To improve the reliability and safety of unmanned aerial vehicles(UAV)flying in complex mountainous environments,an improved snake optimization algorithm for UAV path planning is proposed. Firstly,a drone environment model and a mountain peak threat model were constructed by combining digital elevation information and complex terrain threats.Secondly,an improved snake optimization algorithm is proposed,which integrates traditional snake optimization algorithms with cellular automata for unmanned aerial vehicle path planning.The niche technology and optimal local jitter are introduced to avoid the algorithm falling into local optima and improve global search ability. Finally,simulation experiments are carried out in three scenarios to verify the effectiveness of the proposed method.The experimental results show that the average path length of the proposed method is 2.201,1.801 and 2.187 km respectively in three complex scenarios,and the average convergence time is 14.8,13.9 and 14.9 s.Compared with other path planning algorithms,the generated path has good practical effect on real UAV operation.