In order to improve the speed and accuracy of firefly algorithm, a parameter variance adjustment firefly algorithm is proposed. Firstly, based on the analysis of the firefly algorithm the core idea of parameter variance adjustment algorithm firefly is developed: calculate the population variance of luminance assessment of convergence and divergence of populations, and adjust parameters according to the processing, which aims to improve the firefly algorithm, and then presents the algorithm implementation steps and processes. Secondly, compare and analysis the parameter variance adjustment firefly algorithm with basic firefly algorithm, genetic algorithm and particle swarm optimization with four optimization test functions, and find that the parameter variance adjustment firefly algorithm can quickly find the solutions which meet the accuracy requirements of the test, while the success rate is 100%. Compared with other algorithms, it has a clear advantage and good stability. Finally, the parameter variance adjustment firefly algorithm is applied to a real backgrounds and geographical parameters of three-dimension path planning simulation experiments by constructing the objective function to calculate the energy and it meet the required path planning in the three-dimensional ocean environment, which shows the validity of the parameter variance adjustment firefly algorithm.