Abstract:
|
High Performance Robotic Computing (HPRC) is a new computer science field, which aims at exploiting traditional High Performance Computing (HPC) features, strategies and technologies in the context of multi-robot settings or more generally in vehicles capable of embedding computing units and able to use them to integrate with other similar vehicles as a single cohesive unit. In this work, we present the novel idea of HPRC cooperative flying, aircraft that while flying, join to do so cooperatively, during segments of their trajectories by becoming a HPRC cluster of aircraft. Furthermore, we introduce a simulation platform for HPRC cooperative flights which allows evaluating joining strategies, performance indicators such as fuel consumption, etc. The platform is written using parallel computing libraries, which allows it to run upon HPC clusters of computers and to be used to evaluate any quantity of flights. As a proof of concept, we present results of a simulation composed of up to 256 flights flying to Paris Charles de Gaulle Airport. In addition, we introduce a novel performance indicator based on the ideas of network science, specifically Global Reaching
Centrality (GRC), which we define as Aircraft Network hierarchy (ANH). |