A Local Charged Particle Swarm Optimization to track an underwater mobile source
| Autor: | Charles Coquet, Andreas Arnold, Pierre-Jean Bouvet, Clement Aubry |
|---|---|
| Přispěvatelé: | Laboratoire ISEN (L@BISEN), Institut supérieur de l'électronique et du numérique (ISEN)-YNCREA OUEST (YO), Thales DMS France, SAS, Arnold, Andreas |
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
0209 industrial biotechnology
Flocking (behavior) Computer science [INFO.INFO-RB] Computer Science [cs]/Robotics [cs.RO] Real-time computing Particle swarm optimization Swarm behaviour Context (language use) 02 engineering and technology Function (mathematics) Robotics Tracking (particle physics) Computer Science::Robotics 020901 industrial engineering & automation [INFO.INFO-AU]Computer Science [cs]/Automatic Control Engineering 0202 electrical engineering electronic engineering information engineering Swarm [INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO] 020201 artificial intelligence & image processing Underwater Source tracking [INFO.INFO-AU] Computer Science [cs]/Automatic Control Engineering Index Terms-PSO |
| Zdroj: | MTS/IEEE Oceans MTS/IEEE Oceans, 2019, Marseille, France HAL |
| Popis: | International audience; In this paper, a possible solution to track a mobile underwater source in a closed environment with N Autonomous Underwater Vehicles (AUV) in a swarm formation is adressed. The source tracking algorithm is defined as successful when the range between the source and the swarm is sufficiently low during a given duration, short enough to perform a specified action (for example a source localization). A source is defined as an entity that releases a scalar information affected by transport and diffusion in the environment. We use a generic time-varying information f (pi(t)), where pi at time t is the m-dimensional position of a tracker i and function f (.) is a function that represents sensor information. In this paper, we propose an innovative tracking method inspired by the Particle Swarm Optimization (PSO) algorithm that we call the Local Charged Particle Swarm Optimization (LCPSO). The proposed algorithm is adapted to range-dependant communication that characterizes the underwater context and includes flocking parameters. Comparison of the LCPSO against state of the art methods demonstrate the interest of our approach in an underwater scenario. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|