Simultaneous Multi-Parameter Identification Algorithm for Clearance-Type Nonlinearity
Bingbing Hu,Jie Liu,Shanhui Liu,Bing Li,Xiaofei Lei
DOI: https://doi.org/10.1016/j.ymssp.2019.106423
IF: 8.4
2020-01-01
Mechanical Systems and Signal Processing
Abstract:Clearance which is a typical nonlinear factor is unavoidable in the engineering structure. Clearance would constantly change during service, thus exceeding the reasonable design range. The improper clearance can change the contact state of parts to degrade the dynamical performance of the mechanical system. Nonlinear parameter identification is an efficient tool for understanding the characteristic of clearance nonlinearity and contributing significantly to control the clearance effect. In this paper, the simultaneous identification algorithm is proposed to simultaneously identify two clearance parameters, clearance value and effective stiffness, and the transfer error existing in the previous multi-step identification method (Li et al., 2015) can be avoided so as to improve the identification accuracy. For the proposed simultaneous identification algorithm, on the one hand, Tikhonov regularization method is used to solve a nonlinear force calculation formula which is formed by the transmission characteristic of the clearance-type system to directly obtain the nonlinear force. On the other hand, clearance nonlinear force can also be predicted by the mechanics characteristic of clearance nonlinearity. And then, the clearance parameters can be simultaneously obtained by an iterative parametric optimization process, wherein the objective is to minimize the error between the calculated and predicted nonlinear forces. The feasibility of simultaneous identiffcation algorithm is verified by simulation data from a cantilever with clearance nonlinearity. The influence factors, including initial guess, non-equivalence clearance parameters and other nonlinearities, are fully discussed to illustrate the robustness of simultaneous identification algorithm. The experimental data from the clearance test-bed are utilized to verify the effectiveness of the simultaneous identification algorithm. Numerical and experimental studies show that the proposed identification algorithm can precisely and simultaneously identify the clearance value and effective stiffness, and compared with previous multi-step identification method, simultaneous identification algorithm is more accurate and stable. (C) 2019 Elsevier Ltd. All rights reserved.