Conditional toggle algorithm: an adaptive metaheuristic and its implementation on handling engineering problems

Abstract

There have been numerous new metaheuristic algorithms in this decade. Unfortunately, the attention on taking stagnation is still less considered so that it is difficult to find new metaheuristic algorithms that are enriched with stagnation taking mechanism. This work introduces a new method called conditional toggle algorithm (CTA). CTA is designed to be adaptive on facing enhancement and stagnation during iteration as its novelty. When enhancement occurs, the exploitation-focused look is applied. Meanwhile, the exploration-focused look is applied when stagnation occurs. The efficacy of CTA is then measured by implementing to solve three cases: 23 functions, 4 engineering design problems, and economic emission dispatch (EED) problem in Java-Bali power system in Indonesia. CTA is compared with five new metaheuristic algorithms. The evidence provides that CTA is supreme in taking high dimension functions and competing in taking fixed dimension functions. CTA is also supreme in taking pressure vessel and speed reducer design problems and the EED problem. But its performance is average in taking welded beam and spring design problems. In the future, CTA can be modified with other metaheuristic algorithms to enhance its performance and challenged to take broader problems, especially in electrical engineering fields.