Login / Signup

Beam Search Algorithm for Anti-Collision Trajectory Planning for Many-to-Many Encounter Situations with Autonomous Surface Vehicles.

Jolanta KoszelewJoanna Karbowska-ChilinskaKrzysztof OstrowskiPiotr KuczyńskiEric KulbiejPiotr Wołejsza
Published in: Sensors (Basel, Switzerland) (2020)
A single anti-collision trajectory generation problem for an "own" vessel only is significantly different from the challenge of generating a whole set of safe trajectories for multi-surface vehicle encounter situations in the open sea. Effective solutions for such problems are needed these days, as we are entering the era of autonomous ships. The article specifies the problem of anti-collision trajectory planning in many-to-many encounter situations. The proposed original multi-surface vehicle beam search algorithm (MBSA), based on the beam search strategy, solves the problem. The general idea of the MBSA involves the application of a solution for one-to-many encounter situations (using the beam search algorithm, BSA), which was tested on real automated radar plotting aid (ARPA) and automatic identification system (AIS) data. The test results for the MBSA were from simulated data, which are discussed in the final part. The article specifies the problem of anti-collision trajectory planning in many-to-many encounter situations involving moving autonomous surface vehicles, excluding Collision Regulations (COLREGs) and vehicle dynamics.
Keyphrases
  • deep learning
  • machine learning
  • depressive symptoms
  • neural network
  • artificial intelligence
  • high throughput
  • minimally invasive
  • monte carlo