Gentic Algorithm for curve detection
The original problem was to solve the Brachistochrone problem. The
problem is defined as such. We want to move some ball typically from
the origin (0, 0) to (x, -y), with x >= 0, y >= 0. But we want to
detect the shape of the curve which will minimise the transit time of
the ball. This genetic algorithm is a generalisation to a whole class
of problems where the shape of some curve typically need to be
detected. The code given below can be driven by a user defined
objective function. The objective function to solve the Brachistchrone
is also given in the code. The crossover, mutation, and population
merging operators given are worth investigating themselves. These are
discussed in the report below.
Back | Home