ECE & embedded
Start on the bench. Turn IMU measurements into a stable attitude controller before attempting flight.
Inspired by Cornell University
Your proposed build
A smaller scope. A useful outcome.
Build a simulated quadcopter and a constrained, single-axis attitude rig. Log disturbances and controller response. Free flight is an optional, separately supervised extension.
The scope, timebox and success targets below are Noetic suggestions. They are separate from the university team's work.
The problem
A drone is a useful control-systems project only when its sensing, timing and failure modes can be measured and understood.
Before you start
- C and timed interrupts
- PID control and basic dynamics
- IMU calibration
Components & tools
- Microcontroller and IMU
- Simulation environment
- Constrained test rig
- Serial data logger
Choose parts after checking your lab's equipment and local availability.
Build in stages
Model one axis and compare sensor noise at rest.
Tune the controller in simulation with repeatable disturbances.
Run the constrained rig and compare its measured response with the model.
Document loss-of-signal behavior and a physical stop control.
Define success
- Track a small commanded angle with less than 5 degrees steady-state error on the constrained rig.
- Demonstrate a stop command and log at least five repeatable disturbance trials.
Proposed evaluation targets, not reported university results.
Before you build
Remove propellers for firmware work. Powered rotor tests require a guarded rig, an experienced supervisor and a suitable battery procedure. Check local permissions before any flight.
University inspiration / Student project
Autonomously Hovering Quadcopter
Cornell University / ECE 4760 / Fall 2018
The student report describes a PIC32 quadcopter with IMU feedback, PID attitude control and radio telemetry.