In order to solve the problems that the traditional Firefly Algorithm （FA） is easy to fall into local optimum and has low convergence speed， an improved FA based on multi-strategy fusion， named LEEFA （Levy flight-Elite participated crossover-Elite opposition-based learning Firefly Algorithm） was proposed after integrating Levy flight， elite participated crossover operator and elite opposition-based learning mechanism in the firefly optimization algorithm. Firstly， Levy flight was introduced based on the traditional FA， so that the global search ability of the algorithm was improved. Secondly， an elite participated crossover operator was proposed to improve the convergence speed and accuracy of the algorithm， as well as to enhance the diversity and quality of solutions in the iterative process. Finally， the elite opposition-based learning mechanism was combined to search for the optimal solution， which improved the ability of jumping out of local optimum and convergence performance of FA， and realized the rapid exploration of solution search space. In order to verify the effectiveness of the proposed algorithm， simulation experiments were carried out on the benchmark functions. The results show that compared with algorithms such as Particle Swarm Optimization （PSO） algorithm， traditional FA， Levy Flight Firefly Algorithm （LFFA）， Levy flight and Mutation operator based Firefly Algorithm （LMFA） and ADaptive logarithmic spiral-Levy Improved Firefly Algorithm （ADIFA）， the proposed algorithm performs better in both convergence speed and accuracy.