TY - GEN

T1 - An h-adaptive stochastic collocation method for stochastic EMC/EMI analysis

AU - Yücel, Abdulkadir C.

AU - Bagci, Hakan

AU - Michielssen, Eric

N1 - KAUST Repository Item: Exported on 2020-10-01

PY - 2010/7

Y1 - 2010/7

N2 - The analysis of electromagnetic compatibility and interference (EMC/EMI) phenomena is often fraught by randomness in a system's excitation (e.g., the amplitude, phase, and location of internal noise sources) or configuration (e.g., the routing of cables, the placement of electronic systems, component specifications, etc.). To bound the probability of system malfunction, fast and accurate techniques to quantify the uncertainty in system observables (e.g., voltages across mission-critical circuit elements) are called for. Recently proposed stochastic frameworks [1-2] combine deterministic electromagnetic (EM) simulators with stochastic collocation (SC) methods that approximate system observables using generalized polynomial chaos expansion (gPC) [3] (viz. orthogonal polynomials spanning the entire random domain) to estimate their statistical moments and probability density functions (pdfs). When constructing gPC expansions, the EM simulator is used solely to evaluate system observables at collocation points prescribed by the SC-gPC scheme. The frameworks in [1-2] therefore are non-intrusive and straightforward to implement. That said, they become inefficient and inaccurate for system observables that vary rapidly or are discontinuous in the random variables (as their representations may require very high-order polynomials). © 2010 IEEE.

AB - The analysis of electromagnetic compatibility and interference (EMC/EMI) phenomena is often fraught by randomness in a system's excitation (e.g., the amplitude, phase, and location of internal noise sources) or configuration (e.g., the routing of cables, the placement of electronic systems, component specifications, etc.). To bound the probability of system malfunction, fast and accurate techniques to quantify the uncertainty in system observables (e.g., voltages across mission-critical circuit elements) are called for. Recently proposed stochastic frameworks [1-2] combine deterministic electromagnetic (EM) simulators with stochastic collocation (SC) methods that approximate system observables using generalized polynomial chaos expansion (gPC) [3] (viz. orthogonal polynomials spanning the entire random domain) to estimate their statistical moments and probability density functions (pdfs). When constructing gPC expansions, the EM simulator is used solely to evaluate system observables at collocation points prescribed by the SC-gPC scheme. The frameworks in [1-2] therefore are non-intrusive and straightforward to implement. That said, they become inefficient and inaccurate for system observables that vary rapidly or are discontinuous in the random variables (as their representations may require very high-order polynomials). © 2010 IEEE.

UR - http://hdl.handle.net/10754/564290

UR - http://ieeexplore.ieee.org/document/5562200/

UR - http://www.scopus.com/inward/record.url?scp=78349240752&partnerID=8YFLogxK

U2 - 10.1109/APS.2010.5562200

DO - 10.1109/APS.2010.5562200

M3 - Conference contribution

SN - 9781424449682

BT - 2010 IEEE Antennas and Propagation Society International Symposium

PB - Institute of Electrical and Electronics Engineers (IEEE)

ER -