Feedback Control Theory: Architectures and Tools for Real-Time Decision Making I
summary
Control theory frames real-time decisions as managing dynamics plus uncertainty with feedforward and feedback — memorably DUFF. It uses observers, online optimization and temporal-logic verification; examples include a DARPA Grand Challenge vehicle and a flight demo that computes trajectories every 0.1 s for the next 1–2 s using receding horizon control (early implementation cost ~$30,000). Architectures combine a predictive trajectory generation module with a fast feedback compensator so planned maneuvers are recomputed and corrected in real time.
AI-written summary · about
chapters (5)
- 0:00Start
- 1:47Traditional view
- 8:14Online Optimization-based control
- 12:05Control Systems: Architectures and Examples
- 49:43Reactive compensation
related talks
57mReal-Time Decision Making in Hybrid Systems
49mUsing Deep Learning for Perception in Autonomous Systems: A Perspective from Control Theory
17mData-flow Availability: Achieving Timing Assurance in Autonomous Systems
33mHuman-Like Behavior in Autonomous Systems
30mOn the Aspects of AI and Robotic Autonomy
15mOptimal Control for Electromagnetic Haptic Guidance Systems
19mAirways: Optimization-Based Planning of Quadrotor Trajectories according to High-Level User Goals
20mInformation Gain Regulation in Reinforcement Learning with the Digital Twins' Level of Realism In RlLib