Target localization and multicircular circumnavigation with bearing‐only measurements

Abstract

This article studies the multicircular circumnavigation problem of a stationary target using bearing-only measurements. The proposed estimator-based control strategy drives the networked mobile agents to move around an unknown target along different circles with any prescribed inter-agent angular spacing. The results presented in this article are as follows: First, a new estimator is developed by introducing a linear filter into an estimation algorithm which has been widely used in the existing literature. The novelty of the proposed estimator lies in its enhanced convergence rate. Analytical results show that the exponential convergence rate of the estimator can be adjusted to be arbitrarily high by increasing the estimation gain. Second, a distributed controller is designed such that global exponential convergence of the agents to the desired circular formation can be achieved. Besides, the proposed algorithms guarantee that each agent can keep a safe distance from the target, and high angular velocities of the agents can be avoided even if the circumnavigation radii become small. Finally, several numerical simulations are given to corroborate the theoretical results.

Heterogeneous Formation Control: a Bearing Rigidity Approach

“…Since any v i in ( 3) is modeled as a normally distributed Gaussian random vector, we suppose that also n S in (7) approximates a Gaussian random vector with n i.i.d. components.…”

“…Indeed, in the former, a Maximum Likelihood approach is exploited to face the non-linearities affecting the measurement noise, while in the latter the task is formalized and addressed in a constrained geometric optimization framework resting on the relative direction of the agents' bearings. More recently, similar measurement-aware methods have been proposed for the localization of multiple static targets by means of a multi-agent formation in the context of obstacle avoidance task [5], and for the localization of a single dynamic target moving on a plane and tracked by a group of seeker agents [6], [7]. In these works, the seeker agents are steered along continuous predetermined (elliptical and circular, respectively) trajectories, independently of the target localization uncertainty.…”