14 results about "Pattern search" patented technology

The embodiment of the invention provides an adversarial verification evaluation method, system and device for a fault mode coverage capability boundary of a fault detection and diagnosis model. The adversarial verification evaluation method for the fault mode coverage capability boundary of the fault detection and diagnosis model comprises the steps that an obtained fault mode search domain is divided into at least two sub-domains, and the fault mode search domain is generated by a parameter combination used for describing a target equipment fault mode; determining a sample extraction number corresponding to each sub-domain in the current round, and extracting from the initial sample set according to the sample extraction number to obtain a corresponding target verification sample; processing the target verification sample through a fault detection model to generate a corresponding fault detection result; and based on the fault detection result and a preset evaluation index, verifying and evaluating the fault mode coverage capability of the fault detection model, and generating a corresponding verification and evaluation result.

The application discloses a short-wave infrared high-NA achromatic superlens design method based on Latin square partition and hybrid optimization, and belongs to the technical field of micro-nano optics. Firstly, an electromagnetic response database of a silicon dioxide substrate-silicon nanometer cylindrical unit is constructed, covering a 2.1-2.3-micron wavelength band; then equal-area ring-belt-fan partition is adopted in combination with Latin square wavelength distribution, so that multiple wavelengths are uniformly distributed in the ring-belt and sector directions; a high-NA evaluation function containing a radial quadratic position weighting factor is constructed, so as to strengthen the optimization weight of the edge phase of the lens; finally, a genetic algorithm-pattern search three-stage hybrid collaborative optimization is executed, and the optimal nanometer structure size distribution is obtained through global search, local convergence and pixel-level discrete fine adjustment. The application can significantly suppress wavelength crosstalk and chromatic aberration, realizes broadband achromatic focusing under the condition of a numerical aperture NA=0.915, and the focal length variation coefficient is only 3.45%, the focused light spot is close to the diffraction limit, and the application is suitable for short-wave infrared high-NA imaging applications such as gas detection and environmental monitoring.

The application relates to the technical field of environmental monitoring and discloses a big data-based atmospheric pollutant rapid tracing system, which comprises a diffusion simulation module, a hybrid optimization module, a dynamic algorithm selection module, a multi-source analysis module and a verification and output module. The dynamic algorithm selection module is configured to build a dynamic matching rule of meteorological data features and a tracing algorithm, to adaptively switch an anti-interference algorithm model according to different meteorological conditions, to eliminate meteorological refraction deviation in pollutant concentration field simulation in real time, to reduce the intensity inversion error caused by optical interference in the traditional method, and to guarantee the physical authenticity of source intensity estimation. The hybrid optimization module is configured to use an axial detection and mode moving strategy to monitor the search path state of a genetic algorithm in real time, to automatically identify the parameter space deviation in the mathematical level, and to trigger a mode search correction mechanism and a genetic algorithm restart operation, so that the geometric accuracy and spatial reliability of the tracing result are improved.

The invention belongs to the technical field of swarm intelligence optimization algorithms, and particularly relates to a multi-strategy fusion improved star-graffiti optimization method, which comprises the following specific steps of: generating an initial population by adopting a strategy of combining Logistic chaotic mapping and mirror reflection learning, and enhancing the uniformity and quality of population distribution; in the iteration process, a nonlinear decline function and a periodic fluctuation function are used for adjusting key control parameters respectively, and dynamic balance of global exploration and local development is achieved; introducing a mode search operator based on a dynamic step size in the middle stage of iteration, and carrying out refined mining on an optimal solution region through direction search and disturbance optimization; high-quality gene transfer is accelerated through an information sharing mechanism based on a cross strategy and greedy selection. Through multi-strategy fusion improvement, the risk that the algorithm falls into local optimum is reduced, and the problem that the global search capability of an original algorithm is limited is solved.

A storage device enables a host to search for data stored across channels on a memory device without reading the user data stored on the channels. The storage device includes a memory device including user channels to store user data and a hash channel to store hash data associated with the user data. The hash data is used to identify the user data. A controller on the storage device generates the hash data including hint data during storage of user data on the user channels. The controller receives a hint from the host, generates a request and submits the request to a pattern search engine. The controller sends a hint response including flash translation indices to the host, based on a response from the pattern search engine. The host may read the flash translation indices to request data stored at a specific location in the user channels.

The present invention provides a ray tracing-based method for indoor multi-base station location optimization in a millimeter wave frequency band, comprising: constructing an indoor millimeter wave network model; determining an indoor millimeter wave network model optimization constraint condition; constructing a cost function of multi-base station location deployment on the basis of the constraint condition; determining the initial location of each base station in the indoor millimeter wave network model; and by taking the initial location of each base station as a starting point, determining an optimal location of each base station by means of axial search together with pattern search. According to the present invention, a good balance can be achieved between the accuracy of optimization results and the complexity of optimization algorithms for the multi-base station deployment optimization problem, so that the optimized base station locations can provide high-quality signal coverage for an indoor millimeter wave network.

Embodiments are generally directed to compression for compression for sparse data structures utilizing mode search approximation. An embodiment of an apparatus includes one or more processors including a graphics processor to process data; and a memory for storage of data, including compressed data. The one or more processors are to provide for compression of a data structure, including identification of a mode in the data structure, the data structure including a plurality of values and the mode being a most repeated value in a data structure, wherein identification of the mode includes application of a mode approximation operation, and encoding of an output vector to include the identified mode, a significance map to indicate locations at which the mode is present in the data structure, and remaining uncompressed data from the data structure.

