Authors:
Assia Idrissi
1
;
Arnaud Malapert
2
and
Rémi Jolin
3
Affiliations:
1
Milanamos, 1047 route des Dolines, Sophia Antipolis, France, Université Côte d’Azur, CNRS, I3S and France
;
2
Université Côte d’Azur, CNRS, I3S and France
;
3
Milanamos, 1047 route des Dolines, Sophia Antipolis and France
Keyword(s):
Flight Radius Problem, Time-independent Model, Graph Database, Shortest Path Algorithms.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Artificial Intelligence
;
Industrial Engineering
;
Knowledge Discovery and Information Retrieval
;
Knowledge-Based Systems
;
Methodologies and Technologies
;
Network Optimization
;
Operational Research
;
Optimization
;
OR in Transportation
;
Pattern Recognition
;
Software Engineering
;
Symbolic Systems
Abstract:
In this article, we present the flight radius problem (FRP) on the condensed flight network (CFN). Then, giving a specific flight that is defined by an origin and destination (OD) pair, the problem consists in finding routes that connect the OD pair and satisfy a regret constraint on time, distance or cost. The found routes help airline manager to find business opportunities. This problem arises in the real world, for instance in some air transportation companies. The FRP is formulated as finding a maximal subgraph of nodes belonging to routes satisfying a regret constraint. Such routes can be found using shortest paths algorithms (SPA). The CFN is generated using a time-independent approach and stored in the graph database Neo4j. Implementing SPA in Neo4j is challenging since the graph database stores the weights of the graph in a separate data structure. In this paper, we propose four methods to solve the FRP: these methods combine parallel and sequential processing with more optim
ization to overcome time and memory costs. The experimental evaluation demonstrates that the best algorithm is the extended Dijkstra algorithm which meets the real-time constraints of the targeted industrial application.
(More)