Aiming at the difficult problems such as high real-time performance and strong covert of generated trajectory requirements faced by unmanned aerial vehicle (UAV) trajectory planning algorithms in complex environments, an adaptive nonlinear convergent global sparrow search algorithm (ANC-GSSA) is proposed in this paper to deal with the complex trajectory planning problem. Through Latin hypercube sampling, the problem of low population diversity is solved at the early stage of the algorithm, and through the design of a nonlinear convergence factor and adaptive scouting strategy to maintain the balance between population diversity and convergence speed, to accelerate the convergence speed while ensuring the global optimality of the generated trajectory. Based on ANC-GSSA, the UAV trajectory planning fitness function is designed with the objective of the shortest path by considering the terrain, manoeuvre, and other constraints to ensure the feasibility of the planning results. Finally, ANC-GSSA is compared with other algorithms by numerical comparison simulation to verify the superiority of the proposed algorithm, and the algorithm is designed to simulate the complex environment and applied to UAV trajectory planning to ascertain the effectiveness of the algorithm.