Higher Order Basis Based Integral Equation Solver (HOBBIES)

10

HOBBIES OPTIMIZER AND ITS APPLICATIONS

10.0 SUMMARY

After simulating the project and obtaining the initial results, the Higher Order Basis Based Integral Equation Solver (HOBBIES) Optimizer can be used to adjust automatically the designated model parameters such as model-element coordinates, object length, and similar quantities to achieve improved results for the desired characteristics such as gain or sidelobe performance. The HOBBIES Optimizer settings include choosing optimization methods, iteration numbers and symbols, and setting criteria. There are three optimization algorithms for the user to choose from. They are Powell's method, the Simplex (Nelder–Mead) method, and the Particle swarm optimization method. The number of iteration limits the maximum number of simulation calls for each specific algorithm. Symbols are defined by the user to change the model parameters, and the user needs to define the upper and lower boundaries for each of the symbols to be optimized. Several criteria are defined for the optimization goals. Each criterion can be assigned a specific weight to differentiate its priority in the whole list of criteria to be optimized.

The optimization process is controlled by a fitness function, which provides a fitness value based on the difference between the calculated result and the desired performance. The Optimizer attempts to find the minimum value of this function. Optimization is an iterative process, in which the Optimizer calculates the fitness ...