Abstract: To address the issues of insufficient diversity and difficulty in escaping from local optima during the optimization process of the multi-objective wolf pack algorithm, we propose a clustering multi-objective wolf pack algorithm for constrained optimization problems (CMOWPA-C). Firstly, by integrating a self-adaptive penalty model and a self-adaptive tradeoff model, we propose a new method to convert constrained problems into unconstrained ones. Secondly, we introduce a random disturbance factor to optimize the moving step of the population and prevent it from falling into local optima. Finally, we employ the K-means clustering algorithm to group the population. The population is divided into different clusters based on the distance between individuals and the cluster centers, ensuring that individuals are associated with each cluster center, thereby enhancing population diversity. To verify the algorithm performance, we compare CMOWPA-C with nine emerging algorithms on benchmark problems and with nine constrained multi-objective evolutionary algorithms on practical constrained problems. The results show that CMOWPA-C significantly improves diversity and effectively avoids falling into local optima.