Abstract:Currently, approximation models have been widely adopted to replace the actual computationally expensive experiments or simulations in engineering design optimisation. However, it is not easy to create an accurate approximation model, especially when the computation-intensive high-fidelity (HF) sample is strictly limited to a small size. To ease this problem, a hybrid variable-fidelity global approximation modelling method (hybrid VF) combining tuned radial basis function (RBF) base and kriging correction is proposed in this paper. In the proposed hybrid VF approach, information from both HF and low-fidelity (LF) models are employed. The LF model is first approximated by RBF, and then this RBF LF approximation is pulled to the HF by the linear tuning model to gain a base surrogate. After that, this tuned RBF base is modified by a kriging-based linear correction function. Some numerical and engineering examples are demonstrated to validate the efficiency of the proposed hybrid VF method. The results show that the proposed hybrid VF can achieve a much more accurate approximation than other metamodelling methods in the small HF sample situation.

A Combined Function Approximation Model and Its Application

Quadratic Approximation to Plane Parametric Curves and Its Application in Approximate Implicitization

Fuzzy RBF Neural Network Based Function Approximation

Approximate Merging of a Pair of Bézier Curves

General Approach to Function Approximation

NONLINEAR UNIFORM SIMULTANEOUS APPROXIMATION

A Hybrid Variable-Fidelity Global Approximation Modelling Method Combining Tuned Radial Basis Function Base and Kriging Correction

On Nonlinear Simultaneous Approximation Problems

MULTI-LAYER FUZZY CMAC ADAPTIVE NONLINEAR CONTROL BASED ON AN APPROXIMATION MODEL

Approximation Capability of a Novel Neural Network Model for Dynamic Systems

Nonlinearly Constrained Best Approximation in Hilbert Spaces: the Strong CHIP and the Basic Constraint Qualification.

NONLINEAR SIMULTANEOUS APPROXIMATION IN COMPLETE LATTICE BANACH SPACES

Approximations of Continuous Functionals by Neural Networks with Application to Dynamic Systems

Data-Driven Control Based on Function Approximation with Binomial Coefficients

Smooth Approximation Algorithm Based on Split-Merge Model Trees

On the approximation error in high dimensional model representation

A Novel Nonlinear Approximation Approach- -Enhanced Hinge-Finding Algorithm

Improving Algorithm of Hinging Hyperplanes for Function Approximation

Design Methods of General Fuzzy Systems As Function Approximators

A HOMOTOPY METHOD FOR MULTIKERNEL-BASED APPROXIMATION

Function Approximation with Randomly Initialized Neural Networks for Approximate Model Reference Adaptive Control