The seasons optimization (SO) algorithm is an evolutionary-based optimizer designed to solve numerical and engineering optimization tasks. Despite its promising performance, the SO algorithm has two significant drawbacks: slow convergence and the tendency to become trapped in local optima in certain problems. To address these limitations, we proposed an adaptive version of the SO algorithm named the seasons optimization with chaotic reverse learning and opposition-based learning (SOCO) algorithm. The proposed SOCO incorporates two learning strategies: chaotic reverse learning (CRL) and opposition-based learning (OBL). These techniques establish a proper balance between exploration and exploitation, allowing for an efficient search of the solution space and facilitating escapes from local optima, ultimately enhancing the algorithm’s convergence quality. We evaluated the algorithm using three engineering design problems and 20 numerical functions of varying dimensionalities taken from test suites CEC-2015, CEC-2018, and CEC-2022. The simulation results demonstrate that the SOCO algorithm outperformed its counterparts in terms of global search capability and convergence performance.
