Optimization Strategy Using Dynamic Radial Basis Function Metamodel Based on Trust Region

2014 ◽  
Vol 50 (7) ◽  
pp. 184 ◽  
Author(s):  
Teng LONG
Keyword(s):  
Radial Basis Function ◽  
Basis Function ◽  
Trust Region ◽  
Optimization Strategy ◽  
Radial Basis

scholarly journals Efficient Solving of Boundary Value Problems Using Radial Basis Function Networks Learned by Trust Region Method

2018 ◽  
Vol 2018 ◽  
pp. 1-4 ◽  
Author(s):  
Mohie Mortadha Alqezweeni ◽  
Vladimir Ivanovich Gorbachenko ◽  
Maxim Valerievich Zhukov ◽  
Mustafa Sadeq Jaafar
Keyword(s):  
Radial Basis Function ◽  
Boundary Value Problems ◽  
Basis Function ◽  
Boundary Value ◽  
Trust Region Method ◽  
Trust Region ◽  
Radial Basis Function Networks ◽  
Structure Selection ◽  
Region Method ◽  
Radial Basis

A method using radial basis function networks (RBFNs) to solve boundary value problems of mathematical physics is presented in this paper. The main advantages of mesh-free methods based on RBFN are explained here. To learn RBFNs, the Trust Region Method (TRM) is proposed, which simplifies the process of network structure selection and reduces time expenses to adjust their parameters. Application of the proposed algorithm is illustrated by solving two-dimensional Poisson equation.

Algorithm for Wireless Sensor Network Data Fusion Based on Radial Basis Function Neural Networks

2014 ◽  
Vol 577 ◽  
pp. 873-878 ◽  
Author(s):  
Zi Yang ◽  
Ming Rui Chen ◽  
Wei Wu
Keyword(s):  
Neural Networks ◽  
Wireless Sensor Network ◽  
Data Fusion ◽  
Radial Basis Function ◽  
Sensor Network ◽  
Basis Function ◽  
Wireless Sensor ◽  
Optimization Strategy ◽  
Radial Basis ◽  
Neural Networks Optimization

This paper presents a data fusion method on wireless sensor network based on radial basis function neural networks. In consideration of the hierarchical relationship topology wireless sensor networks, data acquisition, handling and delivery, we proposed a typical classification approach based on radial basis function neural networks. Optimization strategy adopted to process node data for each node indicated different reaction related with energy consumption. Simulation results verify that the method converges fast and effectively.

Bow shape modification through multi-objective hydrodynamic optimization: Methodology comparison between CAD-based FreeForm Deformation and Mesh-based Radial Basis Function approach

2022 ◽  
pp. 147509022110689
Author(s):  
Parviz Ghadimi ◽  
Amin Nazemian
Keyword(s):  
Radial Basis Function ◽  
Basis Function ◽  
Total Resistance ◽  
Optimization Strategy ◽  
Multi Objective ◽  
Global Recession ◽  
Design Variables ◽  
Radial Basis ◽  
Freeform Deformation ◽  
Optimization Methodology

Marine industrial engineering face crucial challenges because of environmental footprint of vehicles, global recession, construction, and operation cost. Meanwhile, Shape optimization is the key feature to improve ship efficiency and ascertain better design. Accordingly, the present paper proposes an automated optimization framework for ship hullform modification to reduce total resistance at two cruise and sprint speeds. The case study is a bow shape of a wave-piercing bow trimaran hull. To this end, a multi-objective hydrodynamic problem needs to be solved. A combined optimization strategy using CFD hullform optimization is presented using the software tools STAR-CCM+ and SHERPA algorithm as optimizer. Furthermore, a comparison is made between CAD-based and Mesh-based parametrization techniques. Comparison between geometry regeneration methods is performed to present a practical and efficient parametrization tool. Design variables are control points of FreeForm Deformation (FFD) for CAD-based method and Radial Basis Function (RBF) for Mesh-based method. The optimization results show a 4.77% and 2.47% reduction in the total resistance at cruise and sprint speed, respectively.

Multi-Disciplinary Optimization of Space Experimental Device Using Radial Basis Function in ModelCenter

2018 ◽  
Author(s):  
Lyu Wang ◽  
Yuan Yun ◽  
Bin Zhang ◽  
Tao Zhang
Keyword(s):  
Radial Basis Function ◽  
Conceptual Design ◽  
Basis Function ◽  
Optimization Algorithms ◽  
Space Station ◽  
Simulation Models ◽  
Optimization Strategy ◽  
Electromagnetic Actuators ◽  
Time Consumption ◽  
Radial Basis

The multi-disciplinary optimization of actuators which is the key component of a Nested Flying Vehicle (NFV) is done through ModelCenter. The NFV is operated inside the space station for space science experiments. The conceptual design of the NFV has been done with expensive simulation models. However, the optimization result of the NFV using the genetic algorithm led to an unaccepted time consumption. In order to solve those problems, a self-programmed optimizer which is embedded the Radial Basis Function (RBF)-based optimization strategy is developed and integrated to ModelCenter and applied to the electromagnetic actuators of NFV. The optimizer helps to integrate more metamodel-based optimization algorithms and enhance the optimization ability of ModelCenter. By comparing with the optimization algorithms supplied by ModelCenter, the RBF-based optimization strategy proposed in this paper costs less time and results in an acceptable design. The conceptual design of the NFV, the methodology of RBF-based optimization strategy and the key components of optimizer are introduced in this paper.

scholarly journals Derivative-Free Multiobjective Trust Region Descent Method Using Radial Basis Function Surrogate Models

2021 ◽  
Vol 26 (2) ◽  
pp. 31
Author(s):  
Manuel Berkemeier ◽  
Sebastian Peitz
Keyword(s):  
Radial Basis Function ◽  
Basis Function ◽  
Optimization Problems ◽  
Trust Region ◽  
Surrogate Models ◽  
Descent Method ◽  
Information Need ◽  
Derivative Free ◽  
Radial Basis ◽  
General Scalar

We present a local trust region descent algorithm for unconstrained and convexly constrained multiobjective optimization problems. It is targeted at heterogeneous and expensive problems, i.e., problems that have at least one objective function that is computationally expensive. Convergence to a Pareto critical point is proven. The method is derivative-free in the sense that derivative information need not be available for the expensive objectives. Instead, a multiobjective trust region approach is used that works similarly to its well-known scalar counterparts and complements multiobjective line-search algorithms. Local surrogate models constructed from evaluation data of the true objective functions are employed to compute possible descent directions. In contrast to existing multiobjective trust region algorithms, these surrogates are not polynomial but carefully constructed radial basis function networks. This has the important advantage that the number of data points needed per iteration scales linearly with the decision space dimension. The local models qualify as fully linear and the corresponding general scalar framework is adapted for problems with multiple objectives.

Sign in / Sign up

Export Citation Format

Share Document