Research on Longitudinal Intelligent Control of Autonomous-rail Rapid Tram

Abstract

Aiming at the intelligent operation requirements of autonomous-rail rapid tram, this paper proposes a longitudinal intelligent control algorithm based on the driver behavior habit model. Firstly, it analyzes the ART operation scenarios and environmental conditions and introduces the multi-sensor perception fusion technology used to guarantee the 360° obstacle recognition without blind spots in the surrounding environment of trains and high-precision train positioning. And then, it presents the multi-mode division of longitudinal intelligent control designed in the finite-state machine (FSM) pattern and from simulation of drivers' behavior characteristics, and proposes the intelligent longitudinal control strategy and algorithm to enable the intelligent train operation control functions, such as autonomous following, automatic operation at traffic light intersections, automatic deceleration on curves, and parking at stations in high-precision alignment. According to the simulation results, using the developed longitudinal intelligent control algorithm, the test train can depart, accelerate, decelerate and stop stably following the train ahead, and automatically decelerate on curves and run passing traffic light intersections, with a speed control accuracy within ± 1 km / h, impact rate less than 0.75 m/s3, and alignment error for parking less than ± 0.21 m, demonstrating the effectiveness to reduce driving intensity for drivers, and improve the train running comfort and operating efficiency.

Keywords

• autonomous-rail rapid tram
• intelligent driving
• precise parking
• speed control
• adaptive cruise control