In order to solve the task allocation problem, put forward a particle based ant colony algorithm. Ant colony algorithm-particle swarm optimization algorithm, the mixed algorithm, makes the ant colony have the character of ‘particle’. Build up a model of extended cooperative multi-tasks assignment problem. Using making up state-transition allowed set, the number of infeasible solutions was reduced sharply. Additionally, this algorithm has better performance in solving the task allocation problems, by relieving the problem of prematurity and falling into local optimum. The simulation results show that this algorithm can solve the distributed task allocation problem for multiple UAVs effectively, with faster convergence rate and better results.