Disturbance Observer-Based Attitude Control

The most existing DO requires external disturbance or modeling error to satisfy some strict conditions. For instance, most ESO are only feasible for the unknown constant disturbance \(\boldsymbol{u}_d\) or the disturbance with slow variation [1–3], i.e., \(\dot{\boldsymbol{u}}_d=\vec{0}\) , \(\dot{\boldsymbol{u}}_d\approx \vec{0}\) , \(\lim \limits _{t \rightarrow \infty }\dot{\boldsymbol{u}}_d=\boldsymbol{0}\) , or \(||\dot{\boldsymbol{u}}_d||\le {{\delta }_1}\) , where \({\delta }_1\) is positive and small scalar.