, and belongs to the technical field of converter control, and the method comprises the steps: building a mathematical model of a three-level ANPC converter under a three-phase static coordinate system based on the Kirchhoff theorem; performing forward Euler discretization on the mathematical model of the three-level ANPC converter to obtain a current prediction equation model; establishing a cost function according to the control target; performing online search on the optimal voltage vector according to a cost function; voltage vector synthesis is carried out, voltage vectors selected when the optimal voltage vector falls in 36 different small sectors, a voltage vector action sequence and action time distribution are obtained through calculation, and then the action time of a pair of redundant positive and negative small vectors in the voltage vectors is adjusted to carry out neutral point potential balance control. According to the invention, rapid optimal voltage vector search and three-level ANPC converter output current and neutral point potential control are realized.

The invention provides a vector optimization algorithm translation method and device oriented to an RISC-V vector extension platform, and relates to the technical field of computer software cross-architecture translations. The method comprises the steps that source codes of vector optimization algorithms written by other platforms and translation cues are input into a large language model; outputting a translation code of a vector optimization algorithm oriented to the RISC-V vector extension platform; the translated codes are input into a compiler to be compiled; under the condition that compiling succeeds, a compiling product is linked to the test suite to obtain an executable program, and unit testing is conducted on the executable program on the RISC-V vector extension device; and under the condition that the unit test is passed, determining that the translation code is a first version of correct translation code of a vector optimization algorithm oriented to the RISC-V vector extension platform. According to the method, a vector optimization algorithm written by using a special vector language for other platforms is efficiently translated to the RISC-V vector extension platform.

The invention belongs to the technical field of building single or multi-view-angle to 3D modeling reconstruction, and provides a 3D reconstruction method based on a real estate building cluster point cloud to MESH neural network. Comprising the steps of original point cloud data acquisition, noise removal, grid filtering, normal vector optimization, density uniformity processing, continuous multilayer 3D convolutional neural network processing, color mapping, texture mapping, color consistency optimization and texture distortion optimization. Through uniform sampling, noise removal and grid filtering, the point cloud data volume is reduced, and the quality of the point cloud data is improved; through the continuous multi-layer 3D convolutional neural network, the local geometric pattern of the building is automatically learned from the data and the features of the building are extracted layer by layer, so that the precision of the three-dimensional grid model is improved; through color consistency optimization and texture distortion optimization, the color and texture mapping effect is optimized, and refined reconstruction of the building is realized.

The invention discloses a permanent magnet brushless direct current motor and a parameter optimization method thereof. A rotor adopts a regular hexagon structure, and permanent magnets are alternately embedded into the rotor. Seven parameters such as the maximum thickness of a permanent magnet, a pole-arc coefficient and the width of a stator notch are selected for optimization, air gap flux density, no-load back electromotive force and no-load cogging torque are taken as optimization targets, a response surface model of a multi-objective function is constructed, parameter weights of the multi-objective function are determined in combination with a CRITIC weight analysis method, the motor parameters are optimized by adopting an improved MOEA / D algorithm, and the optimal parameters of the multi-objective function are obtained. And carrying out crossover variation on the population by directional variation guided by electromagnetic characteristics, and designing a weight vector optimization Chebyshev aggregation method to find an optimal iteration population. According to the weight in the Chebyshev aggregation function designed by the invention, the problems that the traditional Chebyshev aggregation method is not smooth enough for the continuous MOP problem and the obtained solution is not uniform are solved, and the distribution uniformity of the Pareto solution set is optimized. The method has obvious advantages in the aspects of convergence speed, solution set distribution and engineering feasibility.

The invention discloses an anti-interference receiving beam optimization method for a communication and perception fusion system. The method comprises the following steps: estimating an arrival angle of a receiving signal of the communication and perception fusion system by using a spatial spectrum estimation algorithm; according to the angle of arrival of the received signal and the mathematical model of the received signal, carrying out inversion to obtain snapshot information of the received signal; according to the snapshot information of the receiving signal, a receiving beam vector optimization problem is constructed, the receiving beam vector optimization problem aims at maximizing the uplink communication channel capacity, interference suppression constraint is introduced at the same time, and an optimal receiving beam vector is designed; and obtaining an optimal receiving beam vector by solving the receiving beam vector optimization problem. According to the invention, external electromagnetic interference signals can be effectively suppressed, the communication quality and sensing performance of an ISAC system are improved, the bit error rate is reduced, and the accuracy of functions of target detection, positioning, tracking and the like is enhanced.

