Jong-Han Kim - Research Topics

Machine learning based intelligent guidance techniques

• Online -regularized optimization for optical misalignment cancellation of strap-down seekers.

• Target shape identification using RF signals based on unsupervised learning.

Integrated guidance and control for constrained precision homing

• Finite horizon integrated guidance and control for constrained terminal homing of high performance missile systems.

Guidance and control designs for full-scale missile development programs

• Highly reliable guidance and control design/implementation for several state-of-the-art missile systems development programs.

• The works include classical, optimal, or robust control for highly unstable airframes and high-fidelity flight simulation techniques.

Optimal decentralized control for large-scale networked systems

• Theoretical studies on optimal decentralized control using convex analysis and convex optimization.

• Characterizes convex problems and explicitly solvable problems in optimal control of networked systems under information constraints.

Decentralized control for jet engine systems

• Suboptimal decentralized control synthesis.

• Convex relaxation and convex approximation of nonconvex optimal control problems.

• Application to jet engine control systems.

Multiscale/stochastic consensus for decentralized estimation

• Multiscale concepts to applied to consensus problems, for scalable distributed estimation.

• Stochastic message passing for convergence acceleration. Application to decentralized estimation problems.

Tactical missile systems integration

• Mission profile design, operational performance/effectiveness analysis.

• Flight performance analysis.

• Engineering level 6-DOF flight dynamics simulation.

Attitude control for small satellites

• Design and experiments for attitude control of flexible satellites.

• Ground test modules development.

Co-evolutionary algorithms for constrained optimization

• Constrained optimization, minimax optimization using evolutionary computation.

• Applications to design problems or control problems.

• Convergence acceleration by using artificial neural networks