Hybrid aerial–ground robots have recently attracted increasing attention for their versatile mobility and adaptability across complex environments. This paper presents a bias-aware estimation and control framework for a miniaturized dual-tilt aerial–ground robot. A complete nonlinear dynamic model is first established, and a bias-linearized allocation structure is derived to reveal the effect of tilt-axis misalignment on the generalized force mapping. Based on this model, servo installation bias is modeled as a matched disturbance in the control allocation. It is estimated by a fixed-time observer and compensated using an integral terminal slidingmode controller (ITSMC). The controller ensures boundedness and finite-time convergence during estimator transients and under actuator saturation. Simulations based on the full nonlinear model and flight experiments on a 0.86 kg prototype verify the proposed method. Results show that even small installation offsets (on the order of 1°) can cause substantial yaw drift if uncorrected. The proposed observer–controller combination achieves rapid bias compensation and accurate attitude tracking under realistic noise and actuation limits. The overall findings confirm that bias-aware modeling and control are essential for achieving robust and reliable performance in lightweight hybrid aerial–ground systems.
Zhou et al. (Thu,) studied this question.