Self-Adaptive Forensic-Based Investigation Algorithm with Dynamic Population for Solving Constraint Optimization Problems

Abstract

The Forensic-Based Investigation (FBI) algorithm is a novel metaheuristic algorithm. Many researches have shown that FBI is a promising algorithm due to two specific population types. However, there is no sufficient information exchange between these two population types in the original FBI algorithm. Therefore, FBI suffers from many problems. This paper incorporates a novel self-adaptive population control strategy into FBI algorithm to adjust parameters based on the fitness transformation from the previous iteration, named SaFBI. In addition to the self-adaptive mechanism, our proposed SaFBI refers to a novel updating operator to further improve the robustness and effectiveness of the algorithm. To prove the availability of the proposed algorithm, we select 51 CEC benchmark functions and two well-known engineering problems to verify the performance of SaFBI. Experimental and statistical results manifest that the proposed SaFBI algorithm performs superiorly compared to some state-of-the-art algorithms.