Fuzzy Logic-Based Adaptive Control for LiDAR-Equipped Mobile Robots
6th IEEE International Conference on Smart Information Systems and Technologies, SIST 2026, Astana, Kazakistan, 13 - 15 Mayıs 2026, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/sist61674.2026.11596504
- Basıldığı Şehir: Astana
- Basıldığı Ülke: Kazakistan
- Anahtar Kelimeler: Autonomous Obstacle Avoidance, Fuzzy Logic, LiDAR-Based Perception, Mobile Robot, Risk-Aware Navigation
- Ankara Yıldırım Beyazıt Üniversitesi Adresli: Evet
Özet
Due to the uncertainty of the senses, environmental variability, and the risk-averse nature inherent in traditional control strategies, it is hard for mobile robots to autonomously navigate dynamic, unstructured environments. Most fuzzy navigation methods are based on local obstacle avoidance and use fixed rule sets. However, this increases their navigational adaptability and safety in highrisk environments. An adaptive risk-aware fuzzy navigation system for LiDAR-equipped mobile robots thus fills the gap created by fixed rule sets. This paper introduces a new fuzzy risk index (FRI) that provides an integrated assessment of current risk with respect to regulations and a control strategy that changes in response to real-time input motion or position estimation data. A set of fuzzy rules dynamically parameterizes the control design optimizer based on risk level and features such as obstacle distance, obstacle movement direction, and robot speed. This allows a more safety-oriented approach than traditional methods that set points manually without accounting for hazards at different times or places. It also has a causal structure that interfaces directly with the perception system (LiDAR, SONAR, radar, camera image processing), where fuzzy logic and fuzzy controllers are used to implement dynamic control laws, thereby increasing the overall flexibility and comprehensiveness of the navigation system. The proposed system adopts a tiered architecture, separating the perception system, including risk assessment and motion control, making it more modular and comprehensible for navigation. Simulation experiments show that compared to the traditional fuzzy approach, the proposed method significantly reduces collision rates and improves navigation stability, especially in constrained highrisk environments. These results confirm that fuzzy navigation algorithms, which explicitly account for the uncertain risk factors encountered by autonomous mobile robots on a daily basis, can enhance their safety and reliability under such unfavorable conditions.