Presentation Title

Predicting Coordinated Group Movements of Sharks with Limited Observations using AUVs

Start Date

November 2016

End Date

November 2016

Location

HUB 379

Type of Presentation

Oral Talk

Abstract

This presentation offers a method for modeling and tracking the size, location, orientation, and number of individuals in an animal aggregation using Autonomous Underwater Vehicles (AUVs). The AUVs are equipped with sensors that can measure the position states of a subset of individuals from within the aggregation being tracked. A new aggregation model based on provably stable Markov Process Matrices is shown as a viable model for representing aggregations. Then, a multi-stage state estimation architecture based on Particle Filters is presented that can estimate the time-varying model parameters in real-time using sensor measurements obtained by AUVs. The method is generalizable to any stable aggregation movement model constructed using a Markov Matrix. A historical data set of more than 100 trajectories from a leopard shark aggregation is used to validate the approach. Simulations show that the estimated number of sharks in the aggregation has an error of 6% when at least 40% of sharks are tagged. When using historical data, this error increases to 27%.

Keywords: Multi-Target Tracking, Aggregation Modeling, Swarms, Particle Filter

This document is currently not available here.

Share

COinS
 
Nov 12th, 3:30 PM Nov 12th, 3:45 PM

Predicting Coordinated Group Movements of Sharks with Limited Observations using AUVs

HUB 379

This presentation offers a method for modeling and tracking the size, location, orientation, and number of individuals in an animal aggregation using Autonomous Underwater Vehicles (AUVs). The AUVs are equipped with sensors that can measure the position states of a subset of individuals from within the aggregation being tracked. A new aggregation model based on provably stable Markov Process Matrices is shown as a viable model for representing aggregations. Then, a multi-stage state estimation architecture based on Particle Filters is presented that can estimate the time-varying model parameters in real-time using sensor measurements obtained by AUVs. The method is generalizable to any stable aggregation movement model constructed using a Markov Matrix. A historical data set of more than 100 trajectories from a leopard shark aggregation is used to validate the approach. Simulations show that the estimated number of sharks in the aggregation has an error of 6% when at least 40% of sharks are tagged. When using historical data, this error increases to 27%.

Keywords: Multi-Target Tracking, Aggregation Modeling, Swarms, Particle Filter