Environmental boundary tracking and estimation using multiple autonomous vehicles

In this paper, we develop a framework for environmental boundary tracking and estimation by considering the boundary as a hidden Markov model (HMM) with separated observations collected from multiple sensing vehicles. For each vehicle, a tracking algorithm is developed based on Page's cumulative sum algorithm (CUSUM), a method for change-point detection, so that individual vehicles can autonomously track the boundary in a density field with measurement noise. Based on the data collected from sensing vehicles and prior knowledge of the dynamic model of boundary evolvement, we estimate the boundary by solving an optimization problem, in which prediction and current observation are considered in the cost function. Examples and simulation results are presented to verify the efficiency of this approach.