In the last years, collective motions have fascinated and attracted the attention of the scientific community. Indeed, this type of phenomena is ubiquitous in nature and often it is of rare beauty, just think about the dance of a large swarm of starlings during last hours of the day. It is also impressive how ordered patterns and coherent behaviour emerge from the interactions among individuals. In this Thesis, we propose a new microscopic model based on a system of interacting stochastic differential equations of Langevin-type for describing the collective dynamics of a self-propelled particle swarm. The concept of self-propelled particles is of crucial importance in modeling the behaviour of animal groups or of cell populations.