This paper represents a novel smooth trajectory planning algorithm which uses the natural behavior model of a mobile robot (MR). The shortest path in the free configuration space is obtained by using the visibility graph method. It is modified according to dynamic constraints which are implicitly included in natural behavior of the mobile robot. The modified path becomes a smooth, easily tractable time-optimal trajectory. For every points of it, translating/steering velocities and accelerations and reaching times are known. It is applicable to the real time dynamic configuration spaces, because of simplicity and low computational time.