Control |

Omnicopters for Spacecraft Simulation

Wed, 01 May 2024 00:00:00 +0000

Senior design project and MS thesis component, funded by the LINCS Lab at the Air Force Research Laboratory ($5,000 CRADA award).

Omnicopters — multirotor vehicles with full 6-DOF control authority — were used to simulate spacecraft attitude and translation dynamics in a hardware-in-the-loop configuration. Key contributions:

Determined optimal motor placements and orientations for full controllability
Developed a novel control scheme for rejecting air disturbances in indoor environments
Validated the platform as a low-cost proxy for orbital spacecraft dynamics testing

Autonomous Vehicles with Safety Guarantees

Thu, 01 Feb 2024 00:00:00 +0000

A demonstration suite for safe autonomous vehicle trajectory optimization.

Lane, vehicle, pedestrian, traffic light, and sign detection pipelines
Autonomous lane change via nonlinear model predictive control (nMPC)
Safety-guaranteed lane change using Control Barrier Functions integrated into nMPC (CBF-nMPC), with formal guarantees of non-collision with simulated traffic

Adaptive HVAC for Unknown Deployment

Mon, 01 Jan 2024 00:00:00 +0000

A data-driven approach to HVAC control for environments with unknown thermal characteristics.

Created a heat transfer model of a standard office/warehouse environment equipped with an HVAC system
Embedded physical information of a general environment into ARX for model identification
Gradient descent was used to optimize a PID controller to regulate the internal temperature

Automated Turret

Mon, 01 Jan 2024 00:00:00 +0000

An automated turret with vision-based target tracking and feedback control.

Detected a desired target via a blob detection scheme
Performed pixel-to-torque feedback control to aim the turret
Fine-tuned the controller after determining optimal gains from data-driven transfer function discovery

Reinforcement Learning Control Demonstration

Sun, 01 Jan 2023 00:00:00 +0000

An outreach demonstration introducing reinforcement learning control to general audiences.

Gave an introductory presentation at an appropriate technical level for non-specialists
Developed a set of demonstrations of reinforcement learning control
Performed algorithm benchmarking across multiple RL approaches

Trebuchet Competition

Sun, 01 Jan 2023 00:00:00 +0000

Trebuchet design and optimization as part of Rocker Robotics competition activities (2021–2023).

Assisted in trebuchet building and data collection for Rocker Robotics
Performed model identification and regression to determine optimal counterweight masses and tang angles

Terrain Investigational and Navigational Automaton (TINA)

Sat, 01 Jan 2022 00:00:00 +0000

A differential drive robot built for autonomous course navigation using computer vision feedback control.

Built a differential drive robot with camera module
Utilized computer vision techniques to determine desired path through the course
Performed pixel-to-torque control to successfully navigate the course