To address the issue that it is difficult to fully find the equivalent Pareto optimal solution sets in imbalanced multimodal multi-objective optimization, this paper proposes a multi-stage multi-objective evolutionary algorithm driven by an exclusion mechanism. The algorithm divides the entire evolutionary process into three stages and adopts different environmental selection methods in each stage, thereby realizing the optimization task of searching for different equivalent Pareto optimal solution sets in different evolutionary stages. Specifically, the environmental selection method in the first stage only considers the convergence of individuals in the objective space, enabling the population to quickly converge to the most easily found equivalent Pareto optimal solution set. The environmental selection method in the second stage employs a search strategy based on the exclusion mechanism, which adaptively penalizes individuals close to the already found equivalent Pareto optimal solution sets. This helps avoid the population from searching in duplicate regions and reduces the risk of the algorithm getting trapped in a single equivalent Pareto optimal solution set. The environmental selection method in the third stage fine-tunes the equivalent Pareto optimal solution sets obtained in the first two stages by simultaneously considering the convergence and diversity of individuals in both the objective space and decision space, so as to further improve the optimization performance of the algorithm. Experimental results show that the proposed algorithm can find all equivalent Pareto optimal solution sets under the condition of the same number of function evaluations, and compared with seven other similar algorithms, it has certain advantages in terms of comprehensive performance in both the objective space and decision space.