BROWSE
Detail
Journal Article
Real-Time Terrain-Following Guidance for Mitigation of Pilot-Induced Oscillations
Cited - time in 
Share 
Share
- Authors
- Kwangwoo Jang, Youngjae Kim, Sukjae Im, Man Jun Koh, Hyochoong Bang
- Citation
- Journal of Guidance, Control, and Dynamics, pp.1-17
- ISSN
- 0731-5090
- Publisher
- American Institute of Aeronautics and Astronautics
- Language
- English
- Type
- Journal Article
- Abstract
- Terrain-following (TF) guidance is essential for evading radar detection and enabling safe low-altitude flight. While optimal-control-based TF methods have addressed terrain constraints, they remain limited by computational complexity, input saturation, and pilot-induced oscillations (PIOs). PIOs are particularly hazardous, as they induce sustained oscillations that can cause excessive overshoot over high terrain or collisions. This paper presents a real-time guidance framework that embeds a human pilot model into a model predictive path integral (MPPI) controller with terrain-dependent weighted cost functions. A path-integral-based parameter estimation (PIPE) algorithm is developed to adaptively identify pilot gain and delay online, enabling in-flight suppression of PIOs. Simulation results demonstrate that the proposed MPPI–PIPE framework reduces terrain-following oscillation amplitude by 36.8% compared to MPPI without pilot adaptation, while maintaining real-time feasibility through GPU-parallelized trajectory sampling. Hardware-in-the-loop simulation results further confirm effective PIO mitigation and accurate low-altitude tracking across diverse pilot characteristics.
- Keyword
- Pilot Induced Oscillation, Aerospace Engineering, Optimal Control Problem, Radar Detection, Hardware in the Loop SimulationModel Reference Adaptive ControlFlight Control SurfacesAircraft Flight Control SystemMilitary TechnologyHeight Above Ground Level
- KSP Keywords
- Aerospace engineering, Computational complexity, Cost Function, Hardware-in-The-loop simulation, In-flight, Induced oscillations, Model predictive, Parameter estimation, Path integral, Pilot model, Real-time guidance