Aiming at the defect that the Kepler Optimization Algorithm (KOA) is prone to falling into local optima in high-dimensional spaces, this paper proposes an improved KOA (IKOA) based on hybrid strategies. Firstly, the adaptive Levy flight strategy is introduced into the position formula of the distance between planets and the sun. Secondly, the adaptive Brownian motion fused with the golden sine algorithm is incorporated into the formula for updating planetary positions. Then, the dynamic random centroid opposition-based learning strategy is used to replace the original elite strategy. Finally, a simulation comparison is conducted between the IKOA and the original KOA on 12 test functions from the CEC2022 test suite. The simulation results show that IKOA has stronger convergence speed and global optimal optimization performance.