The application relates to a merging vector optimization method in a mobile cell-free scenario, wherein the method comprises the following steps: generating and dividing a data sample set based on a pre-established full connection CF mMIMO network architecture; performing graph construction and initial graph embedding on random channel state information in a training set to obtain an initial graph embedding vector; performing forward propagation calculation on the initial graph embedding vector in combination with a pre-established graph neural network to obtain a current round merging vector; determining an initial graph neural network model when a training bit error rate meets a first bit error rate requirement; testing a test set to obtain a test bit error rate; determining a final graph neural network model when the test bit error rate meets a second bit error rate requirement; and deploying a final full connection CF mMIMO network architecture and performing network scheduling. Thus, the problems of high calculation complexity, low real-time performance and low reliability in the related art are solved, the calculation complexity is reduced, and the robustness, scalability and system fairness are improved.

This invention discloses a GNSS interference source localization method based on ADS-B data. The method involves acquiring ADS-B data, saving data points affected by GNSS interference, and, if no NIC is available, constructing ADS-B data with a NIC. The message data is then sorted by ICAO and timestamp. A diagonal matrix W is calculated based on the ICAO, timestamp, NIC, a table showing the correspondence between NIC and received interference power, and the hyperparameter value τ. A residual vector R is calculated based on the message data and the optimization vector, and the optimized vector is then output. This invention analyzes large-scale aviation ADS-B data, sorts and classifies the original message data to determine whether flights are subject to GNSS interference, and abstracts the interference source localization problem into a least-squares problem. An optimization function is constructed, and an iterative optimization algorithm is used to calculate the possible GNSS interference source locations and transmission power, thus more efficiently locating GNSS interference sources.

The invention discloses a rapid model prediction control method and device for a three-level grid-connected converter, and the method comprises the steps: selecting a vector optimization sector and a corresponding central axis vector according to a system output current prediction model and a single-target value function; judging and selecting a secondary optimization core area through a difference value between the selected central axis vector and a power grid voltage vector, and finally outputting three voltage vectors; in combination with current tracking single-target value function values corresponding to the three selected output voltage vectors, the action time of the three output vectors in the next control period is calculated, and the action time of a small vector in the three output vectors is further adjusted according to the difference value of the upper voltage and the lower voltage; therefore, current tracking and neutral-point potential balance control of the converter system are realized. According to the method, the calculation in the algorithm execution process is greatly reduced, the system response speed is increased, and the fixed switching frequency is obtained at the same time through the prediction error simplification rolling optimization positioning output vector method.

The invention discloses a representation vector optimization method, system and device for sequence data prediction and a medium, and particularly relates to the technical field of representation vector optimization. According to the method, a hidden variable is introduced into each variable, meanwhile, hidden state coding is carried out on original observation variables full of uncertainty, probability distribution obeyed by the hidden variables is estimated by using a neural network, and parameters of the distribution serve as more robust representation vectors of the variables; the method comprises the following steps: adaptively screening out an invariant feature code which is most stable in association with a predicted target from a variable sequence by using a gating mechanism so as to reduce the uncertainty of the invariant feature code; meanwhile, through an asynchronous loop optimization strategy, information gains provided by the node variables and the information variables in the information spreading process are continuously obtained, and representation information of the node variables and the information variables under information coupling is enhanced.

The invention relates to a dual three-phase permanent magnet motor fault-tolerant prediction control method based on virtual vectors and common-mode rejection, and belongs to the technical field of permanent magnet synchronous motor control. In order to solve the problems of fundamental frequency offset and high harmonic content of common-mode voltage after a T-type three-level dual three-phase electric drive system has a single-phase open-circuit fault, a symmetric prediction current model is established by constructing a reduced-order decoupling matrix, a zero-harmonic voltage virtual vector set is generated, and the vector set is alternately input in combination with real-time common-mode voltage offset detection, so that the zero-harmonic voltage is obtained. And a weight-free series-parallel multi-objective optimization cost function is adopted to carry out vector preferential optimization. According to the method, the abc phase common-mode voltage is reduced to Udc / 6 in a high-speed interval, the full-speed domain of the def phase is lower than Udc / 8, the current distortion rate is smaller than 10%, meanwhile, neutral-point potential balance and low switching frequency are considered, and the fault-tolerant capability and the operation efficiency of the system are remarkably improved.

