Improved dwarf mongoose optimization algorithm using novel nonlinear control and exploration strategies

Abstract

The Dwarf Mongoose Optimization Algorithm (DMO) is a popular metaheuristic algorithm utilized to solve real-world problems. However, DMO has limitations such as slow convergence rate, susceptibility to local optima, and poor performance in high-dimensional problems. Aiming at the defects of the DMO, we propose an improved version called IDMO. This paper introduces the best leader and proposes a novel nonlinear control strategy based on the sine function, significantly enhancing the convergence rate while ensuring accuracy. Moreover, we propose a novel exploration strategy to solve the global optimization and high-dimensional problem. To demonstrate the performance of the proposed IDMO, we test on 65 test functions, including CEC2017, CEC2020, CEC2022, and CEC2013 large-scale global optimization. IDMO is compared with three classes of widely recognized algorithms: (1) highly-cited algorithms, namely, GSA, GWO, WOA, and HHO, (2) advanced algorithms, such as CPSOGSA, CSA, AVOA, and SO, (3) high-performance optimizers including L-SHADE, AL-SHADE, LSHADE-SPACMA, and LSHADE-cnEpSin. Moreover, we also compare with the original DMO and variants such as GDNNMOA, CO-DWO, and BDMOSAO. Meanwhile, we apply IDMO to solve 19 engineering design problems from the CEC2020 real-world optimization suite. Experimental results demonstrate that IDMO outperforms other algorithms, exhibiting excellent convergence rate, stability, and accuracy. The effectiveness of IDMO is confirmed by the Friedman mean ranking, showcasing its potential for metaheuristic optimization and real-world engineering design problems.