Authors:
Christoph Endres
and
Christian Müller
Affiliation:
DFKI GmbH, Germany
Keyword(s):
Automotive, Presentation scheduling, Conflict resolution.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Formal Methods
;
Industrial Applications of AI
;
Informatics in Control, Automation and Robotics
;
Intelligent Control Systems and Optimization
;
Planning and Scheduling
;
Simulation and Modeling
;
Soft Computing
;
Symbolic Systems
Abstract:
In this paper we present a problem which is a variation of the resource-constrained project scheduling problem
and a graph-based approach to solve it. The problem is defined as resource-constrained scheduling problem
(RCSP). Particularly, we apply the approach to the problem of scheduling a large number of driver warnings
based on car-to-car communication (also known as cooperative vehicles). Data is presented from the project
SIMTD , a large-scale field test in the area of the Hessian city of Frankfurt, where 120 cars participate in a
number of controlled tests in three main scenarios: the rural road scenario (basic complexity), the motorway
scenario (intermediate complexity), and the urban road scenario (high complexity). We argue that, due to its
run-time behaviour, our graph-based approach is suitable for the particular application domain at hand. Results
are presented in terms of quality of the solution (conflict resolution), runtime behavior and pruning effects to
the size of t
he search tree. In addition to the scenarios derived from the actual field test, a hyper-real stress test
is presented to demonstrate the performance of our solution.
(More)