The present application relates to the technical field of forklift motor, and particularly relates to a forklift motor energy-saving operation method and system based on reluctance torque vector optimization, which collects motor operation parameters through a sensor, fuses multi-source dynamic characteristics by using a parallel time sequence feature extraction network, extracts working condition characteristics in combination with bidirectional domain adaptive transfer learning, determines the number of key factors based on a dynamic working condition correlation matrix, generates a motor control characteristic vector through convex optimization and probability density estimation, generates a control scene in combination with a reluctance torque model and a phase current coupling relationship, calculates an energy efficiency index based on torque response and energy consumption distribution matching degree, and solves optimal control parameters by using a multi-objective optimization algorithm. The parameters are input into a grouping collaborative control unit, control groups are divided according to phase space distribution, optimal reluctance torque vectors are calculated by using an energy consumption dynamic prediction model through an intra-group competition and inter-group cooperation mechanism, and high-efficiency energy-saving operation under low-speed large-torque and frequent start-stop working conditions is realized.

The invention discloses an intelligent legal file retrieval method and system based on a multi-modal heterogeneous large model, and the method comprises the steps: obtaining original legal file data containing a text, an image and a table, and carrying out the multi-modal analysis and structured preprocessing; extracting depth features of each modal through a multi-modal heterogeneous large model, and fusing the depth features by adopting a multi-modal cross attention mechanism; constructing a case element sparse graph and optimizing the case element sparse graph through an optimization algorithm; carrying out weight adjustment and vector optimization by using a robust learning real-value multi-index model algorithm; candidate case subsets are screened through a random recovery sampling technology, and multi-dimensional matching calculation is carried out; and carrying out visual presentation and interactive analysis processing on the similar historical case list. Through the multi-modal fusion understanding, sparse graph optimization, robust learning and other technologies, the legal file intelligent retrieval accuracy and efficiency are effectively improved, and accurate case reference and decision support are provided for legal workers.

The application discloses a linear trajectory planning method for a space manipulator load operation task and relates to the technical field of space manipulator trajectory planning. A multi-objective optimization model is designed, and optimization targets include straightness vector optimization, obstacle avoidance optimization and jerk optimization; a trajectory planner is designed, an individual searched is a middle node configuration of a quintic polynomial trajectory, a quantum state coding mode is adopted to code the configuration of the manipulator, a QBSO algorithm is provided, a complete trajectory of the manipulator is sequentially connected by a plurality of middle node configurations, the next node configuration is searched according to a current configuration position and a target trajectory each time, and finally, a joint trajectory connected according to a quintic polynomial equation by n middle node configurations is formed. By designing a mathematical coding mode of the quintic polynomial trajectory, designing a linear trajectory planner framework based on a sampling algorithm and designing a straightness vector constraint model, an optimal trajectory meeting constraint requirements can be generated.

The application discloses a kind of flat bottom end face milling cutter tool shaft vector optimization method, tool shaft vector optimization is considered from processing interference, cutting force, rotation axis angular displacement and angular velocity etc. Multiple factors comprehensively. First, the improved space partition tool shaft vector global interference detection method based on surface principle is proposed, whether the interference detection algorithm is used to judge tool to occur processing interference and carry out interference-free tool shaft vector space calculation;Second, the relationship model of end milling cutter tool shaft vector and cutting force change is established, and the tool shaft vector optimization is carried out with the tool cutting force optimization as the target;Third, using Dijkstra algorithm, an optimal path algorithm-based machine tool rotation axis angular displacement and angular velocity optimization model is proposed;Using the optimization algorithm for complex surface machining, the machining quality of the surface workpiece after tool shaft vector optimization is obviously improved compared with that before tool shaft vector optimization.

