Modified SA Algorithm

(a) Choose an initial temperature T (β).

(b)P organisms or numbers in binary code like parents were generated randomly.

(c) Children were produced from all parents.

(d)Selection of best one in every family is completed by competition among children.

(e)The parent for next generation is acquired of each family.

Best child is adopted as like parent for next generation, if

                                              ΔME = ME₁  - ME₂

                                 If ΔME < 0   or T (β)/ T²(β) + ΔME²>β

Here, ME₁ = Magnetic potential energy of the best children,

And ME₂ = Magnetic potential energy of its parent,

T (β) = Temperature, and

β = Random number in between 0 and 1. Temperature is reduced as describe as:

                                                       T (β) = T (β) - 0.1 T (β)

If solution is frozen then stop

Temperature is initialized in the starting of the algorithm. Latest population of organism is generated according to the evolution rule as like selection, mutation, and crossover. Superior population is to be produced after the subsequent generation of organisms. Such variant of SA algorithm overcomes the demerits of generic SA in given two ways as:

(a)   Population size is being decreased.

(b)   Local minima can be escaped.