Chinese Pangolin Optimizer

This paper proposes a novel bio-inspired metaheuristic algorithm, the Chinese Pangolin Optimizer (CPO).

This paper proposes a novel bio-inspired metaheuristic algorithm, the Chinese Pangolin Optimizer (CPO), which draws inspiration from the unique hunting behaviors of Chinese pangolins, including luring and predation behaviors. We successfully simulated these behaviors through mathematical modeling and systematically analyzed the algorithm’s convergence using Markov chain theory, theoretically ensuring its optimization efficiency and reliability. The numerical optimization performance of CPO was extensively evaluated using 74 standard benchmark functions, encompassing unimodal, multimodal, fixed-dimensional multimodal functions, and the CEC2017, CEC2019, and CEC2022 test suites. Experimental results and statistical analyses provided strong evidence of CPO’s effectiveness. Furthermore, CPO demonstrated exceptional performance in high-dimensional complex search tasks, as evidenced by tests conducted on 30-dimensional, 50-dimensional, and 100-dimensional CEC2017 test suites. Three standard engineering design optimization problems and twelve feature selection tasks assessed CPO’s capability in solving complex real-world problems. Experimental results show that the proposed CPO algorithm outperforms numerous baseline metaheuristic algorithms regarding optimization performance and robustness. n feature selection tasks, CPO achieved higher average classification accuracy than seven competing metaheuristic algorithms and six mainstream feature selection methods, demonstrating its versatility and superiority in addressing complex practical problems.