Plastic grabber: Underwater autonomous vehicle simulation for plastic objects retrieval using genetic programming
Paper i proceeding, 2019

We propose a path planning solution using genetic programming for an autonomous underwater vehicle. Developed in ROS Simulator that is able to roam in an environment, identify a plastic object, such as bottles, grab it and retrieve it to the home base. This involves the use of a multi-objective fitness function as well as reinforcement learning, both required for the genetic programming to assess the model’s behaviour. The fitness function includes not only the objective of grabbing the object but also the efficient use of stored energy. Sensors used by the robot include a depth image camera, claw and range sensors that are all simulated in ROS.

Genetic programming

Underwater autonomous vehicle

Plastic collector

Författare

Gabriele Kasparaviciute

Chalmers, Teknikens ekonomi och organisation, Entrepreneurship and Strategy

Stig Anton Nielsen

Köpenhamns universitet

Dhruv Boruah

thethamesproject.org

Alexandru Dancu

Massachusetts Institute of Technology (MIT)

Lecture Notes in Business Information Processing

1865-1348 (ISSN) 18651356 (eISSN)

Vol. 339 527-533
978-303004848-8 (ISBN)

21st International Conference on Business Information Systems, BIS 2018
Berlin, Germany,

Ämneskategorier

Robotteknik och automation

Datavetenskap (datalogi)

Datorseende och robotik (autonoma system)

DOI

10.1007/978-3-030-04849-5_46

Mer information

Senast uppdaterat

2021-11-24