Summary: October 10, 2008
To Whom It May Concern,
This letter is to inform you of the progress of the dynamic path planning project
that was proposed September 12th
. Much time has been spent studying the already
existing documentation and code and it has been decided that we will focus on
implementing 2 improvements to the system's fitness functions, as well as a Particle
Swarm Optimization (PSO) path planning method for comparison purposes.
The first fitness function improvement is a radial vision system that works to
improve the already existing collision avoidance system by considering collisions for all
paths (as opposed to just the current path) but placing less importance on potential
collisions further down each path, and greater importance on imminent collisions.
The second fitness function improvement considers the timing aspects of a potential
collision by projecting the area of an obstacle and studying the generated paths that cross
through the potential collision zone to see if a collision will occur based on the velocities
and the entry & exit points of the path.
Finally, based on research, a PSO algorithm could potentially require less
computational complexity and therefore in implementation could increase speed and
reduce power consumption. Therefore we will implement a PSO replacement for the
Genetic Algorithm and study the performance of both algorithms to compare their