Aiming at the problems of low solution efficiency and easy to fall into local optimal values in the course planning of UAVs in three-dimensional complex environment, a real-time track planning method based on the improved sparrow search algorithm (CFSSA) of fusion multi-strategy is proposed. Due to the problems of slow convergence speed and low optimization accuracy of the sparrow search algorithm, firstly, the Circle chaotic mapping is used to initialize the population, the population distribution is more uniform, and the diversity of sparrow population positions is improved. Secondly, the position information of the sparrow is optimized, and the firefly disturbance strategy is used to increase the flexibility and search range of the algorithm, inhibit the premature convergence of the group, and improve the convergence speed and accuracy of the algorithm. Finally, the effectiveness of the CFSSA algorithm is verified by six standard test functions, which can break through the limitation of local optimal solution and obtain higher accuracy. Applying it to solve route planning problems, simulation results show that the CFSSA algorithm is more expensive and has a smaller variance than other algorithms.