Abstract: In order to improve the superiority of monarch butterfly optimization algorithm in dealing with complex problems and avoid the problem of local optimal solution, an enhanced monarch butterfly optimization algorithm based on multiple strategies is proposed. In the first step, non-linear computation based on normal cloud generation is applied to single-parent Monarch butterflies to expand the potential solution set and enhance their local exploration efficiency. Secondly, the optimal cloud-individuals are selected by greedy strategy. Finally, adaptive mechanisms are used to make the population more diverse. Through the comparative analysis of optimization of 8 benchmark test functions and Friedman test result, it can be seen that the improved algorithm has better convergence and stability.