This article addresses the dynamic task allocation problem with limited communication and velocity. The main challenge lie in the selection of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula> fittest winner participants and the participant contention that one winner participant may be selected by multiple targets simultaneously. Existing methods take the distance between the targets and participants as the evaluation index to select winners, which may lead to futile selection since the winner participant locating at the opposite direction of the target cannot intercept the target with limited velocity. By carefully considering both the distance between the targets and participants and the motion direction of the targets, an improved evaluation index for each target is proposed and employed such that the futile selection can be avoided in the executing process of the algorithm. Moreover, an extra evaluation index for each winner participant is presented to select one winner target to overcome the participant contention. Based on these, the control protocols are developed for targets interception, and their stability is proven by the Lyapunov theory under some suitable conditions. Finally, simulation examples are presented to illustrate the effectiveness and advantages of the proposed algorithms.