Particle Swarm Optimization has been widely used to solve real world problems, mainly when there are too many variables to be optimized and these variables are continuous. In nature one can observe many examples of cooperative behaviors that lead to complex problem solving. Recently, some Particle Swarm Optimization variations gracefully incorporate such cooperative features with consequent beneficial new abilities. In this paper we put forward the incorporation of auto-adaptation capability in a cooperative Particle Swarm Optimization algorithm, called Clan Particle Swarm Optimization. Next, we present a deep analysis on the adaptation process for one multimodal function and evaluate the performance of our proposal in some well known benchmark problems. The results revealed that our proposal achieved better performance than other approaches, specially in tough multimodal problems.