Go up to Top
Go forward to 2 Searching on a simple graph
1 Finding Paths in a Grid
Consider the problem of finding a path in the grid shown below from
the position s to the position g. The robot can move on the grid
horizontally and vertically, one square at a time (each step has a
cost of one). No step may be made into a forbidden shaded area.
-
On the grid, number the nodes in the order in which they are removed
from the frontier in a depth-first search from s to g, given that
the order of the operators you will test is: up, left, right, then
down. Assume there is a cycle check.
- Number the nodes in order in which they are taken off the
frontier for an A* search for the same graph. Manhattan distance
should be used as the heuristic
function. That is, h(n) for any node n is the Manhattan distance
from n to g. The Manhattan distance between two points is the
distance in the x-direction plus the distance in the y-direction.
It corresponds to the distance traveled along city streets
arranged in a grid. For example, the Manhattan distance between g
and s is 4.
What is the path that is found by the A* search?
- Assume that you were to solve the same problem using dynamic
programming. Give the dist value for each node, and show which path
is found.
-
Based on this experience, discuss which algorithms are best suited for this
problem.
- Suppose that the graph extended infinitely in all directions.
That is, there is no boundary, but s, g, and the forbidden area
are in the same relative positions to each other. Which methods would
no longer find a path? Which would be the best method, and why?
Computational
Intelligence online
material, ©David Poole, Alan Mackworth and Randy Goebel, 1999