본문 바로가기
  • Home

Implementation and Performance Analysis of a Genetic Algorithm for Dynamic Obstacle Avoidance

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2026, 12(4), 19
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : July 17, 2026
  • Accepted : August 21, 2026
  • Published : August 31, 2026

MyounJae Lee 1

1백석대학교

Accredited

ABSTRACT

This study applied a Genetic Algorithm (GA) to improve the autonomous pathfinding and dynamic obstacle avoidance performance of Non-Player Characters (NPCs) in a game environment. To address the trait interference problem caused by the conventional single-gene structure, the genes were designed by separating them into wandering, targeting, and risk-avoidance traits. To verify the performance of this model, experiments were conducted in a Unity 2D environment by setting the dominant gene selection rate, mutation rate, and total learning generations as variables. Through this, the impact of each parameter's variation on the agent's target arrival count and extinction count was comparatively analyzed. This study provides empirical evidence that parameter control of genetic algorithms can maximize the adaptability of NPCs in complex terrains.

Journal Copyright Policy

No CCL information provided

Citation status

* References for papers published after 2025 are currently being built.