The invention relates to an MISO system secrecy rate maximization method and system based on active RIS and artificial noise assistance, and the method comprises the following steps: S1, calculating the secrecy rate of a system according to the information rates of a legal user and an eavesdropper; s2, judging whether the # imgabs0 # meets the # imgabs1 # or not, and if the # imgabs0 # does not meet the # imgabs1 #, executing S3; if yes, executing S7; s3, executing a beam forming vector and artificial noise vector optimization method by taking the # imgabs2 # obtained in the S2 as an input parameter, and outputting # imgabs3 #; S4, executing an active RIS reflection coefficient matrix optimization method by taking the # imgabs4 # obtained in the S3 as an input parameter, and outputting an active RIS reflection coefficient matrix theta; s5, taking the theta obtained in S4 as an input parameter, executing an auxiliary variable optimization updating method, and updating # imgabs5 # and # imgabs6 #; S6, taking # imgabs7 # imgabs8 # theta, # imgabs9 # and # imgabs10 # as input parameters, calculating the secrecy rate # imgabs11 # of the current system, setting n = n + 1, and skipping to S2; and S7, the optimal value of the # imgabs12 # is obtained. According to the method, the influence of the double-fading effect of the RIS-assisted MISO system is overcome, and the secrecy rate of the system is improved.

The application discloses a kind of legal dossier intelligent retrieval method and system based on multimodal heterogeneous big model, comprising: obtaining original legal dossier data containing text, image and table, carrying out multimodal analysis and structured preprocessing;Through multimodal heterogeneous big model, each modal depth feature is extracted and is fused using multimodal cross attention mechanism;Case element sparse graph is constructed and is optimized by optimization algorithm;Robust learning real value multi-index model algorithm is applied to weight adjustment and vector optimization;Through random recovery sampling technology, candidate case subset is screened and multidimensional matching calculation is carried out;Similar historical case list is visualized and interactive analysis and processing are carried out.The application improves the accuracy and efficiency of legal dossier intelligent retrieval by multimodal fusion understanding, sparse graph optimization and robust learning technology, provides accurate case reference and decision support for legal workers.

The invention discloses a virtual resource dynamic scheduling method and system in a hybrid cloud environment, relates to the technical field of resource scheduling, and translates an abstract service demand protocol into a scheduling strategy object which can be understood and activated by a machine by capturing and analyzing the abstract service demand protocol. Meanwhile, multi-dimensional fusion is carried out on the resource state, the multi-cloud price model and the real-time service KPI flow in the underlying mixed cloud environment to form a globally unified system state view. On this basis, an adaptive mechanism based on a control loop is introduced, and continuous weight vector optimization and generation are carried out by taking an activation strategy analyzed by an intention as a guide and taking a multivariate optimization index as a decision factor capable of being dynamically deployed. And finally, the scheduler performs intelligent weighted scoring on the new task request by using the group of dynamic weights, and makes an optimal placement decision. Therefore, an intelligent decision closed loop from business intention definition to resource scheduling execution can be realized, so that the resource allocation efficiency and business agility of the hybrid cloud environment are improved.

The present application relates to a kind of demand side response potential evaluation and benefit optimization system based on load feature clustering, belong to electric power system and its automation technical field.Therein, the method includes: obtaining industrial park heterogeneous load data and preprocessing, extract load time dimension feature by wavelet transform and fuse static attribute, generate standardization load representation vector after feature screening;Based on vector optimization clustering core point, build three-dimensional clustering evaluation system, construct user portrait in combination with production constraint;Quantify production rigidity and flexible constraint, construct response willingness degree model and dynamically correct;Build multi-time scale optimization framework, scale to interface grid data, output optimal response strategy.Load feature accurate representation and response potential scientific evaluation are realized, the pertinence and feasibility of demand side response strategy are improved, grid dispatching adaptability and user benefit maximization are considered, and technical support is provided for industrial park demand side response large-scale promotion.

The invention discloses a concealed wireless communication beam forming optimization method under limited character input, and belongs to the technical field of wireless communication. According to the method, a transmitting power and beam forming vector optimization strategy for jointly optimizing a transmitting end under different modulation modes is constructed, and a positive semidefinite relaxation algorithm is designed on the basis of utilizing lower bound approximation of an objective function: an initial non-convex objective function and a hidden constraint are equivalently converted into a convex form which is easier to process; and solving by using a CVX toolkit in MATLAB to obtain a design which enables the problem to be optimal in beam forming vector. Furthermore, in order to reduce the calculation complexity, the invention provides a low-complexity optimization strategy, and a simulation result shows that the method can remarkably reduce the calculation complexity while meeting the covert communication performance.

