Cooperative sUAV Collision Avoidance Based on Satisficing Theory

Abstract

The recent growth in small unmanned aerial vehicle (sUAV) applications in public and private sectors has generated a greater interest in effective collision avoidance strategies. This paper presents a cooperative collision avoidance approach for sUAV in the low-altitude uncontrolled airspace based on satisficing game theory. By incorporating both the self and cooperative preferences in making individual heading decision, satisficing framework provides a collision avoidance strategy that increases throughput while decreasing unnecessary collisions. A total of 4356 chokepoint scenarios are simulated with varying satisficing parameters, including number of sUAV agents, minimum separation, action angle, and dual utility parameters of raw preference and negotiation index. Simulation-based sensitivity analysis is conducted to find the combined effect of such parameters. The results show that performance of satisficing framework is dependent on traffic density primarily. Except for the very light traffic scenario, the minimum separation and degree of action angle affected the system efficiency the most, while dual utility parameters played crucial roles in highly conflicted scenarios. In conclusion, sUAV traffic and collision rules require an adaptive approach in regard to the traffic density, and satisficing framework can provide an effective collision avoidance strategy in a medium- to high-density traffic environment.