The term "stochastic dynamical system" has a lot of different meanings. Very loosely speaking, stochastic dynamical system is a system that trajectories behave to be very complicated
We can postulate the randomness directly. Alternatively, a deterministic system may appear random to us because we lack sufficiently complete knowledge of it to enable us to predict its behaviour.
A typical example would be the evolution of financiel products like stock prices or derivatives. These are dynamical systems but are described by so-called Wiener processes, which are described by continuous functions which are nowhere differentiable. The resulting differential equations are not the ones you are used to, but are defined via integral equations in the so-called Ito-calculus. See e.g.