Flight Test Validation of Adaptive Collision Avoidance Algorithms Using Multiple Unmanned Aircraft

Abstract

With the advent of the unmanned aerial systems (UAS) era, it is important that research be conducted into the reliable and safe operation of UAS in urban areas, where aircraft are flying over people and property in spatially constrained environments. A large focus in recent years has been on adaptable collision avoidance systems capable of avoiding fixed and airborne traffic. This paper presents flight test validation of an adaptive collision algorithm known as the morphing potential field algorithm. The morphing potential field algorithm has been modified to allow implementation of more advanced guidance logics such as $L_{2}^{+}$ guidance. The aim is to validate the morphing collision avoidance path planning algorithm in a complex scenario, with multiple aircraft–two fixed wing and two rotary wing–flying the same spatially constrained area. The aircraft were intentionally put in a collision course with each other and other obstacles. Validation flight tests were successfully conducted to assess performance of morphing potential field navigation algorithms and also to quantify the impact of possible communication delays on the safety of collision avoidance methods for high speed and high inertia aircraft.