A Novel Collision Detection and Avoidance system for Midvehicle using Offset-based Curvilinear Motion

TL;DR

This paper introduces MCDAS, a new collision detection and avoidance system for midvehicles that uses offset-based curvilinear motion and constant velocity models to prevent rear-end and side collisions in dense traffic.

Contribution

The paper presents a novel collision avoidance algorithm that effectively handles rear-end and parallel parking scenarios using offset-based curvilinear motion and real-time data validation.

Findings

Effective collision avoidance in simulated dense traffic conditions.

Consistent performance demonstrated through Monte Carlo analysis.

Successful validation with real-time vehicle data.

Abstract

Major cause of midvehicle collision is due to the distraction of drivers in both the Front and rear-end vehicle witnessed in dense traffic and high speed road conditions. In view of this scenario, a crash detection and collision avoidance algorithm coined as Midvehicle Collision Detection and Avoidance System (MCDAS) is proposed to evade the possible crash at both ends of the host vehicle. The method based upon Constant Velocity (CV) model specifically, addresses two scenarios, the first scenario encompasses two sub-scenario namely, a) A rear-end collision avoidance mechanism that accelerates the host vehicle under no front-end vehicle condition and b) Curvilinear motion based on front and host vehicle offset (position), whilst, the other scenario deals with parallel parking issues. The offset based curvilinear motion of the host vehicle plays a vital role in threat avoidance from the front-end vehicle. A desired curvilinear strategy on left and right sides is achieved by the host vehicle with concern of possible CV to avoid both end collisions. In this methodology, path constraint is applicable for both scenarios with required direction. Monte Carlo analysis of MCDAS covering vehicle kinematics demonstrated acute discrimination with consistent performance for the collision validated on simulated with real-time data.