13 results about "Location model" patented technology

The invention discloses a stereoscopic warehouse visualization method and system based on a POI technology, and relates to the field of warehouse management. Constructing an intelligent mapping engine, dividing and managing a physical shelf structure of a real stereoscopic warehouse according to a preset hierarchical system, establishing a deterministic batch mapping relationship between goods allocation codes and three-dimensional space coordinates, and generating goods allocation three-dimensional space coordinates; a dynamic hierarchical rendering mode is established, and intelligent switching among different view hierarchies and targeted rendering of the corresponding hierarchies are realized through view hierarchy identification and data matching mechanisms; constructing a state sensing type material mode, and dynamically updating the display characteristics of the cargo virtual model according to the preset attribute parameters of the cargo; a space positioning detection technology is adopted, and when a user triggers a goods allocation model interaction operation, coordinate adaptation of a three-dimensional space and a data display interface is automatically completed. According to the method, automatic batch mapping is realized, the labor investment is reduced, and the rendering performance is improved by using a dynamic hierarchical rendering system.

The invention discloses a damped part positioning model construction method and device, electronic equipment, a storage medium and a program product. The method comprises the following steps: carrying out a damp experiment on a plurality of transformer oil paper insulating sleeve samples to obtain multi-source original data of the transformer oil paper insulating sleeve samples, and determining a multi-dimensional feature fusion vector corresponding to the multi-source original data; wherein the multi-source original data comprises working condition data, actual measurement data and simulation data; disassembling the transformer oil paper insulating sleeve sample after the wetting experiment, and determining the actual water content of a preset part of a core of the transformer oil paper insulating sleeve sample; wherein the preset part comprises a main screen, an end screen and a flange; and constructing a sample data set according to the multi-dimensional feature fusion vector and the actual water content, and training a machine learning algorithm through the sample data set to obtain a target damped part positioning model. According to the technical scheme provided by the embodiment of the invention, the positioning accuracy of the damped part can be remarkably improved.

A robust pre-disaster reinforcement method for distribution networks considering the temporal nature of disasters and the fault tolerance of fault location includes: establishing a time-varying vulnerability model for distribution network components and generating an uncertain disaster scenario set containing fault states of lines and feeder terminal units across multiple time periods based on the Monte Carlo method; constructing a distribution network segment location model and quantifying the location error under each disaster scenario by comparing the difference between the actual and expected states of nodes; establishing a robust pre-disaster reinforcement model with the goal of minimizing the overall system cost; constructing constraints for the robust pre-disaster reinforcement model and iteratively solving the robust pre-disaster reinforcement model using a column and constraint generation algorithm to output the optimal pre-disaster reinforcement scheme. This method, by meticulously characterizing the time-varying evolution of disasters, incorporates the fault tolerance performance of fault segment location into the pre-disaster reinforcement decision-making framework, seeking the optimal resource allocation scheme under the worst-case disaster scenario, thereby achieving a balance between economy and robustness and improving the resilience of the distribution network to extreme disasters.

This invention discloses a method for selecting and planning electric vehicle charging facilities based on multi-source data. First, a user optimal choice model is constructed. Based on this model, a set of optimal user choices is obtained, which is then used as initial parameters to solve the electric vehicle charging facility location model. This set includes a set of optimal search results and a set of driver optimal addresses. Next, an electric vehicle charging facility location model is constructed. Based on the set of optimal search results and the driver optimal address set, the model is solved, leading to a charging facility planning scheme. This scheme includes adjustments to the number of charging piles in current charging facilities and future charging facility location planning. This invention can provide a scientific number of charging piles for candidate sites without existing charging facilities and solve the problem of resource waste caused by excess charging piles for candidate power stations with existing charging facilities.

This invention discloses a method for optimizing warehouse locations for multiple fabric categories based on sales volume and seasonality. The method includes collecting historical sales data for the smallest inventory unit, calculating a comprehensive score for each smallest inventory unit by combining sales volume, seasonality characteristics, and activity adjustment factors; constructing a three-dimensional warehouse location model, and using a three-dimensional spiral algorithm to calculate the access cost from each warehouse location to the shipping outlet; constructing a correlation graph based on the frequency of occurrence of every two smallest inventory units, using the smallest inventory units as nodes in the correlation graph, and grouping the smallest inventory units based on the distance of each edge; using the access cost of each smallest inventory unit and the access cost between every two smallest inventory units in the same group as the objective function, and using the comprehensive score and the correlation graph as reference coefficients to minimize the objective function; this invention reduces the outbound travel distance for the same order, thereby achieving rapid outbound processing.

The present application relates to the technical field of data processing and spatial operations optimization, and particularly relates to a pension facility spatial layout optimization method based on multi-model and reachability analysis, which obtains the spatial distribution data of the elderly population by constructing a hexagonal grid, introduces a Gaussian distance decay function to improve the two-step moving search method, combines the multi-mode travel time cost and facility service capacity to measure the spatial reachability, identifies the supply-demand mismatch area based on the standard score model, constructs a hierarchical optimization strategy by fusing the P-barycenter location model and the reachability variance minimization model for the area, obtains the location scheme through the location optimization stage respectively, and obtains the configuration scheme by using the quadratic programming algorithm in the service capacity configuration stage, and the present application realizes the organic unification of the pension resource configuration efficiency and spatial fairness through the technical means of multi-model fusion and hierarchical progressive optimization.

