A framework based on a mixed reality simulator for coordinating teams of autonomous Unmanned Aerial Vehicles (UAVs) is been developed. This framework would serve as a tool to facilitate crossing the reality gap for different applications; particularly when using these UAVs teams for air pollution monitoring and measurement. The system is built on a co-evolutionary simulator that makes use of data transmitted from some real UAVs to integrate them within a team of simulated UAVs. The system allows the progressive increase of the number of real UAV in the team. This facilitates the setting-up of a single UAV control system and also of the UAV collaboration schemes for different scenarios. A specific implementation of this system focussed on mapping the pollutant dispersion of a plume in the atmosphere is presented. Implementing an appropriate pollution dispersion model within the simulator is a key aspect of the system. This model should require few computational resources, should be easy to adapt in real time to ambient changes, and it should have a fair accuracy.
[1]
Richard J. Duro,et al.
Autonomous UAV based search operations using Constrained Sampling Evolutionary Algorithms
,
2014,
Neurocomputing.
[2]
Rainer Storn,et al.
Differential Evolution – A Simple and Efficient Heuristic for global Optimization over Continuous Spaces
,
1997,
J. Glob. Optim..
[3]
A. R. Perry,et al.
The FlightGear flight simulator
,
2004
.
[4]
Richard J. Duro,et al.
Differential Evolution in Constrained Sampling Problems
,
2014,
2014 IEEE Congress on Evolutionary Computation (CEC).
[5]
Richard J. Duro,et al.
Evolutionary algorithm characterization in real parameter optimization problems
,
2013,
Appl. Soft Comput..
[6]
Phil Husbands,et al.
Evolutionary robotics
,
2014,
Evolutionary Intelligence.
[7]
Inman Harvey,et al.
Noise and the Reality Gap: The Use of Simulation in Evolutionary Robotics
,
1995,
ECAL.