CDS Lecture Series

ABSTRACT

Andrew J. Newman
The Johns Hopkins University Applied Physics Laboratory

UAV Motion Planning and Resource Management of Heterogeneous Platform and Sensor Ensembles

Ground targets performing unpredictable maneuvers and operating in cluttered environments such as urban areas pose severe challenges to current surveillance systems. It is often necessary to detect, track, and classify multiple targets distributed over broad areas in the face of difficulties such as varying backgrounds, low target observables, occlusions, dense and intersecting traffic, stop-and-go cycles, on-off road transitions, and intentional target countermeasures. The rapidly increasing number, diversity, agility, and capability of surveillance platforms and sensors present an opportunity to deal with the problem. New techniques and algorithms are required to enable effective dynamic collaborative resource management of heterogeneous platform and sensor ensembles that scale to typical problem sizes of interest. This talk will focus on the problem of non-myopic multiple platform trajectory control in a multiple target search and track setting. It will present a centralized receding discrete time horizon controller (RHC) with variable-step look-ahead for motion planning of a heterogeneous ensemble of airborne sensor platforms. The controller operates in a closed feedback loop with a Bayesian Multiple Hypothesis Tracker (MHT) that fuses the disparate sensor data to produce target declarations and state estimates. The RHC action space for each air vehicle is represented via maneuver automaton with simple motion primitives. The reward function is based on expected Fisher information gain and priority scaling of target tracks and ground regions. A customized Particle Swarm Optimizer (PSO) is developed to handle the resulting non-Markovian, time-varying, multi-modal, and discontinuous reward function. The algorithms were evaluated by simulating ground surveillance scenarios using representative platforms and sensors with varying fields of view, typical target densities and motion profiles, and environmental characteristics such as varying backgrounds and occlusions. Simulation results show improved aggregate target detection, track accuracy, and track maintenance for closed loop operation as compared with typical open-loop surveillance plans. This talk will also introduce recent applications of these techniques and algorithms to the space surveillance domain. As our reliance on space systems and space technology grows, the consequences of collisions and other spacecraft interactions are becoming increasingly serious. Even relatively small objects can severely damage or destroy a satellite. It is essential to maintain an accurate catalog of most, if not all, objects in orbit. New capabilities, including automated data fusion and resource management of heterogeneous sensor ensembles, are required to accurately detect, track, and classify small objects in earth orbit at a variety of altitude regimes.


Back to CDS Lecture Series
Back to Intelligent Servosystems Laboratory