Systems and techniques are described herein for determining bounding box coordinates. For example, a computing device can process an image to generate a first plurality of tokens associated with the image. The computing device can process, using a first machine learning model, the first plurality of tokens and a class token associated with a class of an object to generate a probability distribution associated with coordinates of a bounding box within the image. The computing device can determine, based on the probability distribution, target coordinates to position the bounding box within the image.

Systems and techniques are described herein for determining bounding box coordinates. For example, a computing device can process an image to generate a first plurality of tokens associated with the image. The computing device can process, using a first machine learning model, the first plurality of tokens and a class token associated with a class of an object to generate a probability distribution associated with coordinates of a bounding box within the image. The computing device can determine, based on the probability distribution, target coordinates to position the bounding box within the image.

The invention discloses a population flow coupled urban vertical take-off and landing field network site selection method, and belongs to the field of urban air traffic infrastructure planning. Multi-source geographic space big data are fused, and a vertical take-off and landing field candidate point set conforming to urban built-up area environmental constraints is constructed; constructing a population flow coupled low-altitude aircraft vertical take-off and landing field network location model CPF-VNLM, taking the candidate point set as a starting and destination vertical take-off and landing field of the model, and determining an objective function, a multi-constraint condition and a binary decision variable which take maximum air-ground combined travel cost saving as a core; designing a model reduction pruning strategy; and solving the model by combining an integer programming solver to obtain an optimal layout scheme of the vertical take-off and landing field network. According to the invention, through coupling the population space-time distribution data and the urban physical space constraint, the vertical take-off and landing field network layout is optimized, and the operation efficiency and the service quality of the urban air traffic system are significantly improved.

This invention relates to the field of flood disaster technology, specifically providing a deep learning-optimized method for selecting flood monitoring equipment locations. The method includes: S1: defining a coverage area calculation method including terrain correction, establishing a monitoring equipment location model, and determining the objective function and constraints of the model; S2: obtaining a training dataset for the deep learning network based on the monitoring equipment location model; S3: constructing a deep learning network and training a crossover point prediction model using the training dataset; S4: embedding the trained crossover point prediction model into a genetic algorithm framework; S5: fine-tuning the parameters of the optimized genetic algorithm for different monitoring areas, terrain features, and other scenarios, outputting an optimal location scheme for monitoring equipment adapted to specific scenarios. This method can be stably applied to monitoring network planning tasks of different scales and terrains, ranging from simple campuses to complex watersheds.

The invention relates to a Bayesian optimization-based deep random configuration network damage location model optimization construction method. The method comprises the following steps of: obtaining time domain signal data sets corresponding to different damage positions; calculating a permutation entropy value and a Higuchi fractal dimension value of the signal, and screening out the signal with the change rate exceeding the reference to extract a damage feature set; a deep random configuration network DSCN containing a single hidden layer and a node is constructed as an initial model, hidden layer parameters are initialized from uniform distribution and screened by a target function, and an output weight is calculated through a least square method; constructing a Gaussian process model by adopting a mixed kernel function, and performing Bayesian optimization on an initial model hyper-parameter through a time weighted expectation improved acquisition function; and utilizing the optimized model to predict a to-be-detected sample and outputting a damage position coordinate, so as to solve the problem of insufficient positioning precision caused by inaccurate signal feature extraction, insufficient attention of the model to key features and low hyper-parameter optimization efficiency during composite material damage positioning in the prior art.

The embodiment of the invention relates to the technical field of model making, and discloses a model making and developing method and device based on an unreal engine, a medium and a product. The method comprises the following steps: acquiring GIS location data, wherein the GIS location data comprises location edge position coordinates and GIS data for depth rendering; using a vector cross multiplication algorithm to extract drawable point locations in the GIS location data and compiling the drawable point locations into groups; making a location model at a corresponding position in the scene through the data set capable of drawing the point location, and dynamically controlling the material of the location model; and dynamically carrying out secondary processing on the location model according to the positions of different data sequences by using a user-defined rendering depth function in an unreal engine. The method can be at least used for solving the problem of low virtual modeling efficiency of large-scale GIS zone bit data.

This invention discloses a method and system for evaluating the installation location of electricity meter collectors, relating to the field of power Internet of Things (IoT) data acquisition and management technology. The method includes collecting multi-source environmental data of the installation area, constructing a candidate installation location model to screen installation conditions; for each candidate installation location, using multi-factor inference rules to output a candidate location suitability score; and based on the candidate location suitability score, adjusting rules and indicator weights according to scenario weights to output a comprehensive evaluation score for the candidate location. This method, by collecting multi-source environmental data of the installation area, constructing a candidate installation location model, and dynamically optimizing it using multi-factor inference rules and scenario weights, scientifically evaluates the comprehensive suitability of candidate locations. This effectively improves the scientific rigor and precision of installation location planning, optimizes communication performance, reduces electromagnetic interference risks, enhances maintenance convenience and controllability, strengthens adaptability to complex environments, and improves the overall operational quality and reliability of the electricity meter acquisition system.