Efficient Low-Frequency Human Exposure Assessment with the Maximum Entropy Snapshot Sampling

Reduced-order models of high-fidelity linear systems enable efficient and comprehensive numerical assessment of human exposure to low-frequency magnetic fields. These models are often generated by snapshot-based reduced basis methods, which rely on the singular value decomposition of a matrix whose columns are solution vectors of a high-fidelity system. However, most of these solution vectors provide redundant information, which turns the singular value decomposition into a computation-ally and memory-demanding step. With the maximum entropy snapshot sampling strategy, the number of solution vectors can be efficiently reduced to the essential ones. Therefore, this work aims to provide a reduced basis for human exposure assessment in a computationally and memory-efficient manner using this information-theoretic framework.