Three-Dimensional Route Planning for a Cruise Missile for Minimal Detection by Observer.

摘要

We present an algorithm for finding optimal three-dimensional paths above polyhedral models of terrain using a technique we refer to a 'random-ray'. Contiguous sequences of homogeneous airspace volumes are generated using constraints of probability-of-detection and aerodynamic flight models. The flight costs are calculated as in actual mission planning using time, distance, airspeed, and fuel flow. We then try semi-random directions (rays) from the starting point, turning in accordance with Snell's Law at maneuver points (points between volumes). If we ever do not enter the previously specified next volume, we make random adjustments to the ray (in, out, up, or down) with respect to the center of the facet between the two volumes, until either the path will enter the correct next volume or we determine it is impossible. The performance of random-ray technique is an improvement over an earlier approach using local optimization. We have also implemented a moveable display on a graphics workstation, to allow the user the ability to view the terrain and paths from any angle. Keywords: Evasion routing planning; Artificial intelligence; LISP; Three-dimensional searching path planning; Theses. (EDC)