This study proposes an algorithm designed to address Constrained Multi-Objective Optimization Problems (CMOPs) within the framework of Multi-Objective Particle Swarm Optimization (MOPSO). The method integrates a decomposition strategy with specialized mechanisms for constraint management. A central component of the approach is an adaptive penalty function, derived from the annealing penalty method, yet reformulated to induce a linear interaction between feasible and infeasible solutions during the search process. As a result of this integration, the capabilities of the dMOPSO algorithm are significantly extended, yielding a novel variant referred to as CdMOPSO-Annealing Penalty (CdMOPSO-AP). The proposed method was rigorously assessed using the CF benchmark suite from the CEC’2009 competition. Comparative analyses, based on convergence and diversity performance indicators, demonstrate that CdMOPSO-AP exhibits competitive efficiency, frequently surpassing four state-of-the-art algorithms considered in the evaluation.