Page 50 - ASME DSCC 2015 Program

Technical Program

Design of Mobile Microgrid’s Hierarchy for Power Distribution
ConTRIBuTED SESSIon
1-14-2 TM2 Multiagent network Systems 2 Contributed regular paper. DSCC2015-9866
George Bellows B 1:30pm–3:30pm Barzin Moridian, Daryl Bennett, nina Mahmoudian, Rush D. Robinett III,
Wayne W. Weaver, Michigan Technological University, Houghton, MI, United
Session Chair: xinmin Liu, University of Chicago States
Session Co-Chair: Jingang Yi, Rutgers University This paper presents a hierarchy for autonomous mobile microgrids to
ensure robustness and scalability of such systems for power distribution
Distributed navigation Strategy of Mobile Sensor networks With
applications. Developing an autonomous mobile microgrid based on the
Probabilistic Wireless Communication links
proposed hierarchy will create a microgrid system that will autonomously
Contributed regular paper. DSCC2015-9964
connect to different power nodes (generators, loads, storage units, con-
aqeel Madhag, jongeun choi, Michigan State University, East Lansing, MI, verters, etc.) and interact with them accordingly. This system will have the
United States scalability characteristics of an ad hoc system and can reconfigure itself
Mobile sensor networks have been widely used to predict spatio-temporal depending on available power nodes.
physical phenomena for various scientific and engineering applications. To State Emulator-Based Adaptive Architectures for Resilient networked
accommodate the realistic models of mobile sensor networks, we incorpo- Multiagent Systems over Directed and Time-Varying Graphs
rated probabilistic wire-less communication links based on packet reception Contributed regular paper. DSCC2015-9802
ratio (PRR) with distributed navigation. We then derived models of mobile
Gerardo De la Torre, Georgia Institute of Technology, Atlanta, GA, United
sensor networks that predict Gaussian random fields from noise-corrupt-
States, Tansel Yucelen, Missouri University of Science and Technology,
ed observations under probabilistic wireless communication links. For the
Rolla, MO, United States
given model with probabilistic wireless communication links, we derived
In this paper, we present adaptive architectures for networked multiagent
the prediction error variances for further sampling locations. Moreover, we
systems operating over directed networks to achieve resilient coordination
designed distributed navigation that minimizes the network cost function
in the presence of disturbances. Specifically, we consider a class of unfore-
formulated in terms of the derived prediction error variances. Further, we
seen adverse conditions consisting of persistent exogenous disturbances
have shown that the solution of distributed navigation with probabilistic
and present a state emulator-based distributed adaptive control architecture
wireless communication links for a mobile sensor net-work is uniformly
to retrieve the nominal networked multiagent system behavior. The stability
ultimately bounded with respect to that of the distributed one with the
properties of the proposed architecture are analyzed using results from
R-disk communication model. Ac-cording to Monte Carlo simulation results,
Lyapunov stability and matrix mathematics. Illustrative numerical examples
distributed navigation with probabilistic wireless communication links could
are provided to demonstrate the theoretical findings.
approximate distributed navigation with the R-disk communication model,
confirming theoretical analysis. Mode Controllability and Disturbance Rejection in Consensus
An Iterative learning Control Approach to Multi-Agent formations networks
Contributed short paper. DSCC2015-9872 Contributed regular paper. DSCC2015-9819
Rammah Abohtyra, Colorado School of Mines, Colorado School of Mines,
Michael Quann, University of Michigan, Moorestown, NJ, United States, Kira
Barton, University of Michigan, Ann Arbor, MI, United States CO, United States
This paper presents an iterative learning control (ILC) based method for This paper considers a new notion of mode controllability for interconnected
trading off both individual and formation tracking for multiple agents with systems. A new method based on canceling natural system modes that arise
heterogeneous dynamics. The proposed framework provides precise from non-zero initial conditions is proposed. A system of first-order linear
trajectory tracking for systems involving repetitive, cooperative motion. differential equations is used to represent an interconnected system where
The frequency domain based controller has the capability of being able the links are static gains. The paper focuses on the initial conditions of the
to shift weighting between the desire for precise formations amongst the nodes. The system is said to be mode controllable if every initial condition
agents and precise trajectory tracking for each individual agent. Stability and mode can be canceled, while the system is not mode controllable if there
convergence conditions are shown and the proposed controller is validated exists at least one initial condition mode that cannot be cancelled. A test
through simulations. of model controllability is developed, which is equivalent to the standard
controllability. We test this method by rejecting all natural system modes
using the minimum input control. Finally, the paper concludes an illustrative
example with simulation results and future work.

50