In order to improve differential evolution (DE) algorithm's global search ability during later periods, annealing idea was merged into the strategy of selecting DE's amplification factor. Based on the good micro search ability of chaos, the population started adaptive chaotic mutation when it failed into premature. Through comparing this proposed hybrid strategy with original and another improved strategy of DE respectively based sixteen benchmark functions, simulation results show that the proposed hybrid algorithm has not only better ability of departing from local extremes in later periods but also has improved performance of DE in other important aspects simultaneously.