Autonomous navigation in dynamic environments remains a challenging problem due to unpredictable obstacle motion and the limited predictive capability of classical reactive algorithms. The Vector Field Histogram Plus (VFH+) method is widely used for real-time local obstacle avoidance; however, it does not explicitly account for future obstacle states, which may lead to abrupt maneuvers and reduced trajectory smoothness. This paper proposes an adaptive predictive extension of the classical VFH framework designed for safe and smooth navigation in dynamic scenarios. The approach integrates a Constant Acceleration Kalman Filter for short-term motion prediction of dynamic obstacles, an adaptive exponential obstacle magnitude formulation for risk representation in the polar histogram, and a temp

Алсайед Н. (науч. рук. Краснов А.Ю.) An adaptive extension of vector field histogram for local navigation in dynamic environments // Сборник тезисов докладов конгресса молодых ученых. Электронное издание. – СПб: Университет ИТМО, [2026]. URL: https://kmu.itmo.ru/digests/article/16719