Pink Fairy Armadillo Optimization (PFAO): A Novel Bio-inspired Metaheuristic for Solving Optimization Problems

This study proposes a new metaheuristic optimization technique named Pink Fairy Armadillo Optimization (PFAO).The Pink Fairy Armadillo (PFA) Optimization Algorithm is inspired by the unique foraging and digging patterns of the Pink Fairy Armadillo (PFA) in its original environment.PFAO is a novel metaheuristic approach that depends on the two complementary strategies of the Pink Fairy Armadillo (PFA) as a basis.The first strategy consists of quick and deeply digging, which allows for a wide breadth of search-space and the possibility of identifying new promising areas (global exploration).The second strategy is slow and controlled digging, which allows for localized exploitation (granular control) and a stepwise approach toward optimal (gradual) solutions.These strategies (two forms of digging) map onto the two primary phases of the proposed algorithm, exploration, and exploitation.The primary motivation of this study is to evaluate the effectiveness of PFAO in optimizing the trade-off between exploration and exploitation.During the optimization process, a series of detailed metrics were performed to assess the diversity of the population and the patterns of variation in response to the search.The results of this study demonstrated a complete and distinct change in behavioural patterns over a series of iterations.In the initial phases, the population exhibited vast, arbitrary, and stochastic patterns of movement that allowed for the population to explore all quadrants of the search space for potential optimal solutions.Enhanced convergence stability was demonstrated as the pattern of population movement focused on optimal solutions.This adaptive mechanism has clearly shown the ability to maintain equilibrium among global and local searches, and at the same time minimize the chances of premature convergence to the optimum local solution.PFAO is benchmarked against 23 standard functions, a mix of unimodal and multimodal problems, and compared against 9 prominent metaheuristic techniques.PFAO has been shown to perform the best compared to all others on 21 of the 23 functions (91.3%).PFAO has been shown to perform best on unimodal problems and converge significantly faster than all compared.In multimodal, PFAO promotes solution diversity among the population, adjusts its mechanisms to facilitate out of {local} optimum solutions and provide a consistent high quality to solutions.PFAO provides a simple to understand bio inspired and optimization mechanism which is a great alternative to classical or other modern metaheuristic techniques.

Hali tarjima qilinmagan