[L4DC 2025] Automatic hyperparameter tuning for DeePC. Built by Michael Cummins at the Automatic Control Laboratory, ETH Zurich.
Model Predictive Path Integral Control for RoboNav Mars Rover
Code for automated operation, basic (PID) and advanced (optimal) control of an atmospheric pressure plasma device in Helium
A physics-based flight simulator for optimizing airbrake deployment in model rockets to achieve precise target apogee. Uses predictive PID control to dynamically adjust airbrakes during flight based on real-time apogee predictions.
Public experiment log using Get Physics Done (GPD) with Codex to explore predictive control of tokamak plasma turbulence and confinement.
Simulation code, orbital data and analysis scripts for UAV-swarm distributed-aperture reception of LEO satellites — prediction-free geometric control beats a learned predictive one
PMSM dq digital twin comparing PI field-oriented and predictive current control.
ROS2 trajectory-sampling predictive controller for F1TENTH autonomous racing with raceline tracking and LiDAR-reactive obstacle avoidance.
Experimental, unofficial Python controller for Mitsubishi mini-split heat pumps on Kumo Cloud and the Mitsubishi Comfort app. An alternative to Auto mode: stays in heat, steers the setpoint from a 15-minute temperature prediction, cools only for solar gain, and logs every cycle to CSV.
DeePC sandbox — developing, testing, and benchmarking Data-Enabled Predictive Control algorithms
To associate your repository with the predictive-control topic, visit your repo's landing page and select "manage topics."