MARS: A simulation environment for marine swarm robotics and environmental monitoring

Abstract

In this paper, we present a simulation environment for marine robotics with support for swarm robotics and environmental monitoring. A vegetation system and fish swarms in the environment model present indirect bio-feedback. This enables the evaluation of swarm behaviors, e.g. to efficiently cover a vegetation area or monitor and chase fish swarms. The temperature sensor in the sensor model has a thermocline-based function for realistic data and a filter for the camera visually simulates suspended particles. The thruster model allows the use of a steady-state model function to calculate the thrust. For swarms, the underwater communication model is based on multipath propagation and background noise. The energy system supports solar panels that produce energy depending on the sun's position. Finally, a comparison between the depth control of real and simulated autonomous underwater vehicles was carried out.