Abstract
This paper presents a preliminary version of an active learning (AL) scheme for the sample selection aimed at the development of a surrogate model for the uncertainty quantification based on the Gaussian process regression. The proposed AL strategy iteratively searches for new candidate points to be included within the training set by trying to minimize the relative posterior standard deviation provided by the Gaussian process regression surrogate. The above scheme has been applied for the construction of a surrogate model for the statistical analysis of the efficiency of a switching buck converter as a function of seven uncertain parameters. The performance of the surrogate model constructed via the proposed active learning method is compared with that provided by an equivalent model built via a Latin hypercube sampling. The results of a Monte Carlo simulation with the computational model are used as reference.
Highlights
Uncertainty quantification represents a key resource for the design of complex electronic devices, since it allows quantifying statistically the effect of possible uncertain design parameters on the system performance [1].Monte Carlo (MC) simulation can be seen as the most straightforward way to carry out the above statistical analysis
This paper presents a preliminary version of an Active Learning (AL) scheme for the sample selection aimed at the development of a surrogate model for the uncertainty quantification based on the Gaussian Process regression
The AL technique presented in the previous section has been applied for the uncertainty quantification of the DC efficiency η, of the switching buck converter shown in Figure 1, as a function of 7 parameters (i.e., d = 7)
Summary
Uncertainty quantification represents a key resource for the design of complex electronic devices, since it allows quantifying statistically the effect of possible uncertain design parameters (e.g., the components tolerances) on the system performance [1].Monte Carlo (MC) simulation can be seen as the most straightforward way to carry out the above statistical analysis. Within the plain implementation of the MC method, the deterministic simulations are run with the so-called computational model Such deterministic model can be considered as the most accurate synthetic approximation of the system under modeling able of providing, for any configurations of the system parameters, a prediction of the system outputs. Despite its accuracy, such plain implementation of the MC method turns out to be computationally heavy, since, in order to guarantee the convergence of the statistical quantities of interest (e.g., means and standard deviation), it requires to run a large number of simulations (usually in the order of thousands) with the expensive computational model
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.