<p>The Hiking Optimization Algorithm (HOA) is a meta-heuristic approach based on the mathematical modeling of Tobler’s hiking function. It has gained attention owing to its competitive performance and has been widely applied in various fields. However, the original HOA method suffers from slow convergence, low stochasticity, insufficient population diversity, and imbalanced exploration and exploitation. Therefore, this study proposes an MHOA to improve its performance. By introducing a roulette wheel strategy to enhance the global search capability, designing a dual mapping strategy of Logistic mapping and Tent mapping perturbation factor to increase the population diversity, and adopting a flexible position update strategy to improve the convergence accuracy and speed. The performance of the MHOA was evaluated using the CEC2022 and CEC2017 testing functions and various metrics. The Wilcoxon signed rank test results showed that MHOA exhibited significant win rates in complex optimization tasks in both CEC2022 (<i>D</i>10: 56/13/3, <i>D</i>20: 57/9/6) and CEC2017 (<i>D</i>30: 148/18/8, <i>D</i>50: 149/16/9, <i>D</i>100: 156/6/12), verifying its efficiency and robustness under high dimensions. Efficiency and robustness under it tops the list for solving the reducer design, robot gripper design, Himmelblau nonlinearity, and welded beam design. Overall, the MHOA achieves a good balance between exploration and development, outperforms other metaheuristics, and provides a better alternative for solving optimization problems.</p>