Fuzzy Logic In Continual Learning For Autonomous Vehicle Adaptation To Road Conditions
Gundala Swarnalatha,
Mohd Ishaq,
Dr. Tata Sivaiah
Autonomous vehicles must adapt to rapidly changing road conditions to ensure safe and efficient
operation. This research investigates the integration of fuzzy logic into continual learning frameworks
to improve vehicle adaptation to dynamic environments, such as varying road surfaces and weather
conditions. Fuzzy logic provides a means of handling uncertainties, allowing for more flexible decisionmaking
in real-time. By comparing Fuzzy Logic-based Continual Learning (FLCL) with Standard
Continual Learning (SCL) across key performance metrics—speed adjustment, steering correction, and
braking response—under different road conditions (dry, wet, and icy), the study demonstrates that FLCL
offers smoother, more precise adjustments. Results show that FLCL provides more controlled vehicle
speed management, finer steering corrections, and faster braking responses, significantly enhancing the
vehicle’s ability to handle uncertain road conditions. This work highlights the potential of fuzzy logic in
continual learning systems to improve safety and adaptability in autonomous driving.
