Real-Time Area Coverage and Target Localization Using Receding-Horizon Ergodic Exploration

Abstract

Abstract-Although a number of solutions exist for the problems of coverage, search and target localization-commonly addressed separately-whether there exists a unified strategy that addresses these objectives in a coherent manner without being application-specific remains a largely open research question. In this paper, we develop a receding-horizon ergodic control approach, based on hybrid systems theory, that has the potential to fill this gap. The nonlinear model predictive control algorithm plans real-time motions that optimally improve ergodicity with respect to a distribution defined by the expected information density across the sensing domain. We establish a theoretical framework for global stability guarantees with respect to a distribution. Moreover, the approach is distributable across multiple agents, so that each agent can independently compute its own control while sharing statistics of its coverage across a communication network. We demonstrate the method in both simulation and in experiment in the context of target localization, illustrating that the algorithm is independent of the number of targets being tracked and can be run in real-time on computationally limited hardware platforms.

Active Area Coverage from Equilibrium

“…Thus, (16) shows us that ∂ ∂λ J is always negative subject to the schedule of control vectors (13) and the objective is being reduced when µ (t) is applied.…”

Using Neural Networks to Generate Information Maps for Mobile Sensors

“…In some methods, an ergodic trajectory is generated by comparing c to φ using a metric like KL divergence [9]. However, it is more common to decompose c and φ into Fourier coefficients and modify the trajectory until the coefficients are roughly equal [10], [11], [12]. The information map φ is decomposed according to…”

“…Further, there are no calls to the log or measurement functions. The low complexity helps explain why Fisher information maps are common in prior work [6], [16], including a real-time implementation on a robot [12].…”

“…We also generate Fourier coefficients from these information maps. For the bearing modality, we use K = 5 as the highest order coefficient, in line with prior work [12]. In the FOV modality, we use K = 17, the smallest value that captured major features in observed information maps.…”

“…The third quantity, D(φ φ), compares the map generated from the true coefficients to the true information map. Fourier coefficients introduce band-limiting degradation but are still used to guide mobile sensors [6], [10], [11], [13], [12], so this last value is a useful reference of acceptable quality.…”