Uncertainty Quantification: Making useful predictions about antenna systems
An antenna system never behaves exactly as predicted in simulations once it has been built and measured, because of small (or large) deviations in the realized system as compared to the ideal and nominal system analysed in the computer. Simulations and measurements therefore in principle never agree, and most scientists and engineers, in lack of a better approach, resort to a typically subjective assessment of “how good the agreement is”.
Uncertainty Quantification (UQ) is an alternative and fast way of simulating the expected behavior of the antenna system, with uncertainties on parameters of the antenna system – such as dimensions, angles, material properties, etc. – as inputs. The output is intervals around the ideal and nominal behavior of the antenna quantities of interest, for instance the radiation pattern or S-parameters. With this given, measurements can be compared not to the expected nominal behavior, but to the UQ intervals, and so-called “good agreement” is found if measurements fall within this interval – a procedure based purely on an objective assessment.
In the Fall of 2016, Jan S Hesthaven gave a short course at EPFL (École polytechnique fédérale de Lausanne) about so-called non-intrusive UQ, which perfectly fits the needs of TICRA customers to quantify the effects of stochastic design/manufacturing changes on a wide range of parameters. In early 2017, a first simple prototype of the UQ software is implemented at TICRA and compared to conventional Monte Carlo simulations, which shows promising results for the UQ approach.
In 2017-2018, TICRA discusses the use of UQ methods for the European Space Agency – ESA’s Biomass mission, and in early 2020 TICRA wins an ESA GSTP project to further develop the UQ software. This project is successfully concluded in December 2021.
In the first months of 2022, TICRA further matures the UQ software. At the EuCAP 2022 conference in Madrid, the UQ software is demonstrated to the antenna community and industry at TICRA’s industrial workshop, with a strong interest shown.
Later in 2022, the UQ software will be released with TICRA Tools in a beta version to be tested by selected customers.
TICRA anticipates that it can revolutionise how antennas and antenna systems are designed and manufactured, as the UQ software will give antenna engineers hitherto invaluable information about the most sensitive parts of the system. This, in turn, may pave the way for better designs and substantial cost savings in future satellite mission.
More details on TICRA’s UQ project with ESA: Link
Photo Credit: Airbus Defence and Space, ESA’s Biomass satellite