The invention is suitable for the technical field of graph data, and relates to a rule mining method and device for a graph neural network, equipment and a storage medium. Wherein negative nodes in the graph data are determined, and the negative nodes are nodes for making negative prediction by the graph neural network; for each negative node, extracting a negative key sub-graph which enables the graph neural network to make negative prediction from the graph data; identifying a common graph mode between the negative key sub-graphs, and based on the common graph mode, searching a reference graph mode with rule measurement reaching a measurement threshold; and obtaining a mode attribute predicate set corresponding to the reference graph mode, and generating a negative prediction rule corresponding to the graph neural network according to the mode attribute predicate set. According to the invention, the negative prediction rule after the negative prediction of the graph neural network can be mined.

A high-speed target detection method based on a spaceborne measurement radar with binary phase-coded signals comprises the following steps: 1. Dividing the target detection process into stages; 2. Dividing the range segments according to radar detection range requirements to obtain multiple range modes; 3. Classifying the speed according to radar velocity measurement requirements; 4. Determining the priority of range mode selection during full-range mode search and the priority of speed level selection during full-speed level search; 5. Developing search strategies and search-to-tracking strategies for each range mode; 6. Developing speed selection and update strategies when calculating the phase increment of the target echo using Doppler compensation during the target search stage; 7. Developing speed selection and update strategies when calculating the phase increment of the target echo using Doppler compensation during the target tracking stage; 8. Continuously tracking and measuring the target to obtain target information. This invention effectively solves the problem of high-speed target detection, is not limited by radar system, and has a wide range of applications.

The invention relates to an intelligent real estate management system development method based on a large model. The method comprises the steps of performing directional calling on a database based on a unique identifier of system login of an operator to obtain a real estate management system operation table; extracting total user information of the operator by using the login identifier primary key information; obtaining a preference feature set of an operator and a sorting result of functional requirements, and converting text information into a high-dimensional semantic vector through vectorization processing; obtaining a matching degree score and carrying out standardized quantitative assignment; utilizing front-end dynamic rendering to generate an AI function list of an operator; carrying out association mapping on the search service vocabularies, and configuring semantic tags and search words at the same time; performing mode search on the search word through a real estate large model to obtain real estate description data; performing double-region function-description linkage presentation by using a front-end and back-end data synchronization interface; the development of the intelligent real estate management system based on the large model is realized.

The invention discloses a multi-objective optimization method and system based on a full-load passenger car framework in the technical field of passenger cars. The optimization method comprises the following steps: S1, defining geometric parameters of a passenger car framework and taking the geometric parameters as optimization design variables, and establishing a passenger car framework finite element simulation model; s2, determining typical working conditions and constraint conditions of the passenger car, and calculating performance indexes under each working condition through simulation; s3, sensitivity analysis is conducted on the design parameters, and the design parameters are screened and determined; s4, processing the design parameters to generate a design parameter data set, and calculating corresponding performance indexes through the simulation model; s5, constraining the design parameter data set and the performance indexes, and training an AI model; and S6, searching the optimal design parameters meeting the performance requirements by using a pattern search solver. According to the optimization method, through the synergistic effect of the finite element simulation model and the AI model, collaborative optimization of multiple working conditions and multiple targets is achieved, lightweight design and performance balance of the passenger car body structure are achieved, and the design period is shortened.

This invention discloses a novel swarm intelligence optimization algorithm integrating multiple strategies. The method steps are as follows: S1, initialize the Tibetan fox population and dynamic territory; S2, adaptively adjust the territory division using nonlinear dynamic parameters combined with population distribution; S3, integrate PSO, DE, and gradient search strategies to dynamically adjust the selection probability to achieve precise local development; S4, design an adaptive migration mechanism; S5, maintain population diversity through subpopulation co-evolution and periodic information exchange; S6, add a globally optimal guided pattern search to improve convergence accuracy; S7, complete population update using elite retention and dynamic elimination strategies. Compared with existing technologies, this invention's core parameters are dynamically adjusted nonlinearly, balancing exploration and development. Experiments verify that it has better accuracy and robustness in multi-function optimization, and its performance in practical applications is significantly better than PSO and GA. It can be widely used in engineering optimization, machine learning hyperparameter tuning, and other fields.

The application discloses a full-cross visual angle text matching method, which comprises the following steps: firstly, five semantic division modes, i.e., character mode, accurate mode, full mode, search engine mode and Paddle mode, are used to extract three text granularity information, i.e., characters, words and associated phrases, in the text; the division result of each division mode is a representation of the text; the five representations of the text are initially coded; and the context semantic information in the coding is preliminarily extracted by using a bidirectional gate recurrent unit. Then, the initial coding vectors under the five representations are reconstructed into a high-dimensional coding matrix, and the feature information in the high-dimensional matrix is deeply mined by using a convolutional neural network, so that the multi-representation semantic of the text is effectively captured, and the information interaction between the representations is improved. Finally, the multi-representation convolution matrices of two texts are cross-cosine matched by using a full-cross visual angle matching mode, the matching strength of the multi-representation information is strengthened, and the accuracy of the text matching task is improved as a whole.