OPtimization Toolkit for Highly NOn-linear Systems (OPTHiNOS)
OPtimization Toolkit for Highly NOn-linear Systems (OPTHiNOS)
Embedded Files
The OPTHiNOS is a scientific software for Sensitivity Analysis (SA), Uncertainty Quantification (UQ), and optimization of large-scale chaotic dynamical systems. This novel approach leverages machine learning techniques associated with novel algorithms developed from high-level mathematical approaches. The OPTHiNOS works based on the shadowing lemma, a well-known theorem in the dynamical system studies, coupled with non-linear model reduction, which is featured by the minimum residual property.
Recent Findings
Recent Findings
Highlighted Work
Highlighted Work
Sensitivity analysis of chaotic dynamical systems
Conventional sensitivity analysis
Conventional sensitivity analysis
Conventional sensitivity analysis of chaotic dynamical systems usually fail due to strong non-linearity in the corresponding governing equations.
Present sensitivity analysis
Present sensitivity analysis
The OPHTiNOS can efficiently compute the sensitivity solutions within a reasonable uncertainty range. It also reduces the computational memory usage up to 4,000 times than conventional approaches.
PDE-constrained optimization in the presence of massive flow separations
NACA 0012 at AOA=25 degrees.
NACA 0012 at AOA=25 degrees.
A NACA 0012 was considered as a baseline design at AOA=25 degrees. The effort was made to reduce the drag force, while the lift force remains constant.
Optimized airfoil at AOA=25 degrees.
Optimized airfoil at AOA=25 degrees.
The optimized airfoil by the OPTHiNOS can have about 30% less drag force with almost the lift force generation.
"Stability" vs "Instability" for simple dynamical systems
Instability in conventional methods
Instability in conventional methods
Direct solution of the sensitivity function via explicit schemes are highly sensitive to numerical errors, and hence, the sensitivity simulations usually crash.
Stability by the OPTHiNOS
Stability by the OPTHiNOS
The OPTHiNOS converts initial value problems into the boundary value problems, which guarantees the stability of the sensitivity simulations.
Unsteady aerodynamic optimization of flapping wings
Fluid-Structure Interaction (FSI) problems
Fluid-Structure Interaction (FSI) problems
One of well-known FSI problems is birds flying at which complex flow structures are created due to the dynamic stall phenomenon. This problem is called challenging since the flow is highly unstable in the wake of the wings.
Unsteady flow physics in the OPTHiNOS
Unsteady flow physics in the OPTHiNOS
The OPTHiNOS can effectively handle complex optimization problems for flapping wing problems. Here, the thrust is increased, while the lift force remains unchanged to keep the balance with the bird's weight.
Media
Media
Tweeter
Tweeter
Any industrial collaborations are accepted for cutting-edge projects.
Any industrial collaborations are accepted for cutting-edge projects.
Questions?
Questions?
Contact me to get more information on the projects.
Page updated
Google Sites
Report abuse