Pilot Research Projects

1) Light-Tracking Drone Fleet

This is a rather academic problem aimed at testing algorithmic developments for fully autonomous swarm-intelligent drone fleets.
Drone’s self-organisation is based on a memoryless particle swarm algorithm.
Red dot: Drone holding the location of the lightest spot currently found.

Lead Researcher: Dr Mauro S. Innocente
Contributing Student: Luke J. Smith (MSc student, 2019)

Based on Nagel and Schreckenberg’s model (NaSch), with additional features incorporated.
1D model. No junctions, two lanes, one traffic light.

A two-lane road is simply two single lane roads which have an interaction before the update procedure.
The data for each road is passed through a set of lane changing conditions and vehicles swap lanes if they are satisfied.
Once all checks are made regarding lane changes, each lane is then updated independently.
Parameter b_n stands for distance from vehicle to traffic light.
Every car will always try to have v_n = v_max:
There is a random probability (P_slow) that a vehicles’ speed will be decreased by one unit.
Each vehicle is moved forward by v_n cells.
Circular Boundary Conditions.
Probability (P_new) that a new car will appear in first cell.
P_new can then be used to control density of traffic in simulation.