Machine learning surrogates of high-fidelity electrical models
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Parametric sensitivities of dynamic system responses are very useful in a variety of applications, including circuit optimization and uncertainty quantification. Sensitivity calculation methods fall into two related categories: direct and adjoint methods. Effective implementation of such methods in a production circuit simulator poses a number of technical challenges, including instrumentation of device models. This report documents several years of work developing and implementing di- rect and adjoint sensitivity methods in the Xyce circuit simulator. Much of this work sponsored by the Laboratory Directed Research and Development (LDRD) Program at Sandia National Labora- tories, under project LDRD 14-0788.
This report summarizes the methods and algorithms that were developed on the Sandia National Laboratory LDRD project entitled "Advanced Uncertainty Quantification Methods for Circuit Sim- ulation", which was project # 173331 and proposal # 2016-0845. As much of our work has been published in other reports and publications, this report gives an brief summary. Those who are in- terested in the technical details are encouraged to read the full published results and also contact the report authors for the status of follow-on projects.