An MA can solve problems such as slow
convergence, low precision, and the need to avoid local
optima when solving high-dimension optimization
problems with a PSO. Our approach also offers a way
to overcome the sensitivity of a SA to initial values and
parameters.