This paper proposes an Improved Hybrid Genetic Particle Swarm Algorithm (IHGPA) to address the limitations of conventional methods in UAV 3D path planning for mountainous environments. IHGPA effectively combines the mechanisms of Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) into a single hybrid architecture. This system employs adaptive parameter control and integrates genetic operation strategies to maintain population diversity and mitigate early convergence. Furthermore, it utilizes a Gaussian-represented terrain modeling approach and a multi-criteria fitness function to simultaneously optimize path safety, trajectory smoothness, and routing efficiency. Experimental results demonstrate that IHGPA achieves superior convergence speed and solution quality, reducing final cost by 6.0% and 2.1% compared to standard PSO and GA, respectively. This work provides an effective reference for autonomous UAV navigation in complex terrains.