A mirage search optimization algorithm that combines dual-chaos mapping and an opposition-based learning strategy

In this study, we propose a mirage search optimization algorithm that combines dual-chaos mapping and an opposition-based learning tactic, which we call the dual-chaos mapping opposition-based learning mirage search optimization (DMOMSO) algorithm. First, when initializing the population, a chaotic initialization method was used. This method takes advantage of the chaotic characteristics of iterative mapping to generate an initial population with high dispersion. Thus, the algorithm can more reliably reach the global optimum in a shorter time. Before the formal iteration process starts, a chaotic sequence is generated via chaotic mapping, which is used to update the iterative parameters, enabling them to vary dynamically throughout the iterative process. This helps the algorithm avoid becoming stuck in local optima, as well as enhancing its optimization performance and accelerating its convergence. In each iteration, the parameters are adjusted using the opposition-based learning method, which helps to cover the solution space more effectively, find good solutions faster, and explore a wider area. We conducted simulation experiments to assess the DMOMSO algorithm's effectiveness, in terms of its performance on the CEC2017 and CEC2022 benchmark function sets. In this way, we carried out a comparative analysis between the DMOMSO algorithm and seven other algorithms, namely, MSO, SSA, AO, FOX, WOA, HHO, and COA. The experimental results indicated that the DMOMSO algorithm outperforms the original mirage search optimization algorithm, being able to more effectively find optimal solutions and converging faster. The eight considered algorithms were also tested in the context of practical optimization problems, including the design of hydrostatic thrust bearings, the optimization of Himmel Blau's function, and industrial refrigeration system design. Based on the results obtained, DMOMSO is more effective for solving complex engineering problems.