Dragonfly algorithm is a novel meta-heuristic optimization algorithm and was originally proposed for solving continuous optimization problems. To enhance the optimization performance of the Dragonfly algorithm, this paper proposed an improved Dragonfly algorithm, which was based on elite opposition-based learning strategy and exponential function adaptive steps. The elite individual is introduced to generate their opposite solutions by elite opposition-based learning. This mechanism can expand the scope of the search area and is helpful to improve the global exploration capability of this algorithm. At the same time, an adaptive step with exponential function is designed for replacing the original random step, so it can speed up the convergence rate of the algorithm. 15 typical benchmark test functions are applied to verify the effects of these improvements. The experimental results show that the proposed algorithm has faster convergence speed and higher convergence accuracy than basic dragonfly algorithm and some other meta-heuristic optimization methods.