Dynamic adaptation of chaotic motion parameters of unmanned aerial vehicles based on terrain complexity analysis

Abstract

Autonomous unmanned aerial vehicle swarms are increasingly deployed in emergency response and search-and-rescue operations. Existing waypoint-based trajectory planners, however, leave systematic coverage gaps in non-convex areas and cannot adapt once conditions change mid-mission. This paper proposes a terrain-aware adaptive chaotic motion algorithm that tunes swarm motion parameters online from measurable geometric properties of the operational area, including the concavity index, obstacle density, narrow corridor coefficient and real-time coverage progress indicators. A closed-loop adaptive controller matches the terrain profile with nine swarm parameters using a linear interpolation-based rule with smoothing. Experimental evaluation based on fifty simulation tests shows that adaptive mode provides fifteen percents higher average area coverage, reducing the spread between runs by nineteen percents, compared to the baseline variant with fixed parameters. Statistical analysis using the Mann–Whitney U test confirms that improvements in regional-level coverage in terms of total distance and energy consumption are highly significant, whereas differences in redundant scanning and obstacle density are not. These results emphasise that the main operational advantages of the adaptive algorithm are reliable and reproducible, while secondary indicators remain stable in all modes. The results confirm that chaotic adaptation with consideration of terrain relief is a fundamental and operationally viable strategy for autonomous territory scanning.