III. GENETIC ALGORITHMS
Genetic Algorithms is a fast growing optimization technique that imitates the natural selection process and survival of the fittest for finding optimum values of nonlinear multidimensional functions. A predefined objective function is optimized to find its global and/or local maxima and minima. The basic and general steps for genetic algorithm are:
1. Generalization of a random population (variables to be optimized)
2. Calculating the fitness of that population (substituting the variables in the objective function)
3. Selection of population members to proceed and cancelation of population members who did not pass the selection process
4. Cross-Over producing offspring of the past members
5. Mutation (changing the offspring’s properties)
6. Back to point no. 2 and repeat 2-5 until the best fit of variables is found.
Flow charts for a general genetic algorithm and for application for antenna design is shown in figure 2 and 3 respectively.
Based on flow chart figure 3 and formulas for rectangular microstrip antenna using transmission line model, a demonstrative genetic algorithm code was written and used to synthesis rectangular microstrip antennas.
The code is shown in the appendix of this report. The program asks the used to enter the desired frequency, substrate height and substrate dielectric constant, and then it calculates L and W for the user.