Mixed Uncertainty Quantification and Imprecise Sensitivity Analysis for EMC Assessment Based on Probability Box Theory and Polynomial Chaos
Yinhao Wang,Gang Li,Quanyi Yu,Haolin Jiang,Shuai Pi,Tianhao Wang,Yaodan Chi
DOI: https://doi.org/10.1109/temc.2024.3364710
IF: 2.036
2024-01-01
IEEE Transactions on Electromagnetic Compatibility
Abstract:Electromagnetic compatibility (EMC) of electrical equipment is frequently affected by a lot of aleatory and epistemic uncertainties, which may result in a serious decline in the EMC of equipment. Accordingly, the mixed epistemic–aleatory uncertainty quantification (UQ) is crucial in the performance assessment and optimization of EMC. This study proposes a mixed UQ and imprecise sensitivity analysis (ISA) method based on probability box (p-box) theory and polynomial chaos (PC). The method treats the epistemic uncertainty of the parameters in the p-box as a quadratic polynomial function of the PC coefficients, thereby quantifying the mixed uncertainty by calculating the p-box of the response. We also further combine the PC-pbox and Sobol sensitivity analysis methods to conduct ISA on each input variable, quantifying and separating the effect of aleatory and epistemic uncertainties on EMC. Finally, taking the problem of the transmission line crosstalk as an example, this study quantifies the mixed uncertainty and ISA of the induced current. The results show that the proposed method can efficiently and accurately realize the mixed UQ and ISA of the EMC evaluation and provide a theoretical basis for EMC assessment and optimization.
telecommunications,engineering, electrical & electronic