Airborne sense and alert collision warning and avoidance system
The emergence of unmanned aerial vehicles and the ever increasing performance in terms of speed, flight altitude, and endurance have led the debate into allowing such aircraft to fly co-jointly with civilian aircraft. However, due to the absence of the usual flight regulation channel, not to mention the pilot at the aircraft controls, the need for an automated collision avoidance system has arisen in the past few years.
For that we propose to provide a practical solution to equip UAVs with an autonomous sense and avoid capability and an autonomous collision avoidance system, to enable the UAV to fly in a non segregated air space safely and meeting the above regulations.
In this research, we have evaluated different types of mechanism to form this collision avoidance system. The most successful of which concluded of path estimation and the calculation of nearest point of approach. During this research, we developed a collision avoidance mechanism that uses vector algorithms and path estimation methods to increase the efficiency of the logic system and decrease the computation time.
In the results of this experiment, we determined that using few well developed manoeuvres would result in better avoidance efficiency and would require limited change to the flight path of the unmanned vehicle and flight parameters.
Manoeuvres such as changing speed or turning provided the best options for avoiding incoming aircraft, while changing altitude was less successful due to the danger of flying into a different flight level (sharing the same altitude levels with other aircraft) and due to the limited climb and decent rate performances of the model unmanned aerial vehicle used.
More complicated scenarios, such as avoiding multiple aircraft would require a slightly different strategy, where the algorithm would be based upon avoiding a flight path of all aircraft at all times, rather than changing velocity or heading to avoid colliding at a certain point in time.
The main outcome of this experiment, was to prove that such algorithms (with limited complex theory behind it) can prove to be a good option for deriving avoidance systems and ensuring flight safety for manned and unmanned aircraft.
Testing was successfully conducted by the student on simple implementation of the Loss of Separation algorithm to verify, test and expand the algorithm as a pre-preparation for code integration in later stages.
In this document, we present how we will implement the “sense and avoid” algorithm and the logical decision making system that would provide the UAV with the ability to re-route its current path to a safer flight course.