Novel Interval Parameter Identification Method Using Augmented Fourier Series-Based Polynomial Surrogate Model

Abstract

Traditional probabilistic-based parameter identification methods cannot operate in practical engineering systems without sufficient available measurements. To overcome this drawback, a novel augmented Fourier polynomial-based interval parameter identification method is proposed in this paper. First, a novel augmented Fourier series-based polynomial surrogate model is established to precisely describe the mathematical relationship between the identified parameters and system responses. Subsequently, to improve the efficiency of the identification procedure, an improved interval-based multi-objective optimization strategy is designed to seek the nominal and fluctuation value of the parameter interval simultaneously. Eventually, the numerical and experimental examples are provided to verify the feasibility of the proposed method. The proposed method can achieve competitive results on various examples, the mass-spring system, and steel plate structure, compared with the previous methods. Moreover, satisfactory results are also obtained when applied to a space truss structure. The identification results demonstrate the feasibility and reliability of the proposed interval parameter identification method.

Keywords

• Interval theory
• parameter identification
• augmented Fourier series-based surrogate model
• multi-objective optimization strategy