Title: Self-organised prediction method for potential conflicting vehicles at intersections considering vehicle intrusion

Authors: Lingmin Yang

Addresses: Institute of Traffic Management, Hubei University of Police, Wuhan 430034, China

Abstract: To improve the accuracy of predicting advance time and spatial prediction, this study proposes a self-organising prediction method for potential conflict vehicles at intersections during vehicle intrusion. First, analyse the self-organising network system in vehicles and utilise short-range wireless communication technology between vehicles and roadside units to achieve information transmission. Second, vehicle status information is collected through communication between vehicles and RSU, as well as inter-vehicle communication. Finally, the circular danger range theory is introduced to predict potential conflict vehicles at intersections. Analyse the parameters of the circular danger range and the motion state of vehicles, then calculate the approaching time of danger to accurately predict potential conflict vehicles. Experimental results demonstrate that this method maintains a long lead time (7.7 s) in high traffic scenarios, with prediction accuracy ranging between 0.933 and 0.979 under different traffic flows.

Keywords: intersection; invading vehicles; potential conflicts; conflict prediction; self-organising network; circular danger range.

DOI: 10.1504/IJCAT.2026.154040

International Journal of Computer Applications in Technology, 2026 Vol.78 No.4, pp.305 - 313

Received: 12 Feb 2025
Accepted: 10 Jun 2025
Published online: 10 Jun 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article