首页|Compensatory evaluation of frequency evaluation metrics based on empirical mode decomposition
Compensatory evaluation of frequency evaluation metrics based on empirical mode decomposition
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Techno-Press
Experimental evaluation is crucial for ensuring the accuracy of real-time hybrid simulation (RTHS) results. While existing methods can calculate time delay and amplitude error, time-varying delays can destabilize RTHS, requiring a method to account for them. This paper proposes using empirical mode decomposition (EMD) to calculate instantaneous control parameters, such as instantaneous amplitude and time delay. Intrinsic mode functions (IMF) from EMD capture the signal's local characteristics at different time scales, and the Hilbert transform is applied to compute these parameters. After EMD, a different number of IMFs may be obtained for calculated displacements than for measured displacements, and this paper gives advice on the IMFs needed to calculate instantaneous control parameters (ICP), and how they should be matched. The signals obtained from the numerical simulation of the benchmark model without and with compensation are firstly subjected to the calculation of ICP, and the results prove the effectiveness of ICP. Subsequently, the predefined displacement test with a multi-degree-of-freedom structure and the RTHS with a single-degree-of-freedom structure and self-centering viscous damper were subjected to the traditional ICP method and the EMD-based ICP calculation method for ICP calculation, respectively, and the comparative results show that the effectiveness of the EMD-based ICP calculation method is increased, and that EMD effectively solves the negative-frequency issue caused by signals with multiple poles between two crossed zeros. These calculations show great potential in improving experimental evaluations.
empirical mode decompositioninstantaneous control parametersinstantaneous frequencyreal-time hybrid simulationHYBRID SIMULATIONTIME
NETL
NSTL
Techno-Press
