Gilbert’s Potoroo Optimization: A Novel Nature-inspired Metaheuristic Algorithm for Solving Optimization Problems

In this paper, a novel swarm-based metaheuristic algorithm, termed Gilbert's Potoroo Optimization (GPO), is proposed and developed.The algorithm is inspired by the distinctive foraging strategies and soil-digging behaviours of the rare Gilbert's Potoroo, particularly its adaptive movements during food resource exploration and extraction.First, the theoretical foundations and conceptual framework of the GPO algorithm are rigorously established, followed by a detailed description of the population updating mechanism, which is structured into three distinct phases.The first phase corresponds to a local exploration mechanism, modeled based on the short-range movements of Gilbert's Potoroo across neighboring territories, which effectively enhances population diversity and promotes global search capability.The second phase represents a long-range exploration strategy, designed by simulating the animal's occasional long-distance movements toward remote regions; this phase plays a critical role in expanding the search space and mitigating premature convergence.Finally, the third phase is dedicated to exploitation, drawing inspiration from the targeted digging behavior of Gilbert's Potoroo during food extraction, thereby improving local search accuracy and strengthening the quality of the obtained optimal solutions.To assess the effectiveness of the proposed algorithm in real-world practical scenarios, the performance of GPO is evaluated on 22 constrained optimization problems from the well-established CEC 2011 benchmark suite, and its results are comparatively analyzed against those of nine well-known metaheuristic algorithms.The optimization results demonstrate that GPO, owing to its strong capability to maintain an effective balance between exploration and exploitation phases, consistently delivers highquality and competitive solutions across the tested problems.Comparative performance analysis further reveals that GPO achieves the first rank in 16 out of 22 test problems (approximately 73%), clearly outperforming the competing algorithms.These findings indicate that the GPO algorithm, due to its robust performance in solving complex optimization problems, can serve as an efficient and reliable optimization tool for a wide range of applications in science and engineering.