The invention discloses a multi-modal abstract generation method based on dynamic retrieval and self-adaptive query, and relates to a multi-modal data processing technology, and the method comprises the following specific steps: preprocessing a text sequence and a video sequence, extracting features, and projecting the features to the same dimension; constructing a knowledge base, and retrieving and enhancing to obtain enhanced features; fusing the text and video features into an initial mixed query vector and optimizing the initial mixed query vector; then constructing a joint feature sequence, inputting the joint feature sequence into a multi-layer multi-mode Transform, and outputting fusion features; and finally, calculating a correlation score, and generating a multi-modal abstract according to a Top-K strategy. Through multi-modal feature unification and knowledge base enhancement, a modal barrier is broken, and feature pertinence is improved; in combination with query vector optimization and multi-layer cross-modal fusion, interaction between modals is enhanced, high correlation of abstract elements is ensured, the problems of information alignment deviation and content one-sided are effectively solved through the design, the abstract quality and practicability are improved, and the method has wide popularization value.

The invention discloses a workpiece three-dimensional model construction and processing method based on binocular vision detection data, and the method comprises the steps: firstly carrying out the multi-angle synchronous scanning of a workpiece through employing a combination of a binocular camera and a line laser, and obtaining the multi-view three-dimensional point cloud data of the surface of the workpiece; and constructing a density sensing type dynamic registration framework, and performing distance weighted fusion and outlier elimination on the registered multi-view point clouds in an overlapping region to generate a complete workpiece point cloud. A feature retention type point cloud direct modeling strategy is adopted, normal vector optimization is performed on a down-sampling point cloud, key feature marks are manufactured, a lightweight three-dimensional model is generated through color coding rendering, and a large amount of calculation of traditional surface reconstruction is avoided. And finally, based on model coordinate system calibration and point cloud fitting optimization, extracting functional characteristic parameters of edge straightness, datum plane flatness and hole system position degree, and completing workpiece quality evaluation. According to the method, the point cloud processing speed and the key feature analysis precision are improved, and the real-time requirement of industrial online detection is met.

The application belongs to the field of tool axis vector trajectory planning, and particularly discloses a tool axis vector optimization domain planning method and system considering workpiece machining flexibility, which comprises the following steps: establishing a workpiece finite element proxy model, loading cutting force at each tool point to obtain equivalent flexibility of each tool point, and determining a maximum flexibility point; obtaining a maximum cutting force according to a maximum deformation allowed by the workpiece and the equivalent flexibility of the maximum flexibility point, and combining a pre-established cutting force prediction model to obtain a constraint condition; based on the constraint condition and a preset range of machining parameters, performing multi-objective optimization of the maximum flexibility point in coupling with machining parameters to obtain a critical tool axis vector of the maximum flexibility point under critical machining parameters; determining critical tool axis vectors of other tool points under the critical machining parameters to obtain a tool axis vector optimization domain. The application can calculate the tool axis vector optimization domain division within the machining parameter limited range under the influence of force-induced errors and machining efficiency, and provide guidance for tool axis vector trajectory planning.

The application provides an MCR-WPT system plane coil structure parameter optimization method based on an improved four-vector optimization algorithm; an S-S topology compensation structure network is analyzed, and the energy transmission efficiency of the MCR-WPT system is derived; the influence of the coil physical structure on the coil characteristic parameters is analyzed, and the self-inductance value of the coil is calculated; the mutual inductance coefficient value between the coupled coils is calculated; the alternating current resistance is calculated; the coupling coefficient between the coupled coils is calculated, the frequency splitting is analyzed to establish a constraint condition; the optimization variables are determined, and the energy transmission efficiency is taken as the objective function value and the optimization target; the constraint condition is set according to the critical coupling coefficient, and the parameter range is adjusted according to the constraint condition; and the improved four-vector optimization algorithm is used to optimize the energy transmission efficiency, that is, the objective function, according to the coil setting parameters.

