Dynamic PID Controlfor DroneTrajectory Optimization
MATLAB Sandbox for Hybrid Optimization, PID Control, and GMM Target Fields
A compact technical simulation project for trajectory optimization over GMM-style target landscapes, combining optimization logic with a dynamic PID controller to study drone motion, error, and response in one MATLAB sandbox.
A Sandbox, Not a Platform
Simulation-first
The project is a controlled experiment environment, not a deployable autonomy stack. Its value is in making the optimization-plus-control loop inspectable.
One technical loop
Target field, optimizer, controller, simulated motion, and output plots live in one sandbox so behavior can be studied end to end.
Compact scope
The page should stay tool-sized. It should feel technical and real without pretending this repo is broader than it is.
The Technical Loop the Repo Actually Makes Visible
The useful thing about the project is not UI breadth. It is the closed loop between target landscape, optimizer, controller, motion, and output plots.
Why the Controller Layer Was Worth Isolating
Response tuning mattered
The point of the dynamic PID layer was not only stability. It gave a way to shape how aggressively the simulated drone responded to the optimizer’s intended motion.
Optimization needed a controller partner
Optimization logic alone does not say how the simulated vehicle should settle, overshoot, or recover. The PID loop made those behaviors explicit.
Outputs became interpretable
Once trajectory, speed, and error were produced together, the sandbox became a clearer place to compare controller behavior than a loose collection of scripts and plots.
The Outputs That Made the Sandbox Useful
Observe how the simulated path moves across the target landscape under the combined optimizer and controller.
Inspect whether aggressive optimization updates translate into stable or erratic motion commands.
Use error plots to understand settling behavior, transient mismatch, and the practical effect of controller tuning.