We previous proposed a dynamical sequential task allocation and reallocation method to resolve a new task assignment domain: dynamical mobile task allocation, by which tasks can move randomly before they are assigned robots to execute, is the condition in which tasks and robots are allowed to be time-dependent before assigned robots to accomplish the relational tasks. It was demonstrated that our method improved the efficiency for the whole multi-robot coordination system to accomplish all tasks. Moreover, it was more conducive to reduce the numerous computational times and communication costs compared to the existing investigated task assignment methods, while the setting thresholds: large distance threshold and small distance threshold are assigned as a priori values. In this paper, we mainly focus on studying the influence of setting thresholds on our dynamical sequential task allocation and reallocation methodology. Various kinds of large and small distance threshold values and their combination are simulated. The simulation results suggest that it is better to assign both large distance threshold and small distance threshold to a relatively small value according to environment area, the large distance threshold is about one-tenth of environment's length and width, the small distance threshold is below half of large distance threshold, respectively. The results are very significant for us to choose and adjust the values of thresholds to adapt different environment scale.