<p>This study focuses on the path planning problem of unmanned aerial vehicles (UAVs) in complex environments, providing effective solutions for challenging terrains. A path planning model with constraints like path cost, height cost, obstacle avoidance, smoothness, and stability is formulated as a multi-constraint optimization function. To enhance the cost efficiency of path planning for UAVs in complex terrains, this paper presents a collaborative optimization framework that integrates an improved differential evolution (DE) algorithm and dual quaternion. Firstly, to tackle the problems of low convergence accuracy and the tendency to fall into local optima in the standard DE algorithm, a fractional DE (FDE) algorithm is devised. This algorithm models the historical memory effect of population iteration via fractional order calculus and combines a dimensionally mutation mechanism to enhance the algorithm’s capability of escaping local optima. Secondly, the dual quaternion interpolation is incorporated to construct the kinematic constraints of UAV, unifying the three-dimensional path coordinates and attitude parameters into a helical motion description to generate smooth and feasible paths. Finally, in six sets of simulation flight tasks, compared with the standard DE algorithm and six other advanced meta-heuristic algorithms, the FDE algorithm achieves the lowest flight cost among the eight algorithms. Compared with other algorithms, the proposed method reduces the flight cost by an average of 39.93%, improves the stability by an average of 50.82%, and only increases the computational time by an average of 0.739&#xa0;s. Furthermore, the Wilcoxon test confirms that the performance advantage is statistically significant. Simulation experiments demonstrate that this framework can still guarantee path feasibility and significantly improve the overall cost efficiency in complex scenarios with threatening obstacles.</p>