TY  - JOUR
PY  - 2010//
TI  - Optimization of Shared Autonomy Vehicle Control Architectures for Swarm Operations
JO  - IEEE transactions on systems, man, and cybernetics. Part B, cybernetics
A1  - Sengstacken, Aaron J.
A1  - Delaurentis, D. A.
A1  - Akbarzadeh-T, Mohammad R.
SP  - 1145
EP  - 1157
VL  - 40
IS  - 4
N2  - The need for greater capacity in automotive transportation (in the midst of constrained resources) and the convergence of key technologies from multiple domains may eventually produce the emergence of a "swarm" concept of operations. The swarm, which is a collection of vehicles traveling at high speeds and in close proximity, will require technology and management techniques to ensure safe, efficient, and reliable vehicle interactions. We propose a shared autonomy control approach, in which the strengths of both human drivers and machines are employed in concert for this management. Building from a fuzzy logic control implementation, optimal architectures for shared autonomy addressing differing classes of drivers (represented by the driver's response time) are developed through a genetic-algorithm-based search for preferred fuzzy rules. Additionally, a form of "phase transition" from a safe to an unsafe swarm architecture as the amount of sensor capability is varied uncovers key insights on the required technology to enable successful shared autonomy for swarm operations.<p />  <p>Language: en</p>
LA  - en
SN  - 1083-4419
UR  - http://dx.doi.org/10.1109/TSMCB.2009.2035099
ID  - ref1
ER  -