Lisa Hood
Lisa Hood has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Akshay Damany
Department of Energy by National Technology and Engineering Solutions of Sandia, LLC P.O. Box 5800 P.O. Box 969 87185-0101 94551-0969 Albuquerque Livermore NM CA
1 shared publication - Bart van Bloemen Waanders
Sandia National Laboratories, Albuquerque, USA
1 shared publication - Gerardo Cruz1 shared publication
- Joseph Hart
Sandia National Laboratories, Albuquerque, New Mexico, 87185, USA
1 shared publication - Julie Parish
Autonomy for Hypersonics, Albuquerque, USA
1 shared publication
Latest Publications
- 13 March 2023
- Published by Springer Nature in Journal of Optimization Theory and Applications
- Vol. 196 (3) , 855-881
Sensitivity-Driven Experimental Design to Facilitate Control of Dynamical Systems
- 13 March 2023
- Published by Springer Nature in Journal of Optimization Theory and Applications
- Vol. 196 (3) , 855-881
Sensitivity-Driven Experimental Design to Facilitate Control of Dynamical Systems
- Open Access
- 1 October 2021
- Published by Office of Scientific and Technical Information (OSTI)
Autopilot Synthesis and Analysis using H-infinity Optimal Control.
This presentation provides an overview of an autopilot synthesis and analysis framework developed in MATLAB/Simulink based on 𝐻 ∞ optimal control methods. The framework was developed to address questions regarding controllability of an early-stage flight vehicle design in the context of an outer mold line (OML) trade study. Current methods for addressing the controllability of preliminary vehicle designs typically involve "static" analyses, including analyses of stability/control derivatives and trim deflections/hinge moments. These analyses are relatively inexpensive and can be done quickly with only preliminary component models. However, they provide only a qualitative understanding of robustness/performance tradeoffs that fundamentally characterize vehicle controllability in a broad sense. Quantitative characterization of robustness and performance is typically relegated to later stages in the design process and involves time-consuming gain tuning and dynamic stability analysis across expensive Monte Carlo simulations. To push quantitative characterization of autopilot design earlier in the design process, this framework implements a modular design and makes use of 𝐻 ∞ optimal control methods across a given controller architecture. The former enables rapid exploration of competing subcomponent model designs and the latter provides efficient multivariate controller tuning and analysis given known performance specifications and uncertainty models. In contrast to Monte Carlo sampling, 𝐻 ∞ methods enable synthesis and analysis across the continuous space of modeled uncertainty using efficient solution methods, thus enabling rigorous quantification of robustness over a larger design space. Specifically, this framework iteratively applies structured 𝜇-synthesis [1, 2] and 𝜇-analysis for tuning to robust performance and analyzing robust stability, respectively. This occurs across numerous relevant flight conditions and across various competing subcomponent models with minimal manual input from the user, enabling rapid autopilot design turnaround to inform broader early-stage system design. This presentation provides an overview of the framework with an emphasis on formulating the 𝐻 ∞ optimal control problem, including discussion on encoding performance specifications as frequency-dependent weighting functions and representing uncertainty in the 𝐻 ∞ framework. The presentation also covers application of the framework to the OML trade study and provides corresponding autopilot synthesis and analysis results.
- 1 October 2021
- Published by Office of Scientific and Technical Information (OSTI)
Open AccessAutopilot Synthesis and Analysis using H-infinity Optimal Control.
This presentation provides an overview of an autopilot synthesis and analysis framework developed in MATLAB/Simulink based on 𝐻 ∞ optimal control methods. The framework was developed to address questions regarding controllability of an early-stage flight vehicle design in the context of an outer mold line (OML) trade study. Current methods for addressing the controllability of preliminary vehicle designs typically involve "static" analyses, including analyses of stability/control derivatives and trim deflections/hinge moments. These analyses are relatively inexpensive and can be done quickly with only preliminary component models. However, they provide only a qualitative understanding of robustness/performance tradeoffs that fundamentally characterize vehicle controllability in a broad sense. Quantitative characterization of robustness and performance is typically relegated to later stages in the design process and involves time-consuming gain tuning and dynamic stability analysis across expensive Monte Carlo simulations. To push quantitative characterization of autopilot design earlier in the design process, this framework implements a modular design and makes use of 𝐻 ∞ optimal control methods across a given controller architecture. The former enables rapid exploration of competing subcomponent model designs and the latter provides efficient multivariate controller tuning and analysis given known performance specifications and uncertainty models. In contrast to Monte Carlo sampling, 𝐻 ∞ methods enable synthesis and analysis across the continuous space of modeled uncertainty using efficient solution methods, thus enabling rigorous quantification of robustness over a larger design space. Specifically, this framework iteratively applies structured 𝜇-synthesis [1, 2] and 𝜇-analysis for tuning to robust performance and analyzing robust stability, respectively. This occurs across numerous relevant flight conditions and across various competing subcomponent models with minimal manual input from the user, enabling rapid autopilot design turnaround to inform broader early-stage system design. This presentation provides an overview of the framework with an emphasis on formulating the 𝐻 ∞ optimal control problem, including discussion on encoding performance specifications as frequency-dependent weighting functions and representing uncertainty in the 𝐻 ∞ framework. The presentation also covers application of the framework to the OML trade study and provides corresponding autopilot synthesis and analysis results.