Method and device for constructing high-dimensional approximation model in ship type optimization

By decomposing the high-dimensional ship type optimization problem into low-dimensional sub-problems, constructing first-order at most order sub-functions and iteratively optimizing, the problems of parameter explosion and accuracy attenuation in high-dimensional ship type optimization are solved, and efficient modeling accuracy control is achieved.

CN120372822AActive Publication Date: 2025-07-25WUHAN UNIV OF TECH
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202510528763.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-04-25
Publication Date
2025-07-25
Estimated Expiration
2045-04-25

AI Technical Summary

Technical Problem

When dealing with the high-dimensional ship type optimization problem, the existing traditional response function modeling method faces the problem of increasing fitting error caused by the explosion of the number of model parameters and the complex interaction between high-dimensional variables. It is difficult to accurately characterize the combined effect of multivariables, resulting in a decrease in modeling accuracy.

Method used

By decomposing the high-dimensional ship type optimization problem into multiple low-dimensional sub-problems, identifying the variable coupling relationship and constructing first-order, second-order and multi-order sub-functions, combining experimental design methods to construct a mathematical expression model in the local variable space, and iteratively optimized through convergence judgment criteria, and finally combining it into an overall mathematical model.

Benefits of technology

It significantly improves modeling accuracy and calculation efficiency, realizes structured decomposition of complex high-dimensional nonlinear response relationships, ensures controllable local model accuracy, and breaks through the parameters explosion and accuracy attenuation problems of traditional methods in high-dimensional modeling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120372822A_ABST
    Figure CN120372822A_ABST
Patent Text Reader

Abstract

The invention provides a construction method and device for a high-dimensional approximation model in ship form optimization, and relates to the technical field of three-dimensional modelling, and the method comprises the steps: determining a target response function of a ship form optimization problem, and constructing a ship form parameterized model, an optimization variable set for ship form optimization, and a value range of each variable; the optimization variable set is decomposed into a plurality of sub-variable combinations according to the coupling relation between the optimization variables, and a plurality of sub-functions are constructed according to the number of variables contained in each sub-variable combination; for a sub-problem corresponding to each sub-function, selecting a sample point in a corresponding variable space based on an experimental design method, and constructing a low-dimensional mathematical expression model of the sub-problem; and adding all the constructed mathematical expression models according to the combination form corresponding to the sub-functions to obtain an overall mathematical model of the original high-dimensional ship type optimization problem. The method has a dimension decomposition capability and a local precision control capability, so that the problem that current high-dimensional response function modeling is difficult to apply is solved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Global sensitivity analysis method for improving usefulness of complex information system

    CN107886161A

  • Efficient design optimization method and system for super-large variable turbomachinery and application

    CN114169100A

  • Ship motion attitude prediction method based on multiple combinations

    CN115994481A