Study of biology inspired global optimization techniques

Biology inspired global optimization techniques have been widely studied and are promising in many engineering design optimizations. These bio-inspired algorithms typically fall in one of three main categories: evolutionary algorithms (EA), swarm-based algorithms, and ecological algorithms. However, due to their closely related operation they sometimes belong into more than one category. In this report, three promising EAs, revealed by a literature review and computer programming testing via MATLAB, were studied. The first is the differential evolution (DE) algorithm which is based on the natural selection of biological species. Next, the biogeography-based optimization (BBO) algorithm which is based on the natural distribution of species in a habitat. Last, the teacher-learner-based optimization (TLBO) algorithm which is based on the teaching and learning phenomenon which takes place in a classroom. Five established global optimization functions from current literature were used to test and ensure the proper settings for each algorithm. The results from those five mathematical functions were evaluated and compared, revealing strong performances from the algorithms. As a result, an engineering design optimization in electromagnetics was tested using the algorithms. A published design optimization involving an antenna array power pattern synthesis using the genetic algorithm (GA), the first established EA, was optimized using the three algorithms. The results were compared and the three algorithms in the study outperformed the originally used GA. Hence, this study provides support for the need to further expand these algorithms and their application to engineering design optimization problems.