The paper present search stategies based on centroidal voronoi partitions(CVP) for uninhabited aerial
vehicles(UAVs) under such dynamic environments that the targets are uniformly distributed and appear randomly over time
in a bounded region. Firstly, the CVP model is presented, on the basis of that, the search space division algorithm is given and
the convergence of the algorithm is proved. Then the multi-UAVs search strategy under dynamic environment is designed
and the effection of UAVs turning radius to algorithm performance is also considered. Finally, the simulation results show
that CVP search algorithm is effective to search dynamic targets, and the algorithm has good adaptivity.