A novel autonomous bicycle based on control moment gyro technology

Momentum wheel balancing schemes for autonomous bicycles are limited in torque output, load capacity, and continuous-disturbance rejection. This paper presents a dual-rotor single-gimbal control moment gyroscope (SGCMG) stabilizer with dual-mode operation and doubled effective gyroscopic torque. The system is modeled via the Lagrangian method and linearized to a fourth-order state-space form. A fuzzy adaptive PID dual-loop cascade controller achieves the lowest tilt-angle overshoot (-0.05 rad vs. -0.09 rad for cascade PID) and zero steady-state error. Experiments on a 6.77 kg prototype demonstrate swing-up recovery from 8^∘ within 7 s, stable balance under a 2 kg payload (maximum tilt deviation 0.17^∘), convergence to ±0.2^∘ after lateral disturbances, and curved motion at a 15^∘ steering angle with tilt within ±0.6^∘.