Research

Research Topic

AMDC Lab은 자율주행 자동차와 비행체를 위한 의사결정 및 제어 기술을 연구합니다. 인지, 추정, 예측, 경로계획, 제어를 AI와 통합하여 다양한 환경에서 최적 경로 생성과 안정적 제어를 구현을 통해 자율 이동 시스템의 안전성과 성능을 향상시키는 것을 목표로 합니다.

AMDC Lab researches decision-making and control for autonomous driving and flight systems. It integrates perception, estimation, prediction, planning, and control with AI to achieve optimal path generation and stable control in diverse environments.

AMDC Lab

Vehicle Dynamics Modeling
Perception and Estimation
Path Planning and Control
AI Application
Multi-sensor Integration
Minimum Risk Maneuver
Intelligent Control System

Vehicle Dynamics Modeling

Research targets: Passenger vehicles, Off-road vehicle, Heavy-duty buses, Semi-trailer and Double-trailer trucks, 6x6/8x8 specialty vehicles, etc.
Scope: Kinematic and dynamic modeling, multi-articulation modeling, experimental vehicle dynamics analysis, and simulation-based validation
Applications: Autonomous driving, freight transport, military and industrial vehicles, and innovative vehicle control system development

Vehicle dynamics and system modeling are extensively analyzed and mathematically modeled for various vehicles across diverse road, industrial, and defense environments. The focus lies on precise prediction and control of vehicle motion through kinematic and dynamic modeling in complex operational scenarios.

Autonomous Driving

Perception
Localization
Control
End-to-End Network

Prediction/Estimation

Trajectory prediction
Environment
State Estimation
Fault Detection

Motion Planning

Longitudinal & Lateral Motion Planning
Motion Planning Considering Vehicle Dynamics and Ride Comfort
Emergency Maneuver Planning
Optimal Path Generation for Collision Avoidance
Design of Abnormal Driving Scenarios
Real-Time Trajectory Optimization for Dynamic Environments

Control System

Autonomous Driving Control
Optimal Chassis Control Systems
Control Based on Control Theory
Optimal Control Based on Reinforcement Learning
Robust Control Design for Uncertain Systems
Adaptive Control Using Neural Networks

AAV(Advanced Air Vehicle) Autonomous Flight

Robust Real-Time State Estimation and Control System

Enhancing AAV system stability through AI-based state estimation and optimal control in complex urban environments and disturbed conditions.

State Observer Design: Hybrid observer fuses state-space models and deep learning for accurate pose recovery during disturbances.
Optimal Path Planning: Real-time trajectory optimization avoids obstacles and ensures safe recovery paths in abnormal scenarios.
Optimal Motion Control: AI-MPC based control minimizes tracking error while handling actuator limits and urban constraints.
Fail Safety in Complex Urban Environments: Redundant control ensures collision-free flight near crowds and infrastructure.
End-to-End Flight System: Integrated pipeline from state estimation to motion execution enables robust autonomous AAV operations.