THE EFFECT OF MOTION FORMATION ON COOPERATIVE NAVIGATION

Abstract

The effect of formation movement on the performance of cooperative navigation is investigated in this paper. First, the inertial navigation system of each agent with a certain accuracy is modeled and simulated. Initial results showed that the navigation error of each agent increased individually over time, and this problem is more severe for agents equipped with a weaker system. Cooperative navigation is implemented for the agents to resolve this problem. It is shown that the total navigation errors are improved by observing and participating the relative distance between the agents. Various simulations and experimental tests using two real agents supported this assertation. The performance of cooperative navigation can be improved further through appropriate formation. Proper formations are investigated and evaluated through simulations. The collective covariance matrix is employed to form an objective function using an extended Kalman filter (EKF). This function has been minimized using Newton’s method, which could be the solution for the formation. The simulation results show that better accuracy can be achieved by applying the optimal formation trajectory.