Unicycle-hero.gitlab !!top!! ❲2024❳

Self‑balancing unicycles combine the agility of a single‑wheel platform with the challenge of maintaining dynamic stability. They serve as an ideal test‑bed for control theory, sensor fusion, and embedded systems. Existing commercial solutions typically expose only high‑level APIs or proprietary firmware, limiting their use in research and education. Moreover, most academic publications provide only a description of the control algorithm without releasing the accompanying software, hindering reproducibility.

Overall worst‑case execution time (WCET) = , well within the 5 ms deadline. The system employs stack overflow checks and a hardware watchdog that resets the board if the control task fails to execute for > 10 ms. unicycle-hero.gitlab

with (\phi(\vartheta) = \tanh(\vartheta / \theta_\max)). This yields smooth saturation and avoids wind‑up. with (\phi(\vartheta) = \tanh(\vartheta / \theta_\max))

The state vector is . Process and measurement models follow the standard bicycle dynamics (see Section 4.1). The EKF runs at 200 Hz, matching the IMU update rate. unicycle-hero.gitlab

The repository follows the pipeline: