Selecting optimized action strategy can make the task coalition evolve towards the expected operation effect.Considering that the influence strength of events/actions is not consistent in different time horizons, a time-varying Dynamic Influence Nets (DINs) was utilized to model the action strategy optimization in the process of task coalition evolution.Based on causal strength logic, the consistent condition of probability propagation was given and the influence constant was calculated out.Simulation was made to an aerial attack campaign case.The result shows that the action strategy optimization considering the variation of influence constants can improve the capability of cause and effect modeling.