The present application relates to the technical field of space modeling and three-dimensional point cloud data processing, and particularly relates to a point cloud model optimization method and a base station point selection method based on point cloud model feature extraction. The present application first performs down-sampling on a point cloud model, then for each point, obtains a unit normal vector of a fitting plane of the point and its neighborhood points as the unit normal vector of the point; performs normal vector optimization on each point in the simplified point cloud; calculates the feature measure of each point in combination with the optimized unit normal vector of each point in the simplified point cloud, and selects points with feature measures located in a set feature interval to form a feature point cloud. The present application realizes data simplification of the point cloud model while preserving the original features of the point cloud model.

Techniques for using AI to reduce a tensor network are disclosed. A service receives an ANSATZ model that is structured as a DAG. This input DAG includes nodes and edges. The service receives a vector reflective of an optimization problem. The optimization problem identifies parameters related to the ANSATZ model. The service feeds the input DAG and the vector as input to the ML algorithm. The ML algorithm attempts to optimize the parameters by assigning probabilities to the nodes and edges. The probabilities reflect whether corresponding tensors will be included in an output ANSATZ DAG. The service receives an output ANSATZ DAG from the ML algorithm. The service then applies a probability threshold to the output ANSATZ DAG, resulting in removal of nodes and edges from the output ANSATZ DAG.

The application provides a vector optimization algorithm translation method and device for a RISC-V vector expansion platform, and relates to the technical field of computer software cross-architecture translation, and the method comprises the following steps: inputting source code of a vector optimization algorithm written by other platforms and a translation prompt into a large language model, and outputting translation code of the vector optimization algorithm for the RISC-V vector expansion platform; inputting the translation code into a compiler for compilation; in the case of successful compilation, linking the compilation product to a test suite to obtain an executable program, performing unit testing on the executable program on a RISC-V vector expansion device; and in the case of passing the unit testing, determining that the translation code is the first version of correct translation code of the vector optimization algorithm for the RISC-V vector expansion platform. The application realizes efficient translation of a vector optimization algorithm written by using a special vector language of other platforms to the RISC-V vector expansion platform.

The application discloses a reconfigurable intelligent surface assisted movable antenna direction modulation method and system, specifically, an optimization problem model is established on the basis of an active intelligent reflecting surface assisted large-scale multiple-input multiple-output system, confidential information transmitted by a base station equipped with a movable antenna reaches a legal user and an eavesdropper through two main paths; under power constraints, maximum reflection coefficient constraints and movable antenna position constraints, the maximum safe rate of the legal user and the eavesdropper is designed; in the transmission end weight vector optimization stage, the movable antenna position constraints are converted into l2 norm by using a compression sensing method, the transmission rate that can be realized by the eavesdropper is fixed, and a singular value decomposition is used to obtain the transmission end weight vector; in the phase shift matrix optimization stage, the transmission rate that can be realized by the legal user is fixed, and a singular value decomposition is used to obtain the intelligent reflecting surface phase shift matrix. The application greatly improves the achievable safe rate.

The invention provides an InSAR deformation anomaly extraction method and system based on multi-modal fusion and vector optimization, and relates to the technical field of artificial intelligence and geological disasters, and the method comprises the steps: obtaining regional multi-source remote sensing data, calculating earth surface deformation phase data, topographic relief, gradient, vegetation index, water body index and soil index, and carrying out the calculation of the earth surface deformation phase data, the topographic relief, the gradient, the vegetation index, the water body index and the soil index; re-sampling to obtain multi-source remote sensing full-waveband data; performing manual interpretation on the InSAR surface deformation abnormal range of the region based on expert knowledge to obtain vector data and binary mask data of the InSAR surface deformation abnormal range; constructing a double-branch multi-modal feature fusion network, and extracting the surface deformation anomaly information of the InSAR based on the binary mask data and the multi-source remote sensing full-waveband data; and constructing a dynamic vector optimization network, performing boundary optimization on the InSAR earth surface deformation abnormal information based on the multi-source remote sensing full-waveband data, and dynamically converting the information into a vector result. According to the invention, seamless connection from the remote sensing image to the GIS vector is realized.