Robots and autonomous vehicles, are increasingly utilized for Mars exploration, behind enemy lines reconnaissance, firefighting missions, border control and weather forecast. These vehicles are used usually in missions which are dull, dirty and dangerous. During recent years, being capable of replanning their mission details autonomously when facing unpredictable changes, has become one of the critical features of robots. Among all types of robots, path planning of aerial robots or unmanned aerial vehicles, due to their restrictions such as more influence of non-deterministic issues and necessity to consider differential equations, is more challenging. One of the main parts of designing an autonomous UAV is its path planning, especially when it is designed to perform in unknown, changing and controversial environments. In previous global path planning methods, the current position of moving obstacles is addressed, while in local obstacle avoidance methods, an estimate of future position of moving obstacles is considered. In this thesis, a prospective global path planning is proposed to avoid moving and static obstacles or undesirable regions. The goal of this thesis is to propose a method to find a path for an UAV operating in a 3D environment. In this environment obstacles are divided into two categories; hard and soft obstacle. And being time variant and not complete knowledge is true about the latter type of obstacles. Proposed prospective path planning method is performed better in comparison with its present-based counterpart in term of Soundness.