The four-robot configuration seems to be the optimal situation with a shadow area of
zero. The five-robot configuration results in a shadow area of zero also, which is an
intuitive result. As more robots are added to the solution, the better the coverage of the
search space. The algorithm converged after cycling through 100 generations of the
genetic algorithm. Plots of the shadow areas versus time are shown in figure 5.5.
- 2 Robots
3 Robots
4 Robots
- 5 Robots
0 2 4 6 8 10 12 14
Time (sec)
Figure 5.5: Map 2 convergence rates for 2-5 robots after 100 generations
The five-robot configuration converged in 15 seconds and the solution of four robots
converged in 10 seconds. A recurring theme in this work is that the genetic algorithm
will provide 'near-perfect' results. That is, the optimal solution will not always be
obtained but something close to optimal will be obtained. The final solution will be
dependent on how long the algorithm runs for and ultimately the number of generations
that are iterated through to get a final solution.