State estimation for a bio-inspired hovering robot equipped with an angular sensor

The present paper describes a methodology to estimate the attitude and the position of a bio-inspired robot as well the position of a target. The robot is equipped with a decoupled eye yielding an angular measurement relative to a target and it does not use an IMU. An Extended Kalman Filter is designed to estimate robot and target states. Simulation results show that the estimator is able to converge to the actual states, and reject some perturbations. It is important to highlight that the method assumes that the initial position of the target is unknown. As a consequence no initial guess for